Surg Endosc (1997) 11: 10061009

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

Combination of subcutaneous abdominal wall retraction and optical trocar to minimize pneumoperitoneum-related effects and needle and trocar injuries in laparoscopic surgery
L. Angelini, M. M. Lirici, V. Papaspyropoulos, F. L. Sossi
4th Department of Surgery, Policlinico Umberto I, University La Sapienza, Viale del Policlinico, 155, 00161 Rome, Italy Received: 11 January 1996/Accepted: 10 February 1997

Abstract Background: Both pneumoperitoneum and blind needle and trocar insertion may cause complications: because of the well-known physiological effects, CO2 insufflation is not indicated in patients with impairment of cardiorespiratory function and high-risk patients; injuries to underlying viscera and vessels by needles and trocars have been reported even when the open technique is used. Methods: A technique which combines abdominal wall suspension by a new subcutaneous lifter (LaparoTenser) and optical trocar (OptiView) insertion has been evaluated in a random series of 22 patients undergoing various laparoscopic procedures. The optic trocar was inserted without previous insufflation, but low-pressure (15 mmHg) pneumoperitoneum was associated during the course of the procedure in 16 cases. Results: The exposure of the operating field was good or sufficient in 21 cases (95%), while the placement of the optical trocar was always safe. One complication related to the insertion of the subcutaneous needles of the wall lifter occurred (suprafascial hematoma). Conclusions: The subcutaneous retractor allows the use of conventional cannulae and the combination of abdominal wall suspension with or without low-pressure pneumoperitoneum, thus enhancing the quality of exposure with no effect on the hemodynamic and respiratory functions. Key words: Gasless laparoscopy Subcutaneous abdominal wall retraction Optical trocar Trocar and needle injuries

Both pneumoperitoneum and blind needle and trocar insertion may cause complications. Pathophysiology of gas laparoscopy has been well described (reduction of pulmonary function, splanchnic vasoconstriction, increased physiological dead space, ventilation-perfusion mismatch, increase of total peripheral resistance, increased pulmonary wedge pressure, effects on cardiac output, rise in PaCO2) [2]. Whereas in young and healthy people these changes are well tolerated, in patients with impairment of cardiorespiratory functions and in high-risk patients gas-less laparoscopy may be the best option to accomplish surgical procedures by the endoscopic approach. The exposure of the operating field achieved with most abdominal lifting systems is unfortunately poor because of the tenting effect. On the other site, Veress needle and trocars inserted with the conventional blind technique often cause minor or major injuries such as bleeding from the abdominal wall, bleeding from major abdominal vessels, and visceral lesions [1, 36, 8]. Injuries may occur even when the open laparoscopy technique is employed [7]. Furthermore, bleeding from the abdominal wall increases the risk of wound infection. The combination of mechanical exposure by a new subcutaneous abdominal wall retractor (LaparoTenser, LT Lucini, Milano, Italy) and endoscopic guided insertion of optical cannulae (OptiView, Ethicon EndoSurgery, Cincinnati, USA) has been employed in a random series of 22 patients undergoing various laparoscopic procedures, with the aim to minimize the risks of needle/trocar injuries and highpressure pneumoperitoneum.

Materials and methods

LaparoTenser is a new abdominal lifter provided with specially designed subcutaneous needles (PluriPlan) built according to a mathematical model to distribute forces along all their length for a smooth and balanced suspension (Figs. 1 and 2). The exposure achieved is that of a frustum of a cone. The system is compatible with CO2 insufflation and does not required special cannulae or open laparoscopy technique. The trocar OptiView is shown in Fig. 3.

Correspondence to: M. M. Lirici

1007

Fig. 1. LaparoTenser (LT Lucini, Milan, Italy) is an abdominal wall lifter provided with subcutaneous needles which is attached to the operating table and features micro- and macrometric adjustment of the wall suspension. Fig. 2. Special subcutaneous needles (PluriPlan) are designed according to a mathematical model to provide an intraabdominal space with the shape of a frustum of a cone, and deflect if dangerous forces are applied. Fig. 3. OptiView (Ethicon EndoSurgery, Cincinnati, USA) consists of a conventional 10/12-mm disposable cannula and a pistol-grip plastic trocar with a transparent tip, into which the laparoscope is inserted. Fig. 4. Positioning of the access ports and subcutaneous needles for laparoscopic cholecystectomy. Fig. 5. Positioning of the access ports and subcutaneous needles for gynecological procedures (A) and laparoscopic hernia repair (B). Fig. 6. With the abdominal wall lifted up by the LaparoTenser retractor, the OptiView is inserted in the left hypocondrium under visual control with gentle rotating movements and pointing the cannula toward the midline. Fig. 7. The abdominal wall layers are visualized step by step during the insertion of the OptiView: (1) subcutaneous fat, (2) anterior fascia incised, with underlying muscle fibers, (3) last muscle fibers divided and the transparency of the peritoneal leaf.

The technique described here has been employed in 22 cases (12 cholecystectomies, five hysterectomies, one ovarectomy, one TAPP hernia repair [unilateral], one ligature of spermatic vein for left varicocele, one repositioning of inflatable gastric banding, one right colectomy [Table 1]). The position of the subcutaneous suspension needles and that of cannulae is shown in Figs. 4 and 5. In the case of cholecystectomy the points of PluriPlan needle insertion are in the right hypocondrium just below the costal margin. For all procedures in the inferior abdominal cavity, the subcutaneous needles are inserted right above the pubis. The trocar OptiView is inserted in the left hypocondrium for laparoscopic cholecystectomy (Fig. 6) and at the navel site in all other procedures. The quality of exposure achieved during the surgical procedures and the safety of the technique and its advantages and disadvantages have been evaluated.

Results Results are shown in Table 1. The exposure of the operating field was good or sufficient in 21 cases (95%). The placement of the optical trocar was always safe. Low-pressure CO2 insufflation improved exposure and allowed total exploration of the peritoneal cavity in 16 cases (maximum pressure 5 mmHg). Gas insufflation has always started after placement of all operating cannulae. The exposure of the operating field at the site of the anastomosis during right colectomy was insufficient. In this

1008 Table 1. Quality of exposure and ease and safety of trocar insertiona Procedure Laparoscopic cholecystectomy 1. P. S. 2. A. M. T. 3. A. S. 4. C. P. 5. S. C. 6. A. G. 7. A. F. M. 8. C. I. 9. M. L. 10. M. V. P. 11. N. F. 12. A. O. LAVH & laparoscopic semm hysterectomy 1. P. T. 2. C. P. 3. D. M. 4. A. M. P. 5. A. F. Laparoscopic Ovarectomy 1. M. P. P. Laparoscopic TAPP hernia repair 1. G. T. Laparoscopic ligature of spermatic veins 1. G. C. Laparoscopic gastric banding repositioning 1. L. M. Laparoscopic right colectomy 1. A. S.
a

Exposure Sufficient Good + CO2 insufflation Sufficient Good + CO2 insufflation Good + CO2 insufflation Sufficient Sufficient Good Sufficient Sufficient + CO2 insufflation Sufficient + CO2 insufflation Sufficient + CO2 insufflation Sufficient + CO2 insufflation Good + CO2 insufflation Sufficient + CO2 insufflation Good + Co2 insufflation Good + CO2 insufflation Good + CO2 insufflation Good + CO2 insufflation Good + CO2 insufflation Good + CO2 insufflation Insufficient + CO2 insufflation

Trocar insertion Easy and safe Easy and safe Easy and safe Little problems, safe Little problems, safe Easy and safe Easy and safe Little problems, safe Easy and safe Easy and safe Easy and safe Easy and safe Easy and safe Easy and safe Some resistance, safe Some resistance, safe Some resistance, safe Easy and safe Easy and safe Some resistance, safe Easy and safe Easy and safe

The procedure performed and the initials of the patients are reported in column 1. In 16 cases lowpressure pneumoperitoneum (15 mmHg) has been combined with abdominal wall suspension. In cases where little problems were encountered during trocar placement, the access into the peritoneal cavity was initially uncertain. Table 2. Trocar and needle injuries and reference authors who reported them in the last 5 years Injury Bleeding abdominal wall Veress needle Trocar Reference article Fitzgibbons [3] Hulka [4] Fitzgibbons [3] Hurd [5] Soderstrom [8] Fitzgibbons [3] Hulka [4] Hurd [5] Apelgren [1] Fitzgibbons [3] Hurd [5] Soderstrom [8] Fitzgibbons [3] Hulka [4] Reich [6] Soderstrom [8] Reich [6] Soderstrom [8] Sadeghi-Nejad [7]

case (a patient with Crohns disease) the operation was converted because it wasnt possible to bring the ileum up to the transverse colon because of its extensive posterior attachments. One complication related to the described technique occurred in the last cholecystectomy: a parietal hematoma of the left flank likely caused by the subcutaneous needle. This operation was converted because of the severe adhesions due to a previous operation on the stomach, and because it was impossible to carry out a safe dissection. In three cases the access into the peritoneal cavity was uncertain. This occurred in the first patients of this series who underwent laparoscopic cholecystectomies, and was mainly related to the different direction which has to be followed while inserting the trocar with the abdominal wall lifted up: the puncture-cannula must not be pointed downward but toward the midline. Changing the angle of insertion allowed the visualization of the peritoneal cavity in all cases. Discussion The combination of subcutaneous abdominal wall retraction and optical trocars effectively avoids pneumoperitoneum and trocar and needle injuries in laparoscopic surgery. The latter have been reported by several authors in recent years (Table 2). LaparoTenser may cause some space constraints, hindering some movements of the working instruments. To avoid this, the position of cannulae has to be slightly changed.

Bleeding major vessels Veress needle Trocar

Visceral injury Veress needle

Trocar Open laparoscopy technique

The best exposure is achieved by gas insufflation with peritoneal pressures up to 15 mmHg. Nevertheless, CO2 insufflation must be avoided in some patients who may have maximum benefit from abdominal wall suspension. The LaparoTenser may also be employed with or with-

1009

out gas insufflation, and it has been successfully used during laparoscopic assisted vaginal hysterectomy, allowing endoscopic guidance of the vaginal step of the procedure. The insertion of the optical trocar has been always safe, with precise visual identification of the wall layers (Fig. 7). Because of its limited sharpness, when inserted at the navel site, the trocar OptiView should be pointed slightly at the side of the navel to avoid the midline, which often offers a certain resistance. References
1. Apelgren KN, Scheeres DE (1994) Aortic injury. Surg Endosc 8: 689 691

2. Baxter JN, ODwyer PJ (1995) Pathophysiology of laparoscopy. Br J Surg 82: 12 3. Fitzgibbons RJ, Annibali R, Litke BS (1993) Gallbladder and gallstone removal: open versus closed laparoscopy, and pneumoperitoneum. Am J Surg 165: 497504 4. Hulka J, Reich H (1994) Textbook of laparoscopy. WB Saunders, Philadelphia, PA, pp 339345, 353359 5. Hurd WW, Pearl ML, DeLancey JOL, Quint EH, Garnett B, Bude RO (1993) Laparoscopic injury of abdominal wall vessels: a report of three cases. Obstet Gynecol 82(4): 2 6. Reich H (1992) Laparoscopic bowel injury. Surg Laparosc Endosc 2(1): 7478 7. Sadeghi-Nejad H, Kavoussi L, Peters C (1994) Bowel injury in open technique laparoscopic cannula placement. Urology 43(4): 559560 8. Soderstrom RM, Levinson C, Levy B (1993) Complications of operative laparoscopy. In: Soderstrom RM (ed) Operative laparoscopy: the Masters technique. Raven Press, New York

Surg Endosc (1997) 11: 1058

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

The author replies

We are grateful for the important points which Dr. Sungler and Dr. Boeckl raise in relation to the interim results of the EAES Ductal Stone Trial. They are, of course, right in that centers with highly specialized expertise including their own have reported successful endoscopic extraction rates of 95% and more. However, Dr. Sungler and Boeckl are mistaking efficacy for effectiveness. Efficacy reflects the best possible results that can be obtained in major centers by a few individuals who have extensive experience and expertise in the procedure, whereas effectiveness reflects the results when the same procedure is practiced at large. In this respect the overall effectiveness of endoscopic stone extraction is well below 95% and in the UK NHS hospitals it averages 7589%. The problem cases which account for most of the failures are large occluding stones. In the EAES study, ERCP was normal in 23% of patients randomized to group A (two-stage management). As Dr. Sungler and Dr. Boeckl comment, this is in fact on the low side and is probably explained by the fact that 60% of the patients had either clinical jaundice or elevated serum bilirubin. A higher normal ERCP rate in patients at risk of harboring ductal calculi would in fact be an additional argument for single-stage laparoscopic treatment on cost benefit grounds. The suggestion that some of the stones in group B were air bubbles is unfair and is refuted by the data collected in the trial. The higher conversion rate in group B is due to the decision taken by the vast majority of participating surgeons to convert when laparoscopic ductal stone extraction (group B) failed. As indicated in the discussion, endoscopic sphincterotomy and stone extraction, preferably performed in theater while the patient is still anesthetized, is a viable alternative to conversion to open surgery in these patients. It would achieve the target of single-stage management. Finally, the interim results of this prospective randomized trial have shown that single-stage laparoscopic treatment is at least as effective and as safe as the current two-stage management. It should incur savings and reduce the number of unnecessary ERCPs. In the end, however, the management of patients with ductal calculi at the local level must depend on the expertise available.

A. Cuschieri
Department of Surgery Ninewells Hospital and Medical School Dundee, Scotland, UK DD19

Surg Endosc (1997) 11: 10171020

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

Laparoscopic repair of perforated duodenal ulcer

A prospective multicenter clinical trial
M. L. Druart,1 R. Van Hee,2 J. Etienne,3 G. B. Cadie ` re,3 J. F. Gigot,3 M. Legrand,3 J. M. Limbosch,1 B. Navez,3 3 3 3 M. Tugilimana, E. Van Vyve, L. Vereecken, E. Wibin,3 J. P. Yvergneaux3
1 2

Medical and Surgical Unit of Gastroenterology, Hospital Center Etterbeeck-CHEI, Brussels, Belgium Academic Surgical Center Stuivenberg, University of AntwerpUIA, A.Z. Stuivenberg, Lange Beeldekensstraat 267, B-2060 Antwerpen, Belgium 3 The Belgian Group for Endoscopic Surgery, Feluy, Belgium Received: 16 August 1996/Accepted: 1 April 1997

Abstract Background: A series of 100 consecutive patients with perforated peptic ulcer were prospectively evaluated in a multicenter study. The feasibility of the laparoscopic repair was evaluated. Methods: All patients had peritonitis, 20% were in septic shock, and 57% had delayed perforation. Conversion to laparotomy was necessary in eight patients. The morbidity rate was 9% and mortality rate 5%. Results: The mean delay of postoperative gastric aspiration (mean 3.4 days) and resumed food intake (mean 4.4 days) as well as the mean postoperative hospital stay (mean 9.3 days) were comparable to conventional surgery, but postoperative comfort was subjectively increased by laparoscopy and noticed by all laparoscopic surgeons participating in this study. Conclusions: Laparoscopic repair of perforated peptic ulcer proves to be technically feasable and carries an acceptable morbidity and mortality rate, compared with conventional surgery. Key words: Peptic ulcer Perforation Raphy Omentoplasty Peritonitis Laparoscopic treatment

may be indicated and associated to the suture repair by means of a truncal, selective, or highly selective vagotomy, or by means of an anterior seromyotomy or gastric stapling, combined with posterior truncal vagotomy [12, 31, 34]. However, these procedures may be contraindicated in case of peritonitis [4]. They are only justified in approximately one-third of cases in view of the natural history of the ulcer disease and the existence of new medical drugs [27]. The purpose of this study is to evaluate the efficacy of the laparoscopic approach to perforated gastroduodenal ulcer without association of other type of surgical treatment for the healing of the ulcer disease.

Patients and methods

A total of 100 consecutive patients with perforated gastroduodenal ulcer were evaluated prospectively in a multicenter trial set up by the Belgian Group for Endoscopic Surgery. None of the patients underwent vagotomy or seromyotomy for the healing of the ulcer disease. The purpose of our study was only to evaluate the efficacy and safety of the laparoscopic raphy of the gastroduodenal G.D. ulcer perforation. All data were recorded on a specially designed checklist, entered into a computer system, and statistically analyzed. The series consisted of 64 male and 36 female patients with a mean age of 52.5 years (range 1492 years). At clinical presentation signs of an acute abdomen were present in all patients. Associated septic shock, identified in 20 patients and treated preoperatively, was accepted for inclusion in the trial. The white blood cell count was less than 10,000/mm3 in 22 patients, between 10,000 and 15,000/mm3 in 40 patients, and greater than 15,000/ mm3 in 35 patients. An overview of different risk factors is given in Table 1. Previous symptoms of gastroduodenal ulcerative disease were present in 40 patients. Preoperative investigation consisted of plain abdominal X-ray in 97 patients, ultrasound in 33 patients, CT scanning in 27, X-ray investigation after a Gastrografin swallow in 20 patients, and endoscopy in four patients. The delay between perforation and operation is outlined in Table 2. Sixty-seven patients had an empty stomach at the time of operation. At operation, all patients had a peritonitis, either localized (n 35) or gen-

Duodenal ulcer perforation is a serious complication of peptic ulcer disease that occurs in 510% of duodenal ulcer patients and accounts for over 70% of deaths associated with peptic ulcer disease. The treatment of this pathology is essentially surgical [3, 21, 35]. Many authors advocate simple suture of the perforation associated or not with omentoplasty [27, 28, 33]. In some cases, definitive treatment of the ulcer disease
Correspondence to: R. Van Hee

1018 Table 1. Risk factors present in this patient series (n 100) Risk factors Age 70 years Cardiac pathology Chronic respiratory insufficiency Obesity Corticoid treatment Cirrhosis % 25 18 5 5 8 5 Raphy Raphy + omentoplasty Fibrin glue (R/Tissucol) treatment Peritoneal lavage Peritoneal drainage Table 3. Type of laparoscopic treatment, chosen by the individual surgeon n 81 67 7 100 82

Table 4. Reasons for conversion in eight patients Table 2. Delay between perforation and operation Reasons Delay <2 h 26 h 612 h >24 h % 4 38 40 17 Inadequate ulcer localization Posterior location of gastric ulcer Pancreatic infiltration Localized abscess formation Inadequate instrumentation 4 1 1 1 1 n

eralized (n 65). The ulcer location proved to be duodenal in 63 patients, juxtapyloric in 29 patients, and remained unspecified in eight patients. For laparoscopic surgical treatment, the patient is placed in a 1520 reverse Trendelenburg position. The operating surgeon stands between the patients legs. Through a supraumbilical stab incision the pneumoperitoneum is established with a Veress needle and the laparoscope is introduced through a 1012-mm trocar; the other trocars are placed under laparoscopic control: a 5-mm cannula in the epigastrium, used for liver retraction, and two 12-mm cannulae in the right and left subcostal regions, respectively, on the midclavicular and anterior axillary lines. After irrigation with warm saline solution, the perforation is identified. Raphy is performed as in conventional surgery with grasping forceps and needle holder introduced through the lateral 12-mm cannula. The needle is passed through normal duodenum some millimeters from the edge of the perforation to prevent any risk of tearing the ulcer edges and enlarging the perforation. Depending on the choice of the laparoscopic surgeon an omentoplasty was added or fibrin glue was used to seal the closed perforation. Thorough peritoneal lavage is then accomplished by systematic warm saline infusion and aspiration of the peritoneal fluid. Special attention is given to the supra- and subhepatic regions, the left subdiaphragmatic space, and the pelvic cavity. After lavage, all fluid is aspirated and a drain is left under laparoscopic sight in the right subhepatic pouch close to the perforation closure. In case of general peritonitis, a second drain is left in the Douglas recessus and positioned under visual control. Other laparoscopic procedures were associated in three patients respectively, an adhesiolysis, a cholecystectomy, and one liver biopsy. Postoperative management consisted of administration of H2-receptor antagonists, intravenous fluids, antibiotics, and nasogastric aspiration. Subjective well-being of the patients was evaluated postoperatively by the surgeons with respect to abdominal discomfort and rehabilitation.

swallow, showing a suture leak in two patients and a gastric outlet obstruction in two other patients. Nine patients suffered complications, either local or general, or both, resulting in five deaths. As local complications two suture line leaks were observed, necessitating surgical treatment (gastrectomy); two digestive hemorrhages, treated either surgically (gastrectomy) or medically; one Douglas abscess (surgical treatment); one subphrenic abscess treated by CT-guided aspiration under local analgesia; and one parietal abscess treated by antibiotics. Four general pulmonary complications occurred and were successfully treated with medical management. As mentioned, five patients died:

Results Laparoscopic treatment was successful in all but eight cases. The type of treatment and the use of lavage and drainage is shown in Table 3. Conversion to laparotomy was necessary in eight cases. Reasons for conversion are listed in Table 4. The mean operating time, registered in 78 patients, was 80.0 min (range 40135 mins). In the postoperative period, nasogastric aspiration was performed during a mean of 3.4 days (range 110 days). Food intake was resumed after a mean of 4.4 days (range 110 days). The mean postoperative hospital stay lasted 9.3 days (range 240 days) and was less than 5 days in 22 patients and more than 10 days in 19 patients. Ten patients underwent a postoperative gastrographin

An 82-year-old man (ASA 4), in septicemic shock and generalized peritonitis, admitted 6 h after a giant ulcer perforation. A fistula occurred after laparoscopic repair and the patient died in multiple organ failure on the 4th postoperative day. A 74-year-old man (ASA 4), in septicemic shock and generalized peritonitis, lasting more than 24 h. He developed ARDS and died 10 days after laparoscopic repair. A 75 year old man (ASA 4), in septicemic shock and generalized peritonitis, lasting more than 24 h, died of heart failure on the 15th postoperative day. An 85-year-old man with the same clinical conditions as the previous patient died from ARDS on the 10th postoperative day. A 72-year-old man (ASA 4) with similar clinical conditions had a posterior ulcer necessitating a gastrectomy. He developed a digestive hemorrhage and died on the 28th postoperative day from respiratory failure.

Discussion Perforated peptic ulcer remains a challenging disease for the surgeon, occurring far more often than elective peptic ulcer surgery. Not only does perforation occur in almost 10% of all duodenal ulcers, but it often remains the first clinical presentation of the disease.

1019

Ever since the epoch-making work of Taylor [30] conservative treatment of perforated ulcer has had its advocates [22, 23]. As 7580% of such ulcers could eventually heal spontaneously with appropriate nasogastric suction and resuscitation [8], a deliberative approach was suggested, reserving surgical intervention for nonresponders to medical treatment [6]. Broad application of such an approach is, however, destined to lead to a more significant morbidity and mortality, especially in the older age group [24]. Laparoscopic surgical treatment of perforated ulcer seems an attractive alternative for conservative treatment because of the absence of complications compared to conventional laparotomy, especially parietal complications (wound infection and late eventration) and general complications in older patients (pulmonary disease or embolism). Laparoscopic treatment, first reported in 1990 [26], follows the same principles as open surgerynamely, closing the perforation combined with lavage and drainage of the abdominal cavity. For the perforation closure different techniques were proposed: suturing with either intra- or extracorporeal knot tying [9], gastroscopically aided insertion of the ligamentum teres hepatis into the perforation hole [7], stapled omental patch repair [10], or sealing a gelatine sponge or an omental flap into the perforation hole with fibrin glue [1, 29]. In our series we mainly performed suture (n 81), whether or not with omental patch repair (n 67), just as we used to do in open surgery [33], whereas seven patients benefited from fibrin glue sealing. Special attention is given to intraperitoneal lavage, which in our patients was liberally performed with a warm saline solution. Lavage is an adequate measure to counteract the negative effects of peritonitis, which form the major cause of morbidity and mortality in these patients. Some investigators [32] even obtained very satisfactory results performing only laparoscopic lavage and drainage in combination with a conservative Taylor method, confirming the earlier-reported important role of lavage in the perforation management [11]. Our 100-patient series matches favorably with the Hong Kong consecutive series of 100 cases operated by means of either omental patch repair (n 44) or suture patch repair (n 35) or fibrin glue repair (n 21) [25]. As far as parameters can be compared, our series may, however, prove to be subjected to more complications as the mean age is older (52 vs 45 years), presentation delayed for more than 24 hours is more frequent (17% vs 3%), septic shock is more prominent (20% vs 8%), and underlying medical disease is more frequent (33% vs 5%). These differences may account for the higher mortality rate in this series (5% vs 3%), whereas type and number of complications remain similar. Also, the conversion rate was analogous to that in the Hong Kong series [25], 8% and 7%, respectively. Neither study shows a great difference in morbidity or mortality rates, nor in length of hospital stay, compared with conventional surgical treatment, but subjective postoperative comfort of the patients was markedly increased by the laparoscopic approach. Most of the patients could even have been discharged from hospital earlier but remained in the

hospital to receive intravenous antibiotics for 35 days because of the peritonitis, and in view of the Gastrografin swallow on the 4th or 5th postoperative day, before starting to eat. In this patient series, our policy consisted only of raphy of the perforation, without any attempt for definitive ulcer surgery [33]. Laparoscopic types of definitive surgical ulcer treatment may, however, be added to the perforation repair in younger patients with chronic relapsing peptic ulcer disease [20, 34] or may be performed some months later without notable surgical difficulties [20]. Especially anterior gastric stapling, combined with posterior truncal vagotomy, a procedure first introduced by us in 1984 [18, 19] and popularized by Gomez-Ferrer [14], seems to be laparoscopically most appropriate and takes virtually no additional time after perforation repair [15, 17]. The results of this study show the feasability of the laparoscopic approach for perforated peptic ulcer repair with an acceptable morbidity and mortality rate in this multicenter study. Analysis of our results confirm that older patients (70 years) in septic shock and with prolonged peritonitis or other associated diseases remain a high-risk category [5, 24]. Insufflation during a laparoscopic approach has been incriminated as a possible risk factor. Carbon dioxide pneumoperitoneum has indeed increased the incidence of bacterial translocation from the peritoneum into the bloodstream in an animal model [13, 16]. Previous studies equally suggested pneumoperitoneum is a potential risk in cases of preexisting peritonitis [2]. One nonrandomized study showed that two out of 14 patients treated laparoscopically for perforated gastric ulcer (with 15 mmHg pneumoperitoneum) died from peritonitis and septic complications. The increased incidence of bacteremia during insufflation may be related to increasing abdominal pressure as well as to turbulence [2], thereby perpetuating the extent and severity of peritonitis by disseminating contaminated secretions. Notwithstanding the fact that manipulation at open laparotomy may result in similar dissemination, the results of our study made us argue that older patients with septic shock and generalized peritonitis should better be served by conventional surgery. Posterior gastroduodenal perforation should equally be treated by conventional laparotomy (raphy or gastrectomy) because of the difficulty in assessing the posterior side of the gastrointestinal tract by coelioscopy and the risk of postoperative complications like fistula. A randomized prospective large series of patients is needed to confirm the postoperative benefit of this new form of treatment for perforated gastroduodenal ulcer. However, the known advantages of the minimally invasive procedure, such as parietal wall integrity, cosmetic benefit, and early subjective postoperative comfort and rehabilitation, were already noted by all surgeons in this study. References
1. Benoit J, Champault GG, Labhar E, Sezeur A (1993) Sutureless laparoscopic treatment of perforated duodenal ulcer. Br J Surg 80: 1212 (Letter) 2. Bloechle C, Emmermann A, Treu H, Achilles E, Mack D, Zornig C, Broelsch CE (1995) Effect of pneumoperitoneum on the extent and

1020 severity of peritonitis induced by gastric ulcer perforation in the rat. Surg Endosc 9: 898901 Boey J, Wow GJ, Ong GB (1982) A prospective study of operative risks factors in perforated duodenal ulcers. Ann Surg 195: 265269 Bornman PC, Theodorou NA, Jeffrey PC, Marks IN, Essel HP, Wright JS, Terblanche J (1990) Simple closure of perforated duodenal ulcer; a prospective evaluation of a conservative management policy. Br J Surg 77: 7375 Bulut O, Rasmussen C, Fischer A (1996) Acute surgical treatment of complicated peptic ulcers with special reference to the elderly. World J Surg 20: 574577 Cocks R, Kernett RH, Sinclair GWG, McL. Dawson JH, Hong BH (1989) Perforated peptic ulcer: a deliberative approach. Austr NZJ Surg 59: 379385 Costalat G, Dravet F, Noel P, Alquier Y, Vernhet J (1991) Coelioscopic treatment of perforated gastroduodenal ulcer using the ligamentum teres hepatis. Surg Endosc 5: 154155 Crofts TJ, Steele RJC, Cheung SCS, Park K, Li AKC (1987) Prospective randomised controlled trial of conservative therapy versus surgery for perforated duodenal ulcer. Gut 28: A1375 Darzi A, Carey PD, Menzies-Gow N, Monson JRT (1993) Preliminary results of laparoscopic repair of perforated duodenal ulcer. Surg Laparosc Endosc 3: 161163 Darzi A, Cheshire NJ, Somers SS, Super PA, Guillou PJ, Monson JRT (1993) Laparoscopic omental patch repair of perforated duodenal ulcer with an automated stapler. Br J Surg 80: 1552 Delaitre B, Attailla A, Chihaoui M (1988) Ulce ` res gastroduode naux perfore s. Traitement par dialyse pe ritone ale (72 observations). Press Med 17: 12971300 Donovan AJ, Vinson TL, Maulsby GO, Gewin JR (1979) Selective treatment of duodenal ulcer with perforation. Ann Surg 189: 627631 Evasovich MR, Clark TC, Horattas MC, Holda S, Treen L (1996) Does pneumoperitoneum during laparoscopy increase bacterial translocation? Surg Endosc 10: 11761179 Gomez-Ferrer F (1992) Gastrectomie line aire ante rieure et vagotomie tronculaire poste rieure. J Coeliochirurg 4: 3538 Gomez-Ferrer F, Balique JG, Azagra S, Bicha-Castelo H, CastroSousa F, Espalieu P, Rodero D, Estour E (1996) Laparoscopic surgery for duodenal ulcer: first results of a multicentre study applying a personal procedure. Br J Surg 83: 547550 Gurtner GC, Robertson CS, Chung SC, Ling TK, Ip SM, Li AK (1995) Effect of carbon dioxide pneumoperitoneum on bacteraemia and endotoxaemia in an animal model of peritonitis. Br J Surg 82: 844848 Hannon JK, Snow L, Weinstein LS (1993) Linear gastrectomy: an endoscopic staple-assisted anterior highly selective vagotomy combined with posterior truncal vagotomy for treatment of peptic ulcer disease. Surg Laparosc Endosc 2: 254257 18. Hendrickx L, Van de Kelft E, Van Hee R, Hubens A (1984) Anterior lesser curvature stapling and posterior truncal vagotomy: a new method for gastric acid reduction in dogs. Eur Surg Res 16 (Suppl 1): 98 19. Hendrickx L, Van Hee R, Van de Kelft E, Hubens A (1987) Anterior gastric stapling combined with posterior truncal vagotomy: an experimental technique for gastric acid reduction. Eur Surg Res 19: 225232 20. Hendrickx L, Van Hee R (1994) Laparoscopic surgery in the treatment of gastroduodenal ulcer disease. In: Ro her H-D, Heise JW, Verreet PR, Varney M (eds) Update in gastric surgery. Georg Thieme Verlag, Stuttgart, pp 1013 21. Jordan PH, Morrow C (1988) Perforated peptic ulcer. Surg Clin North Am 68: 315329 22. Keane TE, Dillon B, Afdhal NH, McCormack CJ (1988) Conservative management of perforated duodenal ulcer. Br J Surg 75: 583584 23. Kristensen ES (1980) Conservative treatment of 155 cases of perforated peptic ulcer. Acta Chir Scand 146: 189193 24. Kulber DA, Hartunian S, Schiller D, Morgenstern L (1990) The current spectrum of peptic ulcer disease in the older age groups. Am Surg 56: 737? 25. Lau WY, Leung KL, Zhu XL, Lam YH, Chung SCS, Li AKC (1995) Laparoscopic repair of perforated peptic ulcer. Br J Surg 82: 814816 26. Mouret P, Franc ois Y, Vagnal J, Barth X, Lombard-Platet R (1990) Laparoscopic treatment of perforated peptic ulcer. Br J Surg 77: 1006 27. Raimes SA, Devlin HB (1987) Perforated duodenal ulcer. Br J Surg 74: 8182 28. Sawyers JL (1992) Acute perforation of peptic ulcer. In: Scott HW Jr, Sawyers JL (eds) Surgery of the stomach, duodenum and small intestine. 2nd ed. Blackwell Scientific, Boston, MA pp 566572 29. Tate JJT, Dawson JW, Lau WY, Li AKC (1933) Sutureless laparoscopic treatment of perforated duodenal ulcer. Br J Surg 80: 235 30. Taylor H (1946) Perforated peptic ulcer treated without operation. Lancet ii: 441444 31. Taylor TV, Holt S, Headin RC (1985) Gastric emptying after anterior lesser curve seromyotomy and posterior truncal vagotomy. Br J Surg 72: 620622 32. Urbano D, Rossi M, De Simone P, Berloco P, Alfani D, Cortesini R (1994) Alternative laparoscopic management of perforated peptic ulcers. Surg Endosc 8: 12081211 33. Van Hee R (1984) Results of simple suture for perforated peptic ulcer. Acta Chir Belg 84: 97102 34. Van Hee R, Mistiaen W, Hendrickx L, Blockx P (1995) Anterior gastric wall stapling combined with posterior truncal vagotomy in the treatment of duodenal ulcer. Br J Surg 82: 934937 35. Watkins RM, Dennison AR, Collin J (1984) What has happened to perforated peptic ulcer? Br J Surg 71: 774776

3. 4.

5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15.

16. 17.

Original articles
Surg Endosc (1997) 11: 979981

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

Anaesthetic experience with laparoscopic cryotherapy

A new technique for treating liver metastases
C. B. Wallis,1 D. M. Coventry2
1 2

Department of Anaesthetics, Western General Hospital, Edinburgh EH4 2XU, Scotland Department of Anaesthetics, Ninewells Hospital, Dundee DD1 9SY, Scotland

Received: 3 March 1997/Accepted: 28 April 1997

Abstract Background: Laparoscopic cryotherapy is a new technique for treating hepatic tumors that obviates the need for a laparotomy and may reduce the amount of surgical trauma and heat loss associated with the open technique. Liquid nitrogen is applied to the tumor via a cryoneedle probe introduced through a laparoscopic port. The aim of this study was to assess the effect on body temperature and the hematological and biochemical changes associated with this technique. Methods: Five patients who underwent this procedure were studied prospectively under a standardized general anaesthetic. Core and peripheral temperature were measured during the procedure, and blood samples were taken for hematological and biochemical analysis. Results: Freezing time ranged 1957 min and measured blood loss 01000 ml. In one case, bleeding resulted from hepatic surface cracking. Three patients required a blood transfusion. The mean fall in both nasopharyngeal and right atrial temperature was 0.4C. Postoperatively, all patients showed a large rise in alanine aminotransferase (ALT) and a fall in platelet count. A systemic inflammatory response syndrome was seen in some cases, but all patients survived to hospital discharge. Conclusions: Laparoscopic hepatic cryotherapy can be performed without significant temperature changes, but it entails significant morbidity. Key words: Cryosurgery Laparoscopy Anaesthesia

have been shown to be present in 24% of patients at initial colonic resection. Because of their resistance to chemotherapy and radiotherapy they represent a substantial clinical problem [5]. Although local resection of hepatic metastases is sometimes possible, this procedure is unsuitable for the majority of patients due to the extent of disease and coexisting medical conditions. Cryotherapy offers an alternative method of treatment [1]. Liquid nitrogen at 196C is applied to the tumor, causing intracellular ice crystals to form and leading to physical destruction of the cells and death of the tumor mass. In most medical centers, the procedure is performed via a laparotomy. Reported complications include hypothermia, hemorrhage, thrombocytopenia, pulmonary atelectasis, and a need for postoperative ventilation in some patients [6]. There are also anecdotal reports of arrhythmia occurring during cooling, possibly due to a stream of cold blood returning to the heart [6, 4]. Our institution has developed a laparoscopic approach for performing hepatic cryotherapy using specially designed cryoprobes positioned through standard laparoscopic ports [4]. The laparoscopic approach has the theoretical advantage of less heat loss and less surgical trauma than open laparotomy. Therefore we decided to study prospectively the effects of laparoscopic cryotherapy, in particular the intraoperative temperature changes and perioperative changes in hepatic enzymes. Materials and methods
Five patients scheduled for laparoscopic hepatic cryotherapy were studied prospectively. The study had approval from the local research ethics committee, and patients gave written informed consent. Premedication was with oral temazepam 1030 mg. General anaesthesia was induced with propofol, fentanyl, and atracurium. Following tracheal intubation, the lungs were ventilated with isoflurane in oxygen and nitrous oxide. A radial artery cannula was inserted for invasive measurement of arterial pressure and blood sampling. A thermodilution pulmonary artery catheter (Baxter Edwards, Irvine, CA 92714-5686 USA) was introduced via the right internal jugular vein. It was only advanced as far as the right atrium, as judged by the pressure measured at the distal lumen. The thermistor near the tip was

Colorectal cancer is a common disease in developed countries, and 4050% of patients die within 5 years of surgery, mostly with hepatic metastases. Occult hepatic metastases

Correspondence to: C. B. Wallis

980 Table 1. Preoperative and intraoperative serum potassium and arterial pH Preoperative median (range) Potassium, mmol/L normal (3.55) Arterial pH normal (7.367.44) 4.3 (3.94.7) 7.41 (7.327.43) Intraoperative median (range) 4.4 (3.84.5) 7.4 (7.347.45)

Intraoperative potassium is highest value attained; intraoperative pH is measured at end of laparoscopic cryotherapy

109/L, and the lowest value was 74 109/L. There were no clinically significant changes in coagulation as measured by PT or APTT. In the postoperative period, the patients appeared to display some signs of a systemic inflammatory response syndrome (SIRS), especially where cryotherapy had been prolonged or extensive. It was manifested by hypotension, oliguria, and warm peripheries. This condition was treated with additional intravenous fluids. Pulmonary complications occurred in two patients. One had pulmonary edema and a pleural effusion; another had a symptomatic pleural effusion. No patients required postoperative ventilation, and they all survived to hospital discharge.

used to measure right atrial temperature via a cardiac output module (Hewlett-Packard, Andover, MA, 01810 USA). Core body temperature was measured with a nasopharyngeal probe. Carbon dioxide pneumoperitoneum was established by Verres needle and 10.5-mm trocars inserted to allow introduction of a laparoscope to visualize the liver. Implantable needle cryoprobes were inserted into the tumor mass, and the tips were cooled to 196C by a cryogenic unit (Surgical Technology Group, Andover, England). This created an iceball around the tumor, which was assessed using a local ultrasound probe. The iceball was then thawed by irrigation with warm saline (3940C). This cycle was repeated according to the size and number of deposits. Postoperatively, the patients were monitored in a high-dependency area for 2 days using arterial pressure, ECG, and pulse oximetry. During cryotherapy, total freeze time, lowest core, and right atrial temperature and blood loss were recorded. Any arrhythmias were noted. Serum potassium and arterial blood gases were measured during the procedure. Serum albumin, bilirubin, alanine aminotransferase (ALT), alkaline phosphatase, hemoglobin, platelet count, prothrombin time (PT), and activated partial thromboplastin time (APTT) were measured preoperatively and on the 1st and 2nd postoperative days (preop, day 1, and day 2, respectively).

Discussion Laparoscopic hepatic cryotherapy is a new technique for ablation of liver tumors. In this series, we have shown that it appears to be a reasonably safe technique with few anaesthetic problems. The core temperature only fell by an average of 0.4C to 35.9C, which is not clinically significant and in fact is less than might be expected in a laparotomy. A greater fall in temperature was noted in a series of patients undergoing hepatic cryotherapy by an open approach [6]; in this situation, mean temperature fell to 35.2C (range, 33.736.5C). The findings did not support our hypothesis that the proximity of the cryoprobe in the hepatic parenchyma to the inferior vena cava might lead to significant cooling of blood returning to the right atrium. Temperature recordings taken from the right atrium differed little from the nasopharyngeal temperature. This procedure requires active warming measures, such as the use of blood warmers and convective warming blankets. During the thawing stage, it is our practice to irrigate around the supra hepatic vena cava with warmed saline at 3940C to prevent excessive cooling of venous blood. No significant arrhythmias occurred, even though the laparoscopic approach is known to cause bradyarrhythmias [3]. Near-fatal nitrogen embolus has been reported as a complication of hepatic cryotherapy, but it was associated with the use of probes, which allowed direct contact of liquid nitrogen with the tissue [8]. We used probes that delivered liquid nitrogen to the tip via an inner nozzle tube. It then circulates back through the space between the nozzle conduit and the outer metal probe; thus, escape of nitrogen is avoided. There was no significant rise in potassium levels, despite the possibility that cell necrosis would lead to potassium efflux. In addition, some recent work has suggested that prolonged carbon dioxide pneumoperitoneum may cause hyperkalemia [7]. The major biochemical abnormality noted was a large rise in ALT, a result that can be explained by massive hepatocellular necrosis from the iceball. Bilirubin levels rose in all patients, though only in one case did it reach clinically significant levels. This finding has been reported previously [6]. This type of surgery carries a risk of significant hemorrhage, especially from surface parenchymal splits of the liver surface. Fractures can be caused by inadvertent movement of the probes or liver during the freezing stage. It is

Statistical methods
Friedmans analysis for repeated measures was used to compare postoperative to preoperative values. Significance was assumed if p < 0.05.

Results The mean age of patients was 60 (range, 5468) years. The median total freezing time was 32 (range, 1957) min, and the measured blood loss was 100 (01000) ml. In one patient, the liver fractured during freezing, resulting in a 1000ml hemorrhage that required a blood transfusion. Two patients required a postoperative blood transfusion for anemia. The mean nasopharyngeal temperature at the start of surgery was 36.3C; it fell to 35.9C at the end of the procedure. The initial right atrial temperature was slightly higher at 36.5C but fell by the same amount to 36.1C at the end of surgery. The right atrial temperature did not display any sudden reductions, and there were no significant arrhythmias. There was no significant change in serum potassium or arterial pH during the procedure (Table 1). Table 2 shows the changes in serum albumin, bilirubin, alkaline phosphatase, ALT, hemoglobin, platelet count, and coagulation studies in the perioperative period. There was a fall in serum albumin and a small but consistent rise in serum bilirubin in all patients. In one patient, bilirubin rose to 158 mol/L. The most striking finding was a large and significant rise in ALT on day 1, which was beginning to fall by day 2. There was a trend toward significant thrombocytopenia. By day 2, the median platelet count was 126

981 Table 2. Albumin, bilirubin, alkaline phosphatase, ALT, hemoglobin, platelet count, and coagulation studies before and after laparoscopic hepatic cryotherapy Preoperative median (range) Albumin, g/L normal (3650) Bilirubin, mol/L normal (017) Alkaline phosphatase, u/L normal (2085) ALT, u/L normal (035) Haemoglobin, g/dL Platelet count, 109/L normal (150400) PT, s (control 14) APTT, s (control 41)
a

Day 1 median (range) 35 (2540)a 19 (9125) 127 (64287) 1054 (4391204)a 11.2 (9.613.6) 144 (98352) 15 (1417.5) 48 (44.558.5)

Day 2 median (range) 35 (2936)a 20 (14158)a 95 (59246) 671 (154886) 12.5 (7.113.7) 126 (74324)a 15 (1417) 55 (4862)

42 (3143) 7 (615) 127 (53317) 19 (1344) 13.9 (9.515.1) 267 (179522) 15 (1316.5) 45 (4368)

Significant change (p < 0.05) compared to preoperative value.

therefore important that the patient is fully paralyzed and monitored with a peripheral nerve stimulator, because diaphragmatic movement could be catastrophic. Fibrin glue can be used to control any bleeding points. Good intravenous access must be obtained, and central venous pressure monitoring is recommended. In the postoperative period, the patients showed signs of a systemic inflammatory response syndrome (SIRS) [2]. This finding could be explained by hepatic necrosis, release of inflammatory mediators, and activation of the immune system. The thrombocytopenia we observed has been noted previously with hepatic cryotherapy [1] and appears to be a constant feature. It may also be a manifestation of SIRS. The pleural effusions seen in two of our patients has been reported in a previous hepatic cryotherapy series [4]. They may represent a reactive phenomenon due to the proximity of the liver to the pleura or a systemic manifestation of SIRS due to increased capillary permeability. Two patients became anemic postoperatively, one with a right flank hematoma and another with a blood-stained pleural effusion. Therefore, it is crucial to monitor hemoglobin postoperatively. In summary, we have shown that laparoscopic cryotherapy of hepatic tumors can be performed but does carry

some significant morbidity. Although hemorrhage can be significant, hypothermia is not a problem. Postoperatively, further bleeding may occur, with pulmonary complications. References
1. Bayjoo P, Jacob G (1992) Hepatic cryosurgery: biological and clinical considerations. J R Coll Surg Edinb 37: 369372 2. Bone RC, Balk RA, Cerra FB, Dellinger RP, Fein AM, Schein RMH, Knaus WA, Sibbald WJ (1992) Definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis. Chest 101: 16441655 3. Coventry DM (1995) Anaesthesia for laparoscopic surgery. J R Coll Surg Edinb 40: 151160 4. Cuschieri A, Crosthwaite G, Shimi S, Pietrabissa A, Joypaul V, Tait I, Naziri W (1995) Hepatic cryotherapy for liver tumors. Surg Endosc 9: 483489 5. Finlay IG, McArdle CS (1986) Occult hepatic metastases in colorectal carcinoma. Br J Surg 73: 732735 6. Goodie DB, Horton MDA, Morris RW, Nagy LS, Morris DL (1992) Anaesthetic experience with cryotherapy for treatment of hepatic malignancy. Anaesth Intensive Care 20: 491496 7. Pearson MRB, Sander ML (1994) Hyperkalaemia associated with prolonged insufflation of carbon dioxide into the peritoneal cavity. Br J Anaesth 72: 602604 8. Schlinkert RT, Chapman TP (1990) Nitrogen embolus as a complication of hepatic cryosurgery. Arch Surg 125: 1214

SAGES
Surg Endosc (1997) 11: 975978

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

Framework for the future

The 1997 Gerald Marks Lecture Society of American Gastrointestinal Endoscopic Surgeons (SAGES) March 22, 1997
T. L. Dent
Department of Surgery, Abington Memorial Hospital, Abington, PA 19001, USA

President Stiegmann, fellow SAGES members, ladies, and gentlemen, before I begin my remarks, let me recognize Dr. Gerald Marks, in whose honor this lectureship was named. He was honored at this meeting with SAGES first Distinguished Service Award. All of you know Gerry as the Godfather of SAGES. Many of you may not be aware of Gerrys considerable artistic talents. For example, Gerry spent countless hours designing this, our societys logoone that we proudly wear on our SAGES ties. Observe the exquisite details that only an artist of Gerrys caliber could have achieved, and especially the clever way he has spelled SAGES using human intestines as the letters. Seriously, Gerry, this lectureship is a continuing reminder of the debt we owe you for your courage in taking the first steps forward in the leadership of this pioneering organization. My esteemed friend, I salute you. I also want to thank Barbara Saltzman-Berci and her talented staff. Their organizational skills and boundless enthusiasm have helped make our society successful.

his teacher questioned him as to why his autobiography was so short, he replied: Thats all the longer I have lived so far! SAGES, too, is a young organization. Although our history is short in years, it is also long in talent and accomplishments.

Before the revolution In an era when general surgery was driven less by technology, the skills acquired during a general surgery residency were adequate for a lifetime of surgical practice. Postresidency continuing education was a more well-defined and calmer process. Progress in surgical thinking was learned easily by a reading program or by attending lectures. New procedures could be learned in the same way because most were only minor modifications of basic surgical skills. A particularly complex technique occasionally might require the additional effort of watching a videotape or observing another surgeon perform the procedure. Even the incorporation of technology such as surgical staplers in the 1970s was accomplished with little public fanfare, no government scrutiny, and minimal medicolegal reaction. Doubtless, wed still be learning new techniques the same way today had it not been for the endolaparoscopic revolution.

SAGES and the endolaparoscopic revolution For the past several years, I have been intrigued by how the educational issues of endolaparoscopic surgery have been defined in many ways by SAGES, and SAGES as an organization has been shaped by its response to these educational challenges. A review of SAGES educational maturation constitutes the basis of my remarks today. I will lead you on a personal journey, one that begins with SAGES activist origins in flexible endoscopy, continues through its metamorphosis and maturation during the laparoscopic revolution, and ends with some speculation on the future of our Society. As in the line from Shakespeares The Tempest, Whats past is prologue, SAGES short history also is prologue to our future. I am reminded of my cousin, Dennis, who, in fulfilling his fifth-grade assignment to write his autobiography, submitted only one short paragraph. When

Flexible endoscopy: SAGES activist roots With the introduction of flexible fiberoptic endoscopes in the late 1960s, the depths of the upper and lower GI tracts could be probed easily, and new training issues arose. In 1970, my neophyte career in vascular surgery was redirected by seeing Shinya and Wolffs extraordinary film on colonoscopic polypectomy. The effortless way they maneuvered through the colon fascinated me, much like the special effects in the then-popular science fiction movie Fantastic Voyage. Most importantly, these remarkable new instru-

976

ments provided therapeutic access using the bodys natural orifices. When you think about it, operations performed through flexible endoscopes remain the ultimate minimal access procedures. Our colleagues in orthopedics, urology, and otolaryngology recognized the applicability of endoscopic techniques and incorporated them quickly into their practices, effectively precluding other medical specialists from their surgical turf. Most general surgeons, however, were willing to relinquish flexible endoscopy to the gastroenterologists, who quickly exploited the potential of these new instruments. For a while, only a few general surgeons introduced flexible endoscopy into their practices and training programs. We hoped that eventually others would follow suit. The apathy of surgeons toward flexible endoscopy allowed the American Society for Gastrointestinal Endoscopy (ASGE) to promulgate standards for endoscopic training and credentialing, including a full gastroenterology fellowship as a prerequisite for the acquisition of endoscopic skills. As surgical endoscopic techniques expanded, surgeons became increasingly uncomfortable about this state of affairs. Many of you will recall that percutaneous endoscopic gastrostomy was introduced in 1981 by Drs. Gauderer and Ponsky, both surgeons, and the latter a past president of SAGES. A year later, our group at the University of Michigan presented an early surgical experience with PEGs at the Annual Meeting of the Association of Veterans Administration Surgeons. The audience mistakenly thought we were gastroenterologists and was downright hostile during the papers discussion, insisting that most surgeons could do an open gastrostomy more safely and much faster than our reported average of 16 min for PEGs. In 1972, a small group of surgeons with common interests in endoscopy began meeting informally. We resisted the idea of forming yet another surgical society and hoped that the American College of Surgeons or another established surgical group would recognize the importance of gastrointestinal endoscopy and protect us. By 1981, however, several events caused us to believe that the privileging door would soon be closed forever to general surgeons, and SAGES was incorporated. SAGES fundamental mission was clearly stated: to foster and support endoscopy as an integral part of surgical practice. Surgical educators realized that flexible endoscopy required skills that could not be extrapolated easily from prior surgical residency training. A supervised, hands-on experience is the optimal way to learn flexible endoscopy merely watching a videotape, buying a scope, and trying the procedure on patients is not. Even if we had been tempted to use the See one, do one, teach one method, such opportunity was largely precluded by the alreadyentrenched gastroenterologists. Although we did agree that additional hands-on training was necessary, we were convinced that a full 2-year gastroenterology fellowship was not a necessary requirement for surgeons to learn safe endoscopy. The optimal way to integrate any new procedure into surgical practice is first to introduce it into the curriculum of general surgery residency programs. The supervised, progressive education and training in residency programs minimize the complications that are the expected consequences

of any surgical learning situation, commonly called the learning curve. Residents trained in this new procedure could then either become teaching faculty and teach succeeding residents or join surgical practices and teach their peers and partners. SAGES developed and published realistic training guidelines for surgeons. They also helped residency program directors incorporate endoscopic training into their programs by organizing faculty courses and by identifying preceptorships at members institutions. I fondly recall our surgical endoscopy preceptorship at the University of Michigan, where several of you in the audience began your endoscopic careers. The American Board of Surgery also was an ally. By mandating in 1980 that candidates for board certification should be able to adequately perform flexible endoscopic procedures, the Board kept the endoscopic door open while we were getting our educational act together. Surgeons trying to obtain hands-on training in flexible endoscopy encountered a roadblock: There was no organized mechanism to provide such training to large groups of surgeons who had already completed their residency. No surgical society or institution had addressed large-scale, procedural training for surgeons in practice. Most existing residency programs were struggling to teach endoscopy to their current residents and could not offer training to others. A few individual preceptorships existed, but they varied in quality and lacked coordination and standardization. The American College of Surgeons and other surgical societies, although experienced in providing excellent cognitive updates, lacked hands-on laboratories. SAGES stepped forward to address the issue of postresidency endoscopic training. Fortunately, the initial limited interest of surgeons in learning flexible endoscopy allowed SAGES the time to develop and certify hands-on courses. We were aided by the many instrument manufacturers who accepted our fiat that surgeons, not instrument technicians, must be responsible for the content of these training programs. As these courses evolved, we learned better ways of providing postresidency hands-on training by utilizing training models and animate laboratories. This teaching experience, as it turned out, would later serve us well. As a salutary result of the efforts of many individuals and organizations, flexible endoscopy has been successfully retained as an integral part of general surgery. Proof is contained in a recent study by Dr. Wallace Ritchie, the executive director of the American Board of Surgery, who collected the operative experience of 685 practicing general surgeons who sat for the 1996 board recertification examination. Subspecialist surgeons, such as vascular and cardiac surgeons, were excluded from the statistics. These general surgeons individually perform an average of 381 cases per year, and endolaparoscopy comprised fully 25% (13% endoscopic, 12% laparoscopic) of their operative volume. By 1989, SAGES was proud of its modest accomplishments in retaining flexible endoscopy for surgeons and was pleased with the gradual growth of the membership and the modest increases in annual meeting attendance. We breathed a collective sigh of relief, little realizing that the lessons learned about hands-on training in endoscopy were about to be revisited as we were catapulted into the middle of the laparoscopic revolution.

977

Laparoscopy: SAGES metamorphosis Like flexible endoscopy, surgical interest in laparoscopy was minimal until another catalytic depiction captured our attention. In April 1989, Professor Jacques Perrisat, whose presentation had not been accepted for the SAGES main program, displayed his videotape on laparoscopic cholecystectomy in a remote booth of the exhibit area. This videotape quickly attracted a larger audience than did the lecturers in the main auditorium. The clamor of 19,000 practicing surgeons interested in learning this stunning innovation again challenged the educational capabilities of American surgery. Gradually introducing laparoscopic cholecystectomy into residency programs would have been too slow a process for both the public and the surgical community. Even SAGES was initially unprepared to meet this unprecedented educational demand, even with the lessons learned from flexible endoscopy still fresh in our minds. As usually happens when demand exceeds supply, entrepreneurs stepped forward. Surgeons paid exorbitantly for hands-on training courses and even, on occasion, paid handsomely merely to watch a colleague operate. Technician-taught manufacturers courses proliferated. Unfortunately, the incidence of patient injury temporarily increased as the learning curve was negotiated. SAGES expansion into laparoscopy confirmed our societys adaptability and underscored its academic, ethical, and organizational strengths. The primary focus this time shifted from activism to education. Existing guidelines in training and privileging for flexible endoscopy were easily adapted to laparoscopy. This is not a surprising leap, if you think about it, because laparoscopy is endoscopy. SAGES organized 23-day introductory courses for residents and attendings. With general surgeons years of experience in open cholecystectomy, it was only the new method of access and slightly modified instrumentation that had to be learned, a fact frequently forgotten by the media and the legal profession. Most surgeon-attendees of SAGES introductory courses incorporated laparoscopic cholecystectomy into their practices gradually and safely. They typically began with elective, nonemergent procedures, frequently assisted by other surgeons embracing the new technique, and with a low threshold for conversion to an open operation. In my opinion, to malign this method of skills acquisition with the demeaning label of weekend learning because a few surgeons forgot about the steepness of the learning curve does a disservice to the vast majority of general surgeons who seamlessly and safely incorporated basic laparoscopy into their practices. Laparoscopic cholecystectomy was completely incorporated into general surgery in an astonishing 23 years, while the surgical integration of flexible endoscopy is, ironically, still incomplete, even after 25 years. SAGES, and you, its members, were instrumental in introducing laparoscopy into residency training programs. SAGES excellent annual course at the Ethicon facility in Cincinnati, organized and run by Dr. Jeffrey Ponsky, continues to introduce 120 surgery residents a year to endolaparoscopy at no charge. Basic laparoscopy has been firmly integrated into the curriculum of general surgery residency programs. According to ACGME/RRC statistics, the average 1996 surgery

residency graduate performed 83 laparoscopic procedures and 108 endoscopic procedures during his/her 5 years of training. These numbers compare favorably with the numbers for other procedures performed in the primary components of general surgeryspecifically, 101 large intestinal and 32 small intestinal procedures and 76 breast and 21 endocrine procedures.

Advanced laparoscopy: SAGES maturation A natural consequence of surgical enthusiasm was extending laparoscopic techniques from the gallbladder and appendix to other abdominal and retroperitoneal organs, and, as we have seen here in San Diego, into the neck, as well. Such advanced laparoscopy occupies the attention of most SAGES committees at the present time, and the scientific study of these procedures and the validation of their efficacy are high priorities. The Research Committee, with the generous assistance of our corporate partners, has awarded over 40 study grants since 1992. The American College of Surgeons Committee on Emerging Surgical Technology and Education is utilizing the expertise of Col. Richard Satava, a past president of SAGES, in evaluating surgical simulators. The same committee is also launching a long-overdue multi-institutional comparison of laparoscopic and conventional hernia repairs, a study organized and led by Dr. Robert Fitzgibbons, a current SAGES board member. Advanced laparoscopic education and training have not been overlooked by SAGES. The 1996 Surgical Operative Logs from the ACGME/RRC demonstrate that most surgical residents are poorly trained in advanced laparoscopy. Not surprisingly, SAGES is already addressing residency education. An ad hoc Residency Integration Committee, chaired by Dr. Bruce Shirmer, has been charged by President Stiegmann to define the educational standards and facilitate the teaching of advanced laparoscopic procedures within general surgery residency programs. This committee already has outlined the common core skills, that, once mastered, will allow the safe performance of most advanced procedures. These skills include two-handed dissection, intracorporeal suturing, and intra- and extracorporeal knot tying. This core skills concept advocated by the committee is analogous to the RRCs complex interchangeable procedure model and will eliminate the need for separate training and specific credentialing for each and every advanced laparoscopic procedure. Responding to the recommendation of the Residency Integration Committee, SAGES will soon offer residency program directors introductory and subsequent hands-on courses for residents and faculty in advanced laparoscopic techniques, as well as providing coordination of mentoring opportunities for current faculty and selected formal fellowships for future faculty members. The societys commitment to general surgery has been revalidated by its recent decision to table its plans to certify and credential advanced laparoscopic fellowships. These SAGES-sponsored fellowships were originally envisioned as an altruistic response to the needs of general surgery residents for the advanced training that is currently unavailable to them. The SAGES board demonstrated its sensitivity to the concerns of the American Board of Surgery and

978

the Residency Review Committee that SAGES-accredited fellowships might dilute general surgery residents experience in advanced laparoscopy and might further fragment general surgery, as some think occurred when vascular surgery was granted a Certificate of Added Qualifications in 1982. SAGES vision Well, where are we going? Springing from our activist roots, SAGES has evolved into an influential organization that I believe is positioned well to represent our mutual interests as future directions of general surgery education and practice are established nationally. The September 1994 issue of Surgical Endoscopy contains SAGES Framework for Post-Residency Surgical Education and Training. In my opinion, this landmark document, representing the long, hard labor of three SAGES committees, establishes our organization as a leader in American surgical education. If you havent read this remarkable work, you should. Its clear generic guidelines for acquiring safe surgical skills and assessing surgeon competence are a visionary template for accommodating any emerging technology or newly described surgical procedure. Just last week, as Dr. Stiegmann reported in his Presidential Address, SAGES leadership met with the Executive Committee of the Association of Program Directors in Surgery in Atlanta and described SAGES plans to help the program directors with the integration of advanced laparoscopy into surgical residency programs. These plans also will be described to all general surgery program directors at their annual meeting next month in San Diego. More importantly, SAGES began discussions with the APDS leadership concerning ways of increasing the flexibility of surgical residency programs to better accommodate advanced endolaparoscopy. As a result of these discussions, a creative template for advanced education in endolaparoscopy may be developed that could also be applied to other primary components of general surgery (such as trauma/critical care, vascular surgery, and surgical oncology) within the 5-year residency time-frame and possibly to other related interests, such as rural surgery, as well. Recent APDS leaders, including Drs. Robert Barnes and Walter Pories, have championed such residency restructuring, in part because federal funding for postgraduate training may soon become restricted to a total of 5 years, effectively eliminating postresidency surgical fellowships as they exist today. An attractive goal of such residency restructuring, from my perspective, is to provide a more flexible alternative to the current expectation that every general surgery resident

will have an identical experience in all of the primary components of general surgery. Less rigid interpretations of the training regulations by the American Board of Surgery and the RRC might allow a concentrated experience (also known as tracking) in a core surgical area, such as advanced laparoscopy, during the senior resident years, provided the general surgery minimum requirements still are met. This strategy could provide resident expertise in a special area of interest without complicating the credentialing and privileging process with special certificates, and would be analogous to a major in college. An issue relevant to such tracking has been raised by President Stiegmann in the January issue of the American Journal of Surgery and by the Residency Integration Committee in their draft position paper. As both have observed, and as was discussed with the APDS leadership, the numbers of advanced laparoscopic procedures available for resident training are limited. It may be a waste of educational resources to teach advanced laparoscopy to all surgical residentsespecially residents who are planning subspecialty careers and who will never utilize these advanced skills. Tracking of advanced procedures, not only in laparoscopy, but in other components of surgery as well, is an attractive concept for the creative training of surgical residents in skills specifically tailored to their future practice. With these recent initiatives, SAGES is now moving to the next level of its evolution: namely, sharing SAGES education and training vision with other influential organizations who are currently deliberating fundamental changes in the structure of general and subspecialty surgical residency education.

SAGES mission SAGES has come a long way from the tiny band of believers in 1981 to our current membership of almost 3,200 surgeons, one of the largest general surgical organizations in the country. SAGES mission, however, remains as clear as it was when Gerry Marks and its founders committed it to paper: to foster, support, and encourage academic, clinical, and research achievement in gastrointestinal endoscopy as an integral part of gastrointestinal surgery. In preparing this talk and considering how SAGES was created and has become energized by the endolaparoscopic revolution, the words of Victor Hugo seemed appropriate to me: Would you realize what Revolution is, call it Progress; and would you realize what Progress is, call it Tomorrow. Thanks to all of you, SAGES has matured far beyond its activist roots and is solidly prepared to face the challenges of tomorrow.

Surg Endosc (1997) 11: 10261028

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

Laparoscopy in the evaluation and treatment of patients with AIDS and acute abdominal complaints
J. C. Box,1 T. Duncan,1 B. Ramshaw,1 J. G. Tucker,1 E. M. Mason,1 J. P. Wilson,1 D. Melton,2 G. W. Lucas1
1 2

Department of Surgery, Georgia Baptist Medical Center, 303 Parkway Drive, Atlanta, GA 30312, USA Department of Infectious Diseases, Georgia Baptist Medical Center, 303 Parkway Drive, Atlanta, GA 30312, USA

Received: 26 November 1997/Accepted: 7 May 1997

Abstract Background: The evaluation of AIDS patients with acute abdominal complaints (AAC) is quite difficult, and surgical intervention is associated with a high complication rate. The intent of this study is to evaluate the application of laparoscopy in the diagnosis and treatment of AIDS patients with AAC. Methods: This is a retrospective analysis of 10 consecutive AIDS patients who presented with AAC. Each had evaluation by a surgical team with subsequent laparoscopic intervention. The charts were reviewed for age, sex, time with AIDS, AIDS comorbidities, evaluation modalities, findings, treatment modalities, and outcome. Results: Laparoscopy resulted in the successful surgical treatment of four patients, diagnosis of medically treatable conditions in four patients, and alteration of the incision site in the remaining two patients. Each patient thus received direct benefit from laparoscopy. Two complications, in the converted patients, and no mortalities were encountered. Conclusions: Laparoscopy is a safe and effective interventional modality in the diagnosis and treatment of AAC in the AIDS patient. Key words: Laparoscopy Acquired immunodeficiency syndrome Acute abdomen

cases of AIDS, 18 million cases of HIV in adults, and over 1.5 million cases of HIV in children worldwide. In Georgia alone, there were over 14,000 cases of AIDS with over 8,000 attributable deaths as of February 1996 [7]. In Atlanta, tenth in the United States for cumulative cases, we have seen an increasing population of patients with unique needs blended with unusual clinical presentations [7, 8]. The increasing number of patients with AIDS achieving longer survival is destined to increase the number of patients requiring surgical consultation for evaluation of abdominal complaints. The prohibitive morbidity and mortality have increased the difficulty of decision-making in these patients. Laparoscopy is a safe mode of evaluation and therapy well suited to bridge the dilemma of uncertain diagnosis which may differentiate medical versus surgical therapy. In addition, laparoscopy can often provide definitive therapy. We have undertaken a review of some particularly difficult cases to illustrate these issues and help understand the role of laparoscopy. Materials and methods
The admission records of Georgia Baptist Medical Center (GBMC) from January 1993 to December 1995 were reviewed to identify AIDS patients admitted with acute abdominal complaints for whom a surgical evaluation was undertaken. Of this group, 10 patients were identified in whom laparoscopy was utilized in their evaluation or therapy (Table 1). During this time period, 1,930 cases of laparoscopy were performed at GBMC. Age, sex, time with AIDS, evaluation modalities, findings, treatment modalities, and outcome were noted.

The increasing incidence of HIV and AIDS is well recognized in the United States. AIDS has become the leading cause of death for men ages 2544 in the United States [5]. There were over 501,100 cases of AIDS in the United States as of 1995 [10]. This comes from a total of well over 2,000,000 cases of HIV. The problem is not isolated to the United States. It is estimated that there are over 4.5 million

Results There were nine males and one female with an average age of 34 (2840) years. The time with clinical AIDS varied greatly, average 6 (112) years. Two presented as the index presentation of AIDS, requesting testing on diagnostic admission. Preoperative comorbidities included atypical mycobacterium infection in three, CMV retinitis in two, CNS toxoplasmosis in one, cryptosporidiosis in one, tertiary

Correspondence to: J. C. Box, Department of Surgical Oncology, Roswell Park Cancer Institute, Elm and Carlton Streets, Buffalo, NY 14263

1027 Table 1. Patient summary Preop diagnosis Non specific FUO, abd pain FUO, abd pain Abd pain RUQ complaints Acute cholecystitis Acute cholecystitis Acute cholecystitis RLQ complaints Acute appendicitis Acute appendicitis Acute appendicitis perforated Acute appendicitis Postop diagnosis Diffuse peritoneal T-cell lymphoma CMV peritonitis Metastatic melanoma to SB Adenocarcinoma gallbladder metastatic Acute cholecystitis CMV/Crypto Miliary hepatic abscesses Acute appendicitis focal perforation Acute ileitis/typhlitis Perforated ulceration small bowel Infected hemorrhagic ovarian cyst Procedure Diagnostic Laparoscopy Diagnostic Laparoscopy Lap assisted SB resection Lap Cholecystectomy IOC Lap Cholecystectomy IOC Diagnostic Laparoscopy Laparoscopic Appendectomy Diagnostic Laparoscopy Converted Open small Bowel resection Converted Open salpingooophorectomy Open No No No No No No No No Yes Yes Complication None None None None None None None None Small bowel obstruction Pneumonia

Lap, Laparoscopic; FUO, Fever of unknown origin; RUQ, Right upper quadrant; RLQ, Right lower quadrant; IOC, Intraoperative cholangiogram; Abd, Abdominal; CMV, Cytomegalovirus; Crypto, Cryptosporidiosis

syphilis in one, multiple active sexually transmitted diseases in one, and active IV drug abuse in two (Table 2). Based on clinical presentation the patients were placed into one of three general categories: nonspecific abdomen, right upper quadrant complaints, and right lower quadrant complaints. The nonspecific abdomen group consisted of three patients: T-cell lymphoma of the peritoneum, metastatic small-bowel melanoma, and CMV peritonitis. They all had fever of unknown origin and generalized vague abdominal complaints. Their clinical exams all indicated acute processes. However, no clinical exam alone mandated surgical exploration. Extensive workup, including radiologic and laboratory evaluation, was nondiagnostic in each patient except for the one with the T-cell lymphoma. In this case, the gallium scan showed diffuse uptake in the peritoneal cavity. Each patient benefited from surgical exploration and was spared the morbidity of a laparotomy. The patients underwent complete abdominal exploration with cytology, cultures, and tissue procurement. The small bowel with the melanoma metastasis was resected in a laparoscopy-assisted fashion with a 5-cm incision. No complications were encountered in this group. The right upper quadrant complaint group included three patients as well: adenocarcinoma of the gallbladder, cryptosporidiosis/CMV cholecystitis, and multiple hepatic granulomatous abscesses. The first two patients had many comorbid factors, avoided delay in therapy, and received significant palliation. The last patient had a minimally invasive, definitively diagnostic procedure after previous diagnostic modalities made an erroneous diagnosis. A lapa-

Table 2. Preoperative comorbidities in study population Atypical mycobacterium infection CMV retinitis CNS toxoplasmosis GI cryptosporidiosis Tertiary syphilis Multiple STDs Active IVDA 3 2 1 1 1 1 2

CNS, Central nervous system; GI, gastrointestinal; STD, Sexually transmitted disease; IVDA, Intravenous drug abuse

rotomy was avoided in this patient, who did well with medical therapy for a medically treatable condition. No complications arose in this group. The final group presented with right lower quadrant complaints. This group included four patients: acute appendicitis with focal perforation, small-bowel ulceration and perforation, acute ileitis/typhlitis and infected, hemorrhagic ovarian cyst. These patients presented with right lower quadrant pain and a white blood cell count of 0.86.7. The time with AIDS varied from 8 years to index presentation (for the patient with appendicitis). The patients with appendicitis and ileitis had uneventful postoperative courses. The two other patients underwent conversion to open due to the laparoscopic findings of diffuse peritonitis. Both of these patients had complications, pneumonia, and small-bowel obstruction. They had prolonged hospital stays but eventually recovered to baseline and were discharged. No mortality was encountered in these patients. The

1028

morbidity was 20%. Both complications were encountered in the patients converted to open laparotomy. Discussion These cases illustrate the variety of findings one may encounter while evaluating the acute abdomen of a patient with AIDS. The number of patients with HIV and AIDS is growing rapidly in our country. More effective medical management of the condition is advancing HIV into the class of chronic diseases. Surgeons are going to be called on to evaluate patients with HIV and AIDS more frequently. Due to the unique nature of the immunological influence these patients are under, they often present with unique pathological processes [11]. However, they are still subject to more common conditions such as appendicitis and cholecystitis. Many of these AIDS-related processes, including some of the bizarre infectious presentations, i.e., CMV cholangitis and infectious ileitis, are preferably treated medically [6]. However, some are difficult to differentiate from surgical emergencies. In addition, many common surgical processes may present very oddly [1]. The survival of HIV and AIDS patients is being prolonged on a scale of years. Ninety percent of these patients will develop gastrointestinal symptoms prior to their death [9]. Approximately 4% will require surgery. The patient with AIDS carries a substantially increased morbidity and mortality above the baseline. Mortality for urgent operation (1170%), open cholecystectomy (3338%), and even laparoscopic cholecystectomy (033%) are well above the accepted average for non-HIV patients and HIV-positive non-AIDS patients [3,4,9]. The morbidity is extremely high as well. The surgeons awareness of this substantial morbidity and mortality has often led to uncertainty about the timing and type of intervention to best serve the patient without undue risk to the patient, surgeon, and staff [1]. These cases illustrate cooperation with the infectious disease specialists to allow the surgeon an early opportunity to evaluate the patient. This is facilitated by having a safe modality to diagnose and often treat these patients. Laparoscopy can be utilized to complete a thorough examination of the peritoneal cavity and obtain cultures, tissue, and cytology. When conversion is needed, appropriate initial incisions can be made, i.e., avoiding a right lower quadrant incision when a formal laparotomy is needed. We have found that the low morbidity has encouraged earlier intervention in this population. In all of these patients prolonged delay may have proved disastrous.

Conclusion Laparoscopy is a technique proven to be effective in evaluating and treating a variety of abdominal processes. In the AIDS patient it can be used to evaluate and treat both specific and nonspecific clinical presentations [2,12]. The surgeon can obtain fluid for cytology and culture and tissue for inspection. The abdomen can be visualized. Definite therapy may be possible, and if not, it can guide laparotomy. Laparoscopy can be used to bridge the gap of noninvasive diagnostic investigation and laparotomy. Thus delay in surgical investigation is no longer excusable solely due to a concern over prohibitive morbidity and mortality to the patient. As well, as the primary individual at risk, the surgeon (surgical team), should play a major, active role in conjunction with the infectious disease specialist, in evaluating these patients for appropriate timing and method of surgical intervention. In order to do this we must remain educated concerning the course, copathology, and management of the patient with AIDS.

References
1. Binderow SR, Shaked AA (1991) Acute appendicitis in patients with AIDS/HIV infection. Am J Surg 162: 912 2. Bouillot JL, Dehni N, Kazatchkine M, Fernandez F, Piketti C, Salah S, Alexandre JH (1995) Role of laparoscopy in the management of acute abdomen in the HIV-positive patient. J Laparoendosc Surg 5(2): 101 104 3. Burack JH, Mandel MS, Bizer LS (1989) Emergency abdominal operations in the patient with acquired immunodeficiency syndrome. Arch Surg 124: 285286 4. Carroll BJ, Rosenthal RJ, Phillips EH, Bonet H (1995) Complications of laparoscopic cholecystectomy in HIV and AIDS patients. Surg Endosc 9: 874878 5. Detels R (1995) Epidemiologic contributions to the HIV and AIDS literature. Curr Opin Inf Dis 8: 5153 6. Ducreux M, Buffet C, Lamy P, Beaugerie L, Fritsch J, Choury A, Liguory C, Longuet P, Gendre JP, Vachon F, Gentilini M, Rozenbaum W, Le Quintrec Y, Etienne JP (1995) Diagnosis and prognosis of AIDS-related cholangitis. AIDS 9: 875880 7. Georgia AIDS Surveillance Report (February 1996) 8. HIV/AIDS in Georgia, Executive Summary (1994) 9. LaRaja R, Rothenberg R, Odom J, Meuler S (1989) The incidence of intraabdominal surgery in AIDS: a statistical review in 904 patients. Surgery 105(2): 175179 10. Morbidity and Mortality Weekly Report (1995) 44(46): 849853 11. Scott-Conner CEH, Fabrega AJ (1996) Gastrointestinal problems in the immunocompromised host. Surg Endosc 10: 959964 12. Tanner AG, Hartley JE, Darzi A, Rosin RD, Monson JRT (1994) Laparoscopic surgery in patients with human immunodeficiency virus. Br J Surg 81: 16471648

Surg Endosc (1997) 11: 991994

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

Three-dimensional endoscopic ultrasound of the esophagus

Preliminary experience
M. Hu nerbein, S. Gretschel, B. M. Ghadimi, P. M. Schlag
Ro bert Ro ssle Hospital and Tumor Institute, Virchow University Hospital, Humboldt University, Lindenbergerweg 80, 13122 Berlin, Germany Received: 18 February 1997/Accepted: 29 April 1997

Abstract Background: The aim of this study was to develop a technique for three-dimensional endoscopic ultrasound of the esophagus based on standard ultrasonic images. Methods: Endoscopic ultrasound was performed in five esophageal cancer patients using a high-resolution miniprobe (360, 12.5 MHz). For acquisition of threedimensional data sets, the miniprobe was attached to a stepping motor that enabled ECG-triggered withdrawal of the transducer. Three-dimensional images were reconstructed from serial transverse sections on a PC-based 3D work station. Results: Twelve volume scans were obtained in five patients with esophageal cancer. The system enabled the acquisition of accurate three-dimensional ultrasound data within 30 50 s. Computed image processing allowed us to display the data in transverse, longitudinal, and oblique sections, or as a 3D reconstruction. Three-dimensional imaging provided accurate visualization of the tumor and surrounding structures in all cases. The tumor stage was determined correctly in four of five patients. Longitudinal scan planes and 3D views improved the assessment of longitudinal tumor infiltration and the spatial relation of the tumor to relevant mediastinal structures. Conclusion: This study shows that three-dimensional endoscopic ultrasound of the esophagus is technically feasible. The technique allows the assessment of local tumor spread in previously unattainable scan planes and 3D views. This promising preliminary experience should encourage further exploration of this method. Key words: Endoscopic ultrasound Three-dimensional imaging Esophageal cancer Staging

Endoscopic ultrasound (EUS) is the most sensitive technique for the preoperative staging of esophageal cancer. Accuracy rates of 80% in the determination of the infiltration depth and accuracy rates of 70% in the detection of lymph node involvement have been reported [1, 6, 11, 12]. EUS is usually performed with 7.5 MHz transducers that provide 360 transverse scans of the esophagus and the surrounding structures [10]. Although transverse scan planes have proven very useful in determining the infiltration depth of esophageal cancer, this technique has several limitations. Because only discrete 2D images are available to the operator at any given time, EUS provides no direct information on the longitudinal extent of the tumor and its spatial relationships. Consequently, it is necessary to integrate a series of transverse images in order to mentally visualize the real anatomy. Thus, the procedure requires repeated passes of the scan plane over the region of interest, which can be time-consuming and uncomfortable for the patient. Even so, it is still difficult to obtain a spatial impression of the tumor and its relation to relevant mediastinal structures. This study was conducted to evaluate the technical feasibility and the diagnostic value of three-dimensional EUS of esophageal cancer. For this purpose, we developed a technique for 3D data acquisition using miniaturized endoscopic ultrasound probes and a 3D ultrasound processor. Methods
Three-dimensional ultrasound scans were done in five patients with esophageal cancer. EUS was performed with a standard ultrasound unit (B&K 3535; Gentofte, Denmark) and miniaturized ultrasound catheters, which were originally designed for intravascular ultrasound (Endosound; Boston Scientific, Watertown, U.S.A.). Operating with high frequency (12.5 MHz), the mechanical miniprobe provides high-resolution 360 images of the gastrointestinal wall. The penetration depth of the transducer is approximately 24 cm. The dimensions of the catheter (diameter, 6 Fr; length, 2 m) allow to pass it through the working channel of a gastroscope. For 3D data acquisition, the miniprobe was attached to a stepping motor, which enabled standardized withdrawal of the transducer. Threedimensional scanning was achieved by recording a multitude serial images while the ultrasound probe was withdrawn at a defined rate (Fig. 1). Con-

Correspondence to: P.M. Schlag

992

Fig. 1. Principle of three-dimensional endoscopic ultrasound. Consecutive transverse sections of the 2D transducer are sampled at defined times during pullback of the probe. Fig. 2. Longitudinal section of the esophagus reconstructed from three-dimensional data acquired with the continuous pullback method. Cyclic artifacts caused by cardiac motion degrade the image. > tumor, LN lymph node. Fig. 3. Three-dimensional section display. The reference image (A) indicates the orientation of the reconstructed longitudinal scan plane (B). The tumor (>) is displayed with the vena azygos (VA). Fig. 4. Transverse scan (A) and longitudinal reconstruction (B) of esophageal cancer (>) growing along the aorta (AO). Fig. 5. Three-dimensional volume display. Volume rendering provides realistic images of esophagus (1 mucosa, 2 submucosa, 3 muscularis). The tumor disrupts the layers (arrows) and penetrates the muscularis propria (arrowheads). UP ultrasound probe. tinuous speed pullback, at a rate of 1 cm per 3 s with 0.10.2 mm cross sections, was used for image acquisition in the first scans. Later, an ECGgated stepping motor was employed for withdrawal of the miniprobe to avoid artifacts caused by cardiac motion. This device allows stepwise pullback of the transducer with acquisition of images at a specific phase of the cardiac cycle, e.g., R-wave. The time required to traverse the esophageal segment during catheter withdrawal is dependent on the number of steps. The resolution of the final reconstruction, in turn, is determined by the distance between the slices. The examinations were performed as follows: All patients received premedication with 5 mg midazolame and 20 mg butylscopolamine. First, a conventional esophagogastroduodenoscopy was performed to determine the exact location and size of the tumor. Then, the tip of the endoscope was placed 23 cm proximally to the tumor. The esophageal lumen was filled with 4060 ml of deaerated water to achieve an acoustic interface between transducer and esophageal wall. The ultrasound catheter was introduced through the working channel and advanced beyond the tumor. Data acquisition was completed while the probe was pulled back by the stepping motor. Volume data obtained during the scanning process were stored on a hard disk and subsequently processed with a 3D ultrasound system (TomTec, Unterschleiheim, Germany) that incorporated a 133-MHz Pentium processor (64MB RAM) and specialized software.

Results The mechanical acquisition system enabled the generation

993

of accurate three-dimensional data sets for subsequent image analysis on the 3D work station. Twelve threedimensional scans were recorded in five patients with esophageal cancer. Typically, acquisition of a volume scan from esophageal segments of 510 cm length took 3050 s. However, images reconstructed from data sets that were obtained with the continuous pullback method showed cyclic artifacts caused by cardiac motion (Fig. 2). ECGtriggered withdrawal significantly reduced the motion artifacts caused by the heartbeat. Three-dimensional image analysis provided two display modes for the evaluation of three-dimensional data. Section display allows visualization of multiple sections along three orthogonal planes or oblique projections. Individual views are depicted along with a reference image that demonstrates how each section is cutting the volume (Fig. 3). Generally, there is no limitation with respect to the orientation of the scan plane. Postprocessing of volume data permits assessment of reconstructed scan planes that are not available with conventional EUS, e.g., longitudinal and oblique scan planes. In our experience, longitudinal sections of the esophagus have proven valuable for the assessment of local tumor spread. These views provide important information on the length of the tumor and its relation to relevant mediastinal structures, such as major vessels (Fig. 4). Three-dimensional views are produced by volume rendering and provide life-like three-dimensional visualization of the real anatomy (Fig. 5). The esophageal wall is depicted as a tubular structure within the surrounding tissue. The time required for a three-dimensional reconstruction usually ranges between 1 and 5 min. Three-dimensional image analysis allowed accurate assessment of tumor penetration and metastatic lymph node involvement in four of five patients. Histopathologic evaluation of resection specimens showed that three tumors were correctly classified as uT3,N1 and one tumor as uT3,N0. Understaging occurred in a patient with a uT3 tumor. Thoracotomy disclosed advanced carcinoma of the esophagus with infiltration of the chest wall (pT4). This failure in the diagnosis was probably due to inadequate penetration of the 12.5-MHz transducer.

Discussion Although three-dimensional imaging has been used successfully in other areas of medicine, there are only sporadic reports on the use of three-dimensional endoscopic ultrasound in gastrointestinal disease [3, 4, 13]. Limited experience with this technique has been gained with transrectal endoscopic ultrasound [5, 7, 9]. In a pilot study, Mueller et al. performed three-dimensional endorectal ultrasound in three patients with rectal cancer [9]. Recently, a comprehensive study involving 100 patients has shown encouraging results with three-dimensional endorectal ultrasound in the evaluation of rectal cancer [7]. One major reason for the lack of data on 3D endosonography of the esophagus is the problem of data acquisition. In the only available report, Kallimanis et al. confirmed the feasibility of 3D endosonography of the esophagus [8]. However, the quality of the images was compromised by motion artifacts caused by manual withdrawal of the probe

and circulatory pulsations. A similar problem has been also reported with 3D intravascular ultrasound [2]. For this study, we developed a mechanical system for the acquisition of 3D data with EUS. Three-dimensional scans were obtained from serial sectional images that were recorded while the endoprobe was pulled back by a stepping motor. Acquisition of images at a defined phase of the cardiac cycle was achieved by ECG-gated withdrawal of the transducer with the motor. Data sampling with this method allowed accurate reconstruction of three-dimensional ultrasound images without significant motion artifacts. Three-dimensional image analysis provided two display modes. In the section display mode, data are depicted in planes that can be defined deliberately in the scanned volume. Therefore, it is possible to visualize the tumor in previously unattainable projections. Reconstruction of longitudinal scan planes proved to be very valuable for assessment of longitudinal tumor extension. Exact determination of longitudinal tumor infiltration is of major importance, because underestimation of submucosal invasion in the esophagus may lead to nonradical resection. The volume display mode provides 3D views that closely resemble the real anatomy. Three-dimensional projections facilitate the understanding of the three-dimensionality of the tumor and the spatial relationship between the tumor and adjacent structures. Both display modes enhance the diagnostic information yielded by EUS and may improve operation planning in patients with esophageal cancer. In this study, three-dimensional image evaluation allowed accurate classification of tumor penetration and lymph node involvement in four of five patients. Underestimation of the tumor stage occurred in only one patient with infiltration of the chest wall. This was probably due to the limited penetration depth of the transducer. Miniprobes with a frequency of 7.5 MHz may be more appropriate for the evaluation of advanced esophageal cancer [8]. In conclusion, this system enables the acquisition of accurate 3D data sets on the basis of standard ultrasonic images. One of its major advantages is that a conventional ultrasound probe can be used. Three-dimensional image postprocessing improves the staging of esophageal cancer with EUS because additional scan planes and 3D views can be assessed. Original volume data are stored on a hard disk. Real-time reexamination of the data can be done without loss of information at any time. This method may increase the diagnostic confidence in borderline situations that are difficult to assess by conventional examinations (e.g., uT1 versus uT2). Furthermore, computer simulations of realtime examinations promise to offer new perspectives for medical training.

References
1. Dittler HJ, Siewert JR (1993) Role of endoscopic ultrasonography in esophageal carcinoma. Endoscopy 25: 156161 2. Evans JL, Ng KH, Wiet SG, et al. (1995) Accurate three-dimensional reconstruction of intravascular ultrasound data. Circulation 93: 567 576 3. Fishman EK, Magid D, Ney DR, Chaney EL, Pizer SM, Rosenman JG, Levin DN, Vannier MW, Kulman JE, Robertson DD (1991) Threedimensional imaging. Radiology 181: 321337 4. Hamper UM, Trapanotto V, Sheth S, DeJong MR, Caskey CI (1994)

994 Three-dimensional US: preliminary clinical experience. Radiology 191: 397401 Hu nerbein M, Below C, Schlag PM (1996) Three dimensional endorectal ultrasonography for staging of obstructing rectal cancer. Dis Colon Rectum 39: 636642 Hu nerbein M, Dohmoto M, Rau B, Schlag PM (1996) Endosonography and endosonography guided biopsy of upper GI-tract tumors using a curved array echoendoscope. Surg Endosc 10: 12051209 Hu nerbein M, Schlag PM (1997) 3D-endosonography for staging of rectal cancer. Ann Surg (in press) Kallimanis G, Garra BS, Tio TL, Krasner B, al Kawas FH, Fleischer DE, Zeman RK, Nguyen CC, Benjamin SB (1995) The feasibility of three-dimensional endoscopic ultrasonography: a preliminary report. Gastrointest Endosc 41: 235239 9. Mueller MP, Stamos MJ, Cavaye DM, Kopchok GE, Laas TE, White RA (1992) Three-dimensional transrectal ultrasound preliminary patient evaluation. J Laparoendosc Surg 2: 223227 10. Roesch T (1994) Endoscopic ultrasonography. Endoscopy 26: 148 168 11. Ro sch T, Lorenz R, Zenker K (1992) Local staging and assessment of resectability in carcinoma of the esophagus, stomach and duodenum by endoscopic ultrasonography. Gastrointest Endosc 38: 460467 12. Souquet JC, Napoleon B, Pujol B, Keriven O, Ponchon T, Descos F, Lambert R (1994) Endoscopic ultrasonography in the preoperative staging of esophageal cancer. Endoscopy 26: 764766 13. Zonnefeld FW, Lobregt S, van der Meulen JHC (1989) Threedimensional imaging in craniofacial surgery. World J Surg 13: 328 342

5. 6. 7. 8.

News and notices

Surg Endosc (1997) 11: 10591061

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

New Address for the European Association for Endoscopic Surgery (E.A.E.S.)
Effective January 1, 1997, the new correspondence, telephone, and fax numbers of the E.A.E.S. office are: E.A.E.S. Office, c/o Mrs. Ria Palmen Luchthavenweg 81 Unit 1.42 5657 EA Eindhoven The Netherlands or: P.O. Box 335 5500 AH Veldhoven The Netherlands Tel: +31 40 2525288 Fax: +31 40 2523102

Essentials of Laparoscopic Surgery Surgical Skills Unit University of Dundee Scotland, UK

Under the direction of Professor A. Cuschieri the Surgical Skills Unit is offering a three-day practical course designed for surgeons who wish to undertake the procedures such as laparoscopic cholecystectomy. This intensely practical program develops the necessary operating skills, emphasizes safe practice, and highlights the common pitfalls and difficulties encountered when starting out. Each workshop has a maximum of 18 participants who will learn both camera and instrument-manipulation skills in a purpose-built skills laboratory. During the course there is a live demonstration of a laparoscopic cholecystectomy. The unit has a large library of operative videos edited by Professor Cuschieri, and the latest books on endoscopic surgery are on display in our Resource area. Course fee including lunch and course materials is $860. For further details and a brochure please contact: Julie Struthers, Unit Co-ordinator Surgical Skills Unit Ninewells Hospital and Medical School Dundee DD1 9SY Tel: +44 382 645857 Fax: +44 382 646042

Volunteer Surgeons Needed Northwestern Nicaragua Laparoscopic Surgery Teaching Program, Leon, Nicaragua
Volunteer surgeons are needed to tutor laparoscopic cholecystectomy for this non-profit collaboration between the Nicaraguan Ministry of Health, the National Autonomous University of Nicaragua, and Medical Training Worldwide. The program consists of tutoring general surgeons who have already undergone a basic laparoscopic cholecystectomy course. Medical Training Worldwide will provide donated equipment and supplies when needed. For further information, please contact: Medical Training Worldwide Ramon Berguer, MD, Chairman Tel: 707-423-5192 Fax: 707-423-7578 e-mail: berguer.r@martinez.va.gov

Advanced Endoscopic Skills Surgical Skills Unit University of Dundee Scotland, UK

Each month Professor Cuschieri Surgical Skills Unit offers a 412 day course in Advanced Endoscopic Skills. The course is intensely practical with hands on experience on a range of simulated models. The program is designed for experienced endoscopic surgeons and covers advanced dissection techniques, extracorporeal knotting techniques, needle control, suturing, internal tying technique, stapling, and anastomotic technique. Individual workstations and a maximum course number of 10 participants allows for personal tuition. The unit offers an extensive collection of surgical videos and the latest books and publications on endoscopic surgery. In addition, participating surgeons will have the opportunity to see live advanced laparoscopic and/or thoracoscopic procedures conducted by Professor Cuschieri and his team. The course is endorsed by SAGES. Course fee including lunch and course materials is $1850. For further details and a brochure please contact: Julie Struthers, Unit Co-ordinator Surgical Skills Unit Ninewells Hospital and Medical School Dundee DD1 9SY Tel: +44 382 645857 Fax: +44 382 646042

Fellowship in Minimally Invasive Surgery George Washington Medical Center Washington, DC USA
A one-year fellowship is being offered at the George Washington University Medical Center. Interested candidates will be exposed to a broad range of endosurgical Education and Research Center. Active participation in clinical and basic science research projects is also encouraged. For further information, please contact: Carole Smith 202-994-8425 or, send curriculum vitae to: Dr. Jonathan M. Sackier Director of Endosurgical Education and Research George Washington University Medical Center Department of Surgery 2150 Pennsylvania Avenue, N.W. 6B-417 Washington, DC 20037, USA

The Practical Aspects of Laparoscopic Fundoplication Surgical Skills Unit University of Dundee Scotland, UK
A three-day course, led by Professor Cuschieri, designed for experienced laparoscopists wishing to include fundoplication in their practice. The

1060 course covers the technical details of total and partial fundoplication using small group format and personal tuition on detailed simulated models. There will be an opportunity to observe one of these procedures live during the course. Maximum course number is six. Course fee including lunch is $1850. For further details and a brochure please contact: Julie Struthers, Unit Co-ordinator Surgical Skills Unit Ninewells Hospital and Medical School Dundee DD1 9SY Tel: +44 382 645857 Fax: +44 382 646042 egory 1 of the Physicians Recognition Award of the American Medical Association. These programs are also endorsed by the Society of Gastrointestinal Endoscopic Surgeons (SAGES). For further information, please contact: Wanda Toy, Program Administrator Microsurgery & Operative Endoscopy Training (MOET) Institute 153 States Street San Francisco, CA 94114, USA Tel: (415) 626-3400 Fax: (415) 626-3444

Courses at George Washington University Endosurgical Educational and Research Center

George Washington University Endosurgical Educational and Research Center is proud to offer a wide range of surgical endoscopy courses. These courses include advanced laparoscopic skills such as Nissen fundoplication, colon resection, common bile duct exploration, suturing, as well as subspecialty courses. Individual surgeons needs can be met with private tuition. The Washington D.C. area is a marvelous destination to visit for recreational pursuits which can be arranged by the facility to suit your personal agenda. For further details please contact: Carole Smith: Department of Surgery 2150 Pennsylvania Avenue NW 6B Washington, DC 20037, USA Tel: (202) 994-8425

Courses at WISE Washington Institute for Surgical Endoscopy Washington, DC, USA
The Washington Institute of Surgical Endoscopy is pleased to offer the following courses: Laparoscopic antireflux and hiatal hernia surgery (July 1415, 1997); Laparoscopic management of the common bile duct and difficult cholecystectomy (May 1516, August 1112, November 1011, 1997); Laparoscopic colon and rectal surgery (June 2021, September 1516, December 45, 1997). Also, courses for operating room nurses and technicians will be run on a monthly basis and personal instruction and preceptorship is available. For further information, please call: Carole Smith Washington Institute of Surgical Endoscopy 2150 Pennsylvania Avenue, N.W. Washington, DC 20037 Tel: 202-994-9425

Courses at the Royal Adelaide Centre for Endoscopic Surgery

Basic and Advanced Laparoscopic Skills Courses are conducted by the Royal Adelaide Centre for Endoscopic Surgery on a regular basis. The courses are limited to six places to maximize skill development and tuition. Basic courses are conducted over two days for trainees and surgeons seeking an introduction to laparoscopic cholecystectomy. Animal viscera in simulators is used to develop practical skills. Advanced courses are conducted over four days for surgeons already experienced in laparoscopic cholecystectomy who wish to undertake more advanced procedures. A wide range of procedures are included, although practical sessions can be tailored to one or two procedures at the participants request. Practical skills are developed using training simulators and anaesthetised pigs. Course fees: $A300 ($US225) for the basic course and $A1,600 ($US1,200) for the advanced course. For further details and brochure, please contact: Dr. D. I. Watson or Professor G. G. Jamieson The Royal Adelaide Centre for Endoscopic Surgery Department of Surgery Royal Adelaide Hospital Adelaide SA 5000 Australia Tel: +61 8 224 5516 Fax: +61 8 232 3471

Call for Abstracts Society of American Gastrointestinal Endoscopic Surgeons (SAGES) 1998 Annual Meeting April 14, 1998 Seattle, WA, USA
Abstract deadlines: Oral and Poster abstracts: September 12, 1997 Video Submissions: September 18, 1997 For further information, or to obtain an abstract form, please contact: SAGES Program Committee Society of American Gastrointestinal Endoscopic Surgeons Suite #3000 2716 Ocean Park Boulevard Los Angeles, CA 90405 Tel: (310) 314-2404 Fax: (310) 314-2585 e-mail: SAGESMail@AOL.com

European Course on Laparoscopic Surgery (English language) November 1821, 1997 Brussels, Belgium
Course director: G.B. Cadiere For further information, please contact: Administrative Secretariat Conference Services s.a. Avenue de lObservatoire, 3 bte 17 B-1180 Bruxelles Tel: (32 2) 375 16 48 Fax: (32 2) 375 32 99

Advanced Laparoscopic Suturing and Surgical Skills Courses MOET Institute San Francisco, CA, USA
Courses are offered year-round by individual arrangement. The MOET Institute is accredited by the Accreditation Council for Continuing Medical Education (ACCME) to provide continuing medical education for physicians and designates these CME activities for 2040 credit hours in Cat-

1061

Courses Offered at the University of Minnesota Minneapolis, Minnesota, USA September 17, 1997: Fourth Annual Conference, Molecular Biology of Colorectal Cancer September 17, 1997: Sixth Annual Conference, Endorectal Ultrasonography September 1820, 1997: Sixtieth Annual Conference, Principles of Colon and Rectal Surgery
For further information, please contact: Continuing Medical Education University of Minnesota 615 Washington Avenue SE, Suite 107 Minneapolis, MN 55414 Tel: 800-776-8636 Fax: (612) 626-7766

nationally represented and includes leading experts in the field. Simultaneous Spanish and Italian translation is available. For more information, please contact: Cleveland Clinic Florida Department of Education 2950 West Cypress Creek Road Fort Lauderdale, FL 33309-1743 Tel: 800-359-6101, ext. 6066 Fax: 954-978-5539

6th World Congress of Endoscopy Surgery Roma 98 6th International Congress of European Association for Endoscopic Surgery June 36, 1998 Rome, Italy
The program will include: the latest, original high quality research; symposia; plenary lectures; abstract presentations (video, oral, and posters); EAES and SAGES postgraduate courses, OMED postgraduate course on therapeutic endoscopy; working team reports; educational center and learning corner; meeting of the International Society of Nurses and Associates; original and non original scientific reports; and a world expo of new technology in surgery. For further information, please contact: Congress Secretariat: Studio EGA Viale Tiziano, 19 00196 Rome, Italy Tel: +39 6 322-1806 Fax: +39 6 324-0143

Colorectal Disease in 1998 February 1921, 1998 Fort Lauderdale, FL, USA Symposium Director: Steven D. Wexner, MD
Cleveland Clinic Florida presents its ninth annual postgraduate course. Provides an intensive, in-depth, analytical review of all aspects of colorectal disease, including laparoscopy; colorectal carcinoma screening and genetics, inflammatory bowel disease; and pouch surgery. There will be a review of both basic and advanced principles of diagnosis and management of disease. Video techniques will be shown as well. The faculty is inter-

Surg Endosc (1997) 11: 1057

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

EAES ductal stone study

We congratulate the authors for this prospective randomized trial comparing two-stage vs single-stage management of ductal stones (CBDS) in the era of laparoscopic cholecystectomy. Nevertheless, we look at the data and some of the literature reportspartly not mentioned in this papera little differently. As stated in this paper, the participating surgeons had varying experience in laparoscopic stone clearance. The same might apply for the experience of endoscopists in dealing with poor results (80%) on flexible endoscopic stone extraction (ESE), as Cotton [4], Rieger [8] Kra henbu hl and Bu chler [6], Bonatsos [2], Coppola [3], Boeckl [1], and Sungler [9], documented an extraction rate of 95% up to 100%. The rate of unnecessary endoscopic retrograde cholangiography (ERC) in group A is fairly low, as only the presence of stones is mentioned and no other pathology like benign papillary stenosis or signs of passed stones are recorded. Group B encountered 17% more CBDS, a fact raising concern and suggesting that at least some of these stones were air bubbles on intraoperative cholangiogram and consequently easy to treat laparoscopicallyas already concluded earlier by McSherry on a similar occasion [7]. With ESE one or more stones are virtually and directly visualized on extraction; in laparoscopic duct clearance at least some of the suspected stonesair bubbles are only flushed or pushed through the papilla without evidence of their substantial nature. Changing only the criticized data for ESE in group A and the numbers and implications of the questionable cholangiograms in group B in Table 6 [5], the results and conclusions substantially change and definitely favor the twostage procedure, not to mention the conversion rate of 16% in group B. Thus the role of ERC should not change to selective use after laparoscopic cholecystectomy, as the prospects of postoperative ESE in the operating room at the end of a frustrating laparoscopic procedure is a nightmare for every busy unit, as one operating room might be blocked for hoursnot to mention whether an experienced endocopist would be available at that moment. The authors goalnamely, to accomplish all the necessary biliary procedures at the time of laparoscopic surgeryis most desirable, but so far not at all proven to be superior to a two-stage procedure. Laparoscopic common bile duct exploration requires additional instrumentation and a caseload to acquire the skills; otherwise, the incidence of injury will be unacceptably high [7]. The same applies for ESE, and the question arises, where surgical endoscopists should get their routine and experience, when only called into the operating theaters for the most difficult last option procedures, if laparoscopy has failed. We therefore conclude that the therapeutical splitting with selective preoperative ERC is certainly at present the standard procedure and laparoscopic duct exploration is limited to incidental cases with false-negative preoperative diagnostics. P. Sungler O. Boeckl
Surgical Department Mu llner Hauptstr. 48 A-5020 Salzburg, Austria

References
1. Boeckl O, Sungler P, Heinerman PM, Lexer G (1994) Choledocholithiasistherapeutical splitting. Chirurg 65: 424429 2. Bonatsos G, Leandros E, Polydorou A, Romanos A, Dourakis N, Birbas C, Golematis B (1996) ERCP in association with laparoscopic cholecystectomy. Surg Endosc 10: 3740 3. Coppola R, DUgo D, Ciletti S, Riccioni ME, Cosentino L, Magistrelli P, Picciocchi A (1996) ERCP in the era of laparoscopic biliary surgery. Surg Endosc 10: 403406 4. Cotton PB (1984) Endoscopic management of bile duct stones (apples and oranges). Gut 25: 587597 5. Cuschieri A, Croce E, Faggioni A, Jakimovicz J, Lacy A, Lezoche E, Morino M, Ribeiro VM, Toouli J, Visa J, Wayand W (1996) EAES ductal stone study. European Association of Endoscopic Surgeons (EAES) Ductal Stone Co-operative Group. Surg Endosc 10: 11301135 6. Kra henbu hl L, Bu chler MW (1996) Invited commentary to: laparoscopic cholecystectomyhow to diagnose common bile duct stones? Acta Chir Austriaca 28: 101102 7. McSherry CK, Salky BA (1994) Laparoscopic management of common bile duct stones. Editorials. Surg Endosc 8: 11611162 8. Rieger R, Sulzbacher H, Woisetschla ger R, Schrenk P, Wayand W (1994) Selective use of ERCP in patients undergoing laparoscopic cholecystectomy. World J Surg 18: 900904 9. Sungler P, Heinerman PM, Mayer F, Boeckl O (1993) Laparoskopische Cholecystektomie bei CholecystoCholedocholithiasis. Therapeutisches Splitting oder konventionelles chirurgisches Vorgehen? Chirurg 64: 10121017

Case reports
Surg Endosc (1997) 11: 10291031

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

Herniation of the stomach and necrotizing chest wall infection following laparoscopic Nissen fundoplication
A. Viste,1 H. Vindenes,2 S. Gjerde3
1 2 3

Department of Surgery, University of Bergen, Haukeland Sykehus, N-5021, Norway Department of Plastics and Reconstructive Surgery, University of Bergen, Haukeland Sykehus, N-5021 Bergen, Norway Department of Anaesthesiology and Intensive Care Medicine, University of Bergen, Haukeland Sykehus, N-5021 Bergen, Norway

Received: 12 April 1996/Accepted: 26 November 1996

Abstract. This paper addresses gastric herniation following laparoscopic fundoplication for reflux esophagitis. Case history: A 46-year-old woman underwent Nissen fundoplication. Two days postoperatively she developed gastric herniation and perforation with subsequent pleural effusion and necrotizing fasciitis of the chest wall. A patent crural repair might reduce the occurrence of paraoesophageal herniation. Key words: Oesophagitis Laparoscopy Nissen Paraoesophageal hernia Necrotizing fasciitis

The efficacy of laparoscopic Nissen fundoplication in controlling reflux disease with acceptable perioperative morbidity and mortality has been documented in several studies [3, 6]. Excellent or good results are obtained in about 90% of patients and the morbidity is generally low, dysphagia being the most common complaint [13, 14]. The most serious complications are associated with herniation of the stomach and perforations of the esophagus or cardia [6, 12]. Whether the frequency of these complications is increased when the procedure is performed laparoscopically has been an open question. The purpose of this paper is to focus on acute gastric herniation with perforation and subsequent necrotizing fasciitis of the chest wall following laparoscopic Nissen fundoplication.

Case history
A 46-year-old substantially overweight but otherwise healthy woman had for several years suffered from an advanced reflux esophagitis marginally effected by medical treatment. Preoperative workup showed esophagitis grade II (Savary-Miller), hiatal hernia, lower esophageal sphincter pressure

of 6 mm, and a 17% total time of pH less than 4. She was admitted to the hospital for surgery and a laparoscopic Nissen fundoplication was performed. The operation was uneventful; operating time was 145 min. Minimal dissection was performed in the lower mediastinum. The upper short gastric vessels were divided, the opening in the hiatus was judged to be acceptable, and the hiatus was not closed. The immediate period after the operation was normal until the patient 2 days later suddenly experienced intense pain in the chest. A plain chest x-ray examination revealed an effusion in the left hemithorax. Foul-smelling fluid was drained and a 32-French drain was inserted. Antibiotic therapy with cefuroxime, netilmicin, and metronidazole was started immediately. X-ray examination of the esophagus and stomach with water-soluble contrast medium showed no leak. The next morning the patient had septic temperature and leucocytosis and needed ventilatory support and vasopressors. Laparotomy was undertaken and revealed that most of the stomach including the whole fundoplication was positioned in the mediastinum. The greater curvature was partly necrotic with perforation. The fundoplasty was taken down and the greater curvature was resected and closed. At this time pink erythema was observed around the drain exit on the chest wall. This erythema spread rapidly, the patient developed lactacidosis, and 1 day later an incision along the opening of the drain showed necrotic fascia. Bacteriological culture showed streptococci type C. At this time the erythema included the left hemithorax, the left side of the abdomen, and the left side of the back. Wide incisions were performed and necrotic fascia and muscles were excised (Figs. 1 and 2). The next day the erythema had spread further and a second-look operation with even wider incisions was performed, but without finding more necrotic tissue (Fig. 3). During the following days the patient developed multiple organ failure affecting lungs, circulation, liver, and kidneys. Due to severe problems in achieving a satisfactory ventilation the patient had a tracheostomy and was ventilated in a prone position. Two weeks later she had a wound rupture of the abdominal incision which was closed and another week later the jejunostomy catheter dislocated to the peritoneal cavity and she required another reoperation. She developed an immense catabolism and was treated with growth hormone 24 IE daily for 34 days. Later on she required several split skin graftings combined with a transposition of the major pectoral muscle to cover the intercostal muscle defect. Five months following the primary operation she is still at a rehabilitation center and is improving gradually (Fig. 4).

Discussion Laparoscopic operations for reflux disease is now a well established procedure, although technical details like mobi-

Correspondence to: A. Viste

1030

Fig. 1. Skin erythema prior to operation for necrotizing fasciitis. Fig. 2. Necrotic fascia in the left flank 3 days after fundoplication. Fig. 3. Left side of truncus after soft tissue debridement. Fig. 4. Left side of thorax 212 months later.

lization of the short gastric vessels and closure of the hiatus are still debated. Herniation of the fundic wrap into the mediastinum is reported in several series following open surgery [11, 12], and is probably caused by episodes of extreme increase in intraabdominal pressurefor instance, vomiting [3, 7]. We have no reports of such episodes in the actual patient. Mobilization of the short gastric vessels may increase the susceptibility of herniation of the stomach into the chest as well as reduce the circulation of the great curvature. Division of these vessels is, however, considered to be of great importance in order to obtain a floppy fundoplication [1, 4, 14]. The importance of closing the hiatus as part of the operation for reflux disease has been debated for years. Al-

though no comparing studies are available, several authors now advocate the importance of approximation of the crural muscles [6, 12]. Our current practice is also to close the hiatus with one or two unabsorbable sutures. Necrotizing soft-tissue infections usually involve the abdominal wall, perineum, or lower extremities [5, 10] but they have also been reported following chest tube insertions and esophageal surgery [2, 9]. So far, we are not aware of any cases published following laparoscopic surgery. Management includes fluid infusions, cardiorespiratory support, broad-spectrum antibiotics, and extensive surgical debridement of necrotic tissue. As thoracic muscles are sacrificed during debridement a specific problem arises concerning coverage of the chest wall defect. Muscle flap closure will still be the procedure of choice as prosthetic closure risks

1031

further infectious complications. The role of hyperbaric oxygen therapy is debated but should always be secondary to surgery [5, 8]. In conclusion, this case should remind surgeons treating patients with laparoscopic procedures that these so-called minimal invasive procedures may in some cases inflict major complications on the patients. Treatment of these critically ill patients should involve close cooperation between the surgeon, anesthesiologist, plastic surgeon, and specialist in infectious diseases.

References
1. Cadiere GB, Houben JJ, Bruyns J, Himpens J, Panzer JM, Gelin M (1994) Laparoscopic Nissen fundoplication: technique and preliminary results. Br J Surg 81: 400403 2. Chen YM, Wu MF, Lee PY, Su WJ, Perng RP (1992) Necrotizing fasciitis: is it a fatal complication of tube thoracostomy?Report of three cases. Respir Med 86: 249251 3. Coster DD, Bower WH, Wilson VT, Butler DA, Locker SC, Brebeck RT (1995) Laparoscopic Nissen fundoplicationa curative, safe, and cost-effective procedure for complicated gastroesophageal reflux disease. Surg Laparosc Endosc 2: 111117 4. DeMeester TR, Bonavina L, Albertucci M (1986) Nissen fundoplication for gastroesophageal reflux disease. Ann Surg 204: 920

5. Gozal D, Ziser A, Shupak A, et al. (1986) Necrotizing fasciitis. Arch Surg 121: 233235 6. Hinder RA, Filipi CJ, Wetscher G, Neary P, DeMeester TR, Perdikis G (1994) Laparoscopic Nissen fundoplication is an effective treatment for gastroesophageal reflux disease. Ann Surg 220: 472483 7. Johansson B, Glise H, Hallerback B (1995) Thoracic herniation and intrathoracic gastric perforation after laparoscopic fundoplication. Surg Endosc 9: 917918 8. Kaiser RE, Cerra FB (1981) Progressive necrotizing surgical infections: a unified approach. J Trauma 21: 349355 9. LoCicero J, Vanecko RM (1985) Clostridial myonecrosis of the chest wall complicating spontaneous esophageal rupture. Ann Thorac Surg 40: 396397 10. McHenry CR, Piotrowski JJ, Petrinic D, Malangoni MA (1995) Determinants of mortality for necrotizing soft-tissue infections. Ann Surg 221: 558565 11. Takahashi T, McElvein RB, Aldrete JS (1994) Esophageal hiatus hernia after fundoplication. Am Surg 60: 869871 12. Watson DI, Jamieson GG, Devitt PG, Matthew G, Game PA (1995) Paraoesophageal hiatus hernia: an important complication of laparoscopic Nissen fundoplication. Br J Surg 82: 521523 13. Watson DI, Jamieson GG, Devitt PG, Matthew G, Britten-Jones RE, Game PA, Williams RS (1995) Changing strategies in the performance of laparoscopic Nissen fundoplication as a result of experience with 230 operations. Surg Endosc 9: 961966 14. Weerts JM, Dallemagne B, Hamoir E, Demarche M, Markiewicz S, Jehaes C, Lombard R, Demoulin JC, Etienne M, Ferron PE, Fontaine F, Gillard V, Delforge M (1993) Laparoscopic Nissen fundoplication: detailed analysis of 132 patients. Surg Laparosc Endosc 3: 359364

Surg Endosc (1997) 11: 10401044

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

The sensitivity of new color systems in blood-flow diagnosis

The maximum entropy method and angio-color-comparative in vitro flow measurements to determine sensitivity
C. Sohn,1 H. P. Weskott2
1 2

Department for Prenatal and Gynecological Ultrasound Diagnosis and Therapy, Clinic of Gynecology, University of Heidelberg, Germany Medical Clinic II, Siloah City Hospital, Hannover, Germany

Received: 9 December 1996/Accepted: 11 March 1997

Abstract Background: Two new blood-flow-diagnosis techniques have recently been developed as supplements to the established color techniques: the MEM (maximum entropy method) technique and color flow amplitude (power Doppler). These are capable of representing blood flow in distinctly more slowly flowing areas than is possible with the conventional Doppler technique. Methods: Both methods make use of the Doppler technique in part, yet analyze the reflected signal in a different manner, in so doing optimizing the relation between the noise and the signal. Measurements were obtained on two anatomic flow models to test the sensitivity of both techniques under slow flow conditions. Results: The slowest flow the MEM technique was capable of recording was 0.5 mm/s, whereas the angiotechnique was able to detect 0.4 mm/s, albeit utilizing a 5-MHz transducer for the MEM technique and a 10-MHz transducer for the angio technique. One may thus assume that the MEM technique would be still more sensitive when utilizing a 10MHz transducer. The advantage of the MEM technique is its real-time flow representation: The angio technique requires a few seconds of acquisition time. This could have serious consequences during clinical utilization. Doppler sonography was merely capable of detecting a minimum flow velocity of 15 mm/s. The angio technique is less dependent on the angle during flow representation than the MEM technique. Conclusions: The distinctly higher sensitivity of these two new color techniques offers new possibilities in the clinical sector as far as the perfusion diagnosis of organs and tumors is concerned.

Key words: MEM technique Color flow amplitude Sensitivity to slow blood velocities

Correspondence to: C. Sohn Universita tsfrauenklinik, Voss-Str. 9, D69115 Heidelberg, Germany

Blood-flow diagnosis of tumors aims to detect neovascularization, an important criterion in determining the malignancy of pathological changes [9, 10]. Neoangiogenesis leads to the formation of the minutest of new vessels without a tunica. This results in extremely low vascular resistance within this vascular system. A low vascular resistance in turn causes a slow blood flow. The detection of the blood flow in these minutest of vessels which develop as a result of the neovascularization is desirable in tumor blood-flow diagnosis. Further problems are concerned with the diagnosis of the perfusion of organs. The conventional Doppler technique used to date is not capable of disclosing blood flow in these minute anatomic structures, the blood flow being too slow for the Doppler to detect [5, 8, 9]. Two sonographic color techniques to determine blood flow, which aim to notably improve the detection limit for slow-flowing blood, have been available for some time nowthe MEM technique (maximum entropy method) and the color flow amplitude (synonyms: power Doppler, Doppler amplitude ultrasound) [11]. Both techniques make use of the Doppler principle yet are capable of detecting distinctly slower blood flow thanks to the evaluation algorithms [1, 6, 14]. The higher sensitivity in the slow flow range can be theoretically deduced. Futhermore, first clinical experience exemplifies this distinctly greater sensitivity [811]. Reliable data concerning the minimum blood flow velocity which is in fact determinable is, however, not yet available. Experimental examinations were conducted on anatomic flow models in order to objectify the assumed higher sensitivity of these new color techniques when compared with the Doppler technique [11]. The sensitivity of the

1041

MEM technique (maximum entropy method)

The maximum entropy method (MEM) is a nonlinear method of spectral estimation originally developed by J. P. Burg. Essentially, the method chooses the spectrum whose time series outside the observed samples is the most random. By maximizing the randomness, or entropy, of the unknown samples the technique marks a nonbiased judgment of the underlying spectrum. Since the spectrum is nonbiased, it also exhibits information at the highest degree of accuracy allowed by the data. MEM actually assesses the distribution of noise or usable information in an ultrasound sample. By accurately modeling the noise in the sample, MEM maximizes the separation of real velocity information from noise. The performance of the MEM algorithm in an acoustically noisy environment is so impressive that it can detect signals which are significantly weaker than the background noise present in the sample. Intense clinical evaluation of the color Doppler spectrum with MEM is now underway by several investigators. Existing color Doppler algorithms and MEM are being evaluated side by side using eight key performance criteria: sensitivity, penetration, spatial resolution, frame rate, accuracy, velocity range, velocity resolution, and unresolved flow (flow in those vessels not visible in two-dimensional imaging). Results thus far have proven that MEM has [14, 7, 1013): Minimized each noise source as required in order to facilitate measurement of low-frequency shifts applicable to very slow flow Extracted the Doppler shift information from a minimum number of sample points in order to improve the color frame rate without sacrificing lateral resolution or color region of interest Extracted Doppler shift information in a low signal-to-noise environment, enhancing the penetration of a color Doppler system Accurately detected Doppler shifts in presence of noise and/or clutter, which is of critical significance Provided signal extraction in the presence of noise with very few time samples and provided an opportunity to increase the spatial resolution without degrading performance in other key areas Given robustness in a noisy environment with minimal Doppler signal strength, which is fundamental in detection sensitivity for a color Doppler system Detected unresolved flow and differentiated Doppler shift echoes from higher-level echoes due to reflections other than blood in the transducer resolution cell First clinical experiences show that MEM is more robust to movement artefacts and system noise while maintaining spectral resolution [3]. As seen in the eight performance criteria, the MEM algorithm has a marked effect on the performance of a color Doppler system, especially in the areas of tumor flow, tissue perfusion, and studies of systemic venous and arterial stenosis. The MEM system from Acoustic Imaging is characterized by 40-MHz data speed, 80-MHz clock speed, a versatile processor, and 128 channels.

Fig. 1. Comparison of principle of power Doppler and conventional color Doppler imaging.

MEM technique was, furthermore, compared and contrasted with that of the power Doppler.

Materials and methods Sonographic angiography/power Doppler (Fig. 1)

Sonographic angiography is a method which gives a real-time color-coded representation of the blood flow superpositioned on the B-mode scan, as is also the case in conventional color Doppler sonography. In contrast to color Doppler sonography, other physical parameters are, however, taken from the reflected ultrasound ray for the colorcoding; the principles on which the color allocation is based thus differ greatly. Sonographic angiography analyzes the amplitude and the color Doppler the deviation in frequency of the reflected, Doppler-shifted signal. The amplitude is determined by the amount and/or density of the hemocytic aggregates registered by the measured volume of the ultrasound beam, and as such it is independent of the angle between the blood flow direction and the incident ultrasound beam, except for wall filter and other Doppler frequency cutoff effects. The deviation in frequency is, on the other hand, determined by the speed of the reflectors [8]. The amplitude signal is not as noisy as that of the deviation in frequency. This characteristic is particularly pronounced for wide-band signals which arise as the sum of numerous frequencies of comparitive intensity. The amplitude is obtained by adding these varying frequency portions; the deviation in frequency is obtained by calculating the mean values which are less well determined with broadband signals. Sonographic angiography, being a less noisy method, can be carried out with greater system gain. More sensitive and continuous vascular representation is hence possible. Signals with a high amplitude often are represented in yellow, with small amplitudes coded dark blue. The noise within the ultrasound signal occurs stochastically over the entire frequency band, but is, however, characterized by a low amplitude. It thus appears as a mosaic-like coloration on the color Doppler, whereas the noise in the sonographic angiography has a particular colornamely, dark blue. This color can be faded out phototechnically. Only the signals with an intensity greater than the noise are then still depicted. Persistent attempts are currently being made to improve the signal-tonoise relationship of sonographic angiography. This method of averaging the signals is distinctly more effective for sonographic angiography than for color Doppler sonography. The reason for this is to be found in the minor dependence of the amplitude on the velocity and the less pulsatile character of the amplitude. The amplitude signals are thus more alike from scan to scan. A further important part of this system is the tissue discriminator, which identifies ultrasound signals which return from the human body as being either an echo from tissue or from flowing blood. Tissue signals are used for B-mode computation and blood signals are separated for sonographic angiography and appropriate further processing. This techniqueunder ideal conditionshas a notably higher sensitivity than color Doppler sonography. Furthermore, it is independent of the angle between the ultrasound wave and the direction of the blood flow; aliasing is not necessary. We used the Angio color system VST Masters (manufactured by Diasonic/Sonotron).

Doppler color flow imaging (Figs. 2 and 3)

By placing multiple sample volumes along the sector of the B-scan image, a two-dimensional distribution of Doppler frequency shifts can be represented. Signals from each sample volume are thus calculated seperately by Doppler spectral analysis. When using the method of the fast fourier transformation (FFT) for the spectral analysis, there are normally 128 plus echo cycles computed for each scan line. From the frequency analysis 128 values per time unit are delivered. There is no possibility for real-time processing because of the time expense for the data aquisition. It is also not possible to represent a Doppler spectrum for each of these sample volumes. Instead the mean flow velocity is calculated and is encoded in color. For this kind of calculation an autocorrelation processing is mostly used. In that process, the Doppler signals recorded at one time are compared (correlated) with the respective recordings. In the easiest case it is possible to create the color Doppler information with only two samples. For each measuring point a vector determined by the amount and the phase angle is calculated. The difference in the phase angles between two or more following measuring points is directly proportional to the Doppler frequency. The calculated Doppler frequencies are encoded in color and superimposed on the B-scan image.

1042

Fig. 2. A Color Doppler imaging: time and velocity of blood flow at one spatial point. B Spatial distribution of mean velocities depending from the time. Fig. 3. Autocorrelation (explained in the text).

There are no binding rules for color assignment. Usually, however, flow toward the transducer is encoded in red and flow away from the transducer in blue. The examinations were conducted utilizing the following equipment: MEM technique: CSA (manufactured by Acoustic Imaging, Phoenix, AZ, USA) Power Doppler technique: VST Masters (Diasonics/Sonotron) Conventional color Doppler technique: same ultrasound systems used before: VST Masters/Diasonics, Sonotraon, and CSA (Acoustic Imaging, Phoenix, AZ, USA) Two anatomic flow models with which a continuous flow was obtained were used: 1. Cardiac Doppler flow model 523 A made by ATS (attenuation coefficient 0.5 dB/cm/MHz; velocity of sound 1,540 m/s; diameter of the lumen 2 and 6 mm). A test solution (density 1.04 0.01 g/ml; viscosity 1.66 0.1 centiStokes; particle size 30 3 m) was pumped through the tubing at a flow rate between 2 and 974 ml/min with a precision of 2% and a reproducibility of 1%. Doppler angle: 45 2. A model constructed by Dr. Weskott [12], Hannover, displaying the following properties: human blood (no longer utilizable concentrated erythrocytes with a PCVpacked cell volumeof 40 g%) was pumped through 10 parallel silicone tubes with an internal diameter of 0.3 mm each. Doppler angle: 45. An electronic perfusor pump was used. A flow rate between 1 and 1,000 ml/min was possible (deviation +4 to 7%). The detected velocities were calculated as follows: 10 tubes of 0.3 mm diameter each gives a total diameter of 3 mm Q flow volume (ml/h) d total diameter (0.3 cm) V mean flow velocity V 4 Q/xd2 A 7.5-MHz linear transducer (Acoustic Imaging, Phoenix) was used for MEM technique examinations, a 10-MHz linear transducer (Diasonics/ Sonotron) for the angio-color technique, and the same linear transducer in both ultrasound systems for conventional color Doppler sonography. The Doppler frequency was, in both systems, 6 MHz. The minimum detectable velocity of the equipment was regulated via the pulse repetition frequency (PRF, in 100-KHz steps). The lowest and highest PRF during which a flow of 0.5 cm was still detectable on the screen were registered. The acquisition time was measured manually in seconds. Seven tests were carried out per instruments for every PRF and the mean values were calculated from the time needed to construct the image.

Results Comparison between the angio and MEM color techniques and the conventional Doppler technique 1. Anatomic tube model with a single tube. Exact flow volumes could be obtained using a pump. Flow medium: concentrated erythrocytes, PCV 40 g%. Penetration depth: 4.5 cm. Diameter of the tube: 6 mm. Criterion for blood flow detection was positive when a flow of 0.5 cm was registered in the tube. With the help of this phantom, blood flow could always be detected with all three color techniques. A difference was registered in the acquisition times. With a velocity of 127 ml/min, blood flow was detected in power Doppler by an acquisition time of between 0.5 and 6 s. With MEM techniques and conventional Doppler technique in all cases blood flow was detected real time. With a velocity of 102 ml/min the power Doppler needed an acquisition time of between 1 and 13 s. MEM technique and conventional Doppler showed the blood flow in real time again. The detection of blood flow in velocities of 81 ml/min, 40 ml/min, and 26.5 ml/min was successful with the MEM technique and conventional Doppler technique real time, while the power Doppler needed acquisition times of between 2 and 6 s. The acquisition times correlated to the different pulse repetition frequencies (PRF) by using the power Doppler are summerized in Table 1. The MEM technique and conventional Doppler technique showed the different blood flow velocities in all cases in real time. 2. An anatomic model for extremely slow flow velocities. Ten tubes with an internal diameter of 0.3 mm each are placed parallel to one another and perfused with living blood. These tubes are placed in a water bath. A precise flow volume with an exactly predetermined velocity can be perfused using an electric pump. The criterion for blood flow detection was positive when a flow of 0.5 cm was registered in the tube.

1043 Table 1. Acquisition time (in seconds) for power Doppler PRF (Hz) 2,500 Flow velocity 127 ml/min: 102 ml/min: 81 ml/min: 40 ml/min: 26.5 ml/min: 2 3 2 2 2 900 0.5 1.5 2 2.5 3 600 1 1 1 3 3.5 300 1.5 3 2 4 4.5 200 2 6 2 3.5 5 100 6 13 5 5 6 Table 2. 45 vs 86 Degree: 45 Color technique MEM technique Angio-color Degree: 86 Color technique MEM technique Angio-color Minimum flow velocity 0.4 mm/s 0.3 mm/s Minimum flow velocity 3.0 mm/s 2.2 mm/s Acquisition time Realtime 8s Acquisition time Realtime 10 s

Slowest discernible velocities Angle between flow direction and direction of sound waves: 45

VST: 0.3 mm/s, acquisition time: 8 s, PRF: 100 kHz MEM: 0.4 mm/s, real time, PRF: 230 kHz Doppler technique: 12 mm/s

Angle between flow direction and direction of sound waves: 86

VST: 2.2 mm/s, acquisition time: 10 s, PRF: 100 kHz MEM: 3.0 mm/s, real time, PRF: 230 kHz Doppler technique: 38 mm/s

Discussion Blood flow diagnosis is increasingly being directed toward the representation of the perfusion of organs and of blood circulation within tumors [5, 9, 10]. This can, however, only be a successful undertaking if the sonographic techniques utilized are capable of discerning the slow flow velocity prevalent in such cases. Conventional Doppler sonography is not capable of doing this satisfactorily, as this study has confirmed by the determination of accurate flow velocities [11]. Both of the two new sonographic color techniques tested here are, however, able to detect extremely slow flow velocities which are distinctly below the detection limit of the Doppler. In this experiment the MEM technique and the angio-color technique could discern flow velocities of less than 1 mm/s, whereas the conventional Doppler technique only detected velocities greater than 15 mm/s. Both new color techniques are thus extraordinarily more sensitive than conventional Doppler sonography. Distinct differences between the MEM and the angio techniques did, however, become evident during the course of this experiment. Blood flow is often already evident on a black and white image (i.e., without color display) using the MEM system: This exemplifies the extremely high resolution of the Bscan [11]. The decisive advantage of the MEM technique was to be found in the real-time demonstration of the blood flow, whereas the angio-color technique required numerous seconds to show the blood flow on the screen. This could lead to severe problems in routine diagnosis; it is unlikely that one would be capable of holding the transducer still for a number of seconds in order to acquire the desired blood flow image. The advantage of the angio-color technique is that it operates to all intents and purposes, independent of the angle; the MEM technique displayed a greater angle

dependence. The angio technique was; furhtermore, also minimally more sensitive than the MEM technique. Conventional Doppler sonography was merely included in this study to demonstrate the differences between the conventional technique and the newly developed ones. Both the MEM and angio-color techniques proved to be approximately 40 times more sensitive than conventional Doppler sonography. In our opinion the following consequences will arise in the future as far as clinical utilization is concerned: One wont be able to talk about color Doppler sonography any more without having to supply more detailed information about the particular technique with which the flow was detected. These new techniques are capable of discerning blood flow in the capillary range, thus also supplying new information concerning organ and tumor perfusion. It is conceivable that the neovascularization of tumors, especially of malignant ones, will be far more sensitively detected in the future. First clinical experience shows that these new color techniques demonstrate a correlation to biological tumor behavior (aggressive or less aggressive), that the diagnosis of transplantation reaction can be carried out earlier and with greater certainty, that venous diagnosis is improved, that the flow diagnosis of lymph nodes is greatly improved as far as the differentiation between infectious and malignant is concerned, that placenta blood flow will be seen well enough for significant clinical assessment, etc. [810]. One must, however, prevent one misunderstanding concerning clinical utilization from arising: the representation of slow blood flow using the appropriate equipment fails to take the quick blood flow into consideration. Color pixels are no longer shown, since other blood flow qualities are being demonstrated. References
1. Bude RO, Rubin JM, Adler RS (1994) Power versus conventional color Doppler sonography: comparison in the depiction of normal intrarenal vasculature. Radiology 192: 777780 2. Burg JP (1967) Maximum Entropy Spectral Analysis, 37th Annual International Meeting, Soc. Explor. Geophys., Oklahoma City, Oklahoma 3. Dougherty G (1994) Spectral analysis of laser Doppler signals in real time using digital processing. Med Eng Phys 16: 3538 4. Haykin S, Kesler S (1976) The complex form of the maximum entropy method for spectra estimation. Proc IEEE 64: 822823 5. Newman JS, Adler RS, Bude RO, Rubin JM (1994) Detection of soft tissue hyperemia. AJR Am J Roentgenol 163: 385389 6. Rubin JM, Bude RO, Carson PL, Adler RS (1994) Power Doppler US: a potentially useful alternative to mean frequency-based color Doppler US. Radiology 190: 853856 7. Shinozuka N, Yamakoshi Y (1993) Measurements of fetal movements

1044 using multichannel ultrasound pulsed Doppler: autorecognition of fetal movements by maximum entropy method. MBEC Kyoto World Congr Suppl 6066 8. Sohn C, Kru nes U, Becker D, Gu nter E, Mutze S, Steinkamp W, Mu ller F, Weskott HP, Gebel M (1995) Mo glichkeiten und Grenzen einer neuen Farbtechnik: die Ultraschall-Angiographie-Ergebnisse des Heidelberger Rundtischgespra ches. Bildgebung 62: 5363 9. Sohn C, Meyberg G (1995) Erste Erfahrungen mit einer neuen Farbtechnik: die Ultraschall-Angiographie. Zentralbl Gyna kol 117: 9096 10. Sohn Ch, Holzgreve W (1995) Ultraschall in Gyna kologie und Geburtshilfe. Thieme 11. Sohn Ch, Weskott HP, Schieer M (1996) Sensitivita t neuer Farbsysteme: Maximum entropy method. Ultraschall Med 17: 138142 12. Strand ON (1977) Multichannel complex maximum entropy method (autoagressive) spectral analysis. IEEE Trans Automatic Control AC22 13. Ulrych TJ, Bishop TN (1975) Maximum entropy spectral analysis and autoagressive decomposition. Rev Geophys Space Phys 13: 185200 14. Weskott HP, Knuth C ( ) Power Doppler Ultraschall: Phantommessungen von langsamen Blutflu ssen. Bildgebung, Imaging Publikation in Druck

Surg Endosc (1997) 11: 10451051

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

Deflectable endoscopic instrument system DENIS

A. Melzer,1 K. Kipfmu ller,2 B. Halfar3
1 2

Institute for Diagnostic and Interventional Radiology, University Witten/Herdecke, 45468 Mu lheim/Ruhr, Germany Department of Surgery, St. Bernward Hospital, Treibestrasse 9, 31134 Hildesheim, Germany 3 Department of Surgery, Sta dtische Kliniken, Beurlhousstrasse 40, 44137, Dortmund, Germany Received: 28 January 1997/Accepted: 19 April 1997

Abstract Background: The degrees of intraoperative movement with rigid standard instruments during laparo-endoscopic surgery are limited to translation, rotation, and pivoting within the insertion point. Additional distal angulation and rotation of the instrument jaws are a potential improvement. Methods: Different types of articulated instruments have been developed and tested in phantom and animal experiments. The final prototype was used on 30 patients during laparoscopic surgery following a standardized test protocol. Results: The final design incorporates elastically linked tubular segments, 0120 variable curvature, and +/360 rotation of the jaws element. All functions can be operated with one hand. Testing on phantom and in laparoscopic surgery showed improved handling of organs and tissue with no complications. Conclusion: We were able to demonstrate the feasibility of the technical design and the clinical applicability of a deflectable endoscopic instrument system. Although our initial results indicate an improvement in laparoscopic tissue manipulation, the current deflection and jaw rotation require further technical refinement. Key words: Laparoscopic surgeryarticulated instrumentsdeflectable instruments

The kinematic system of the human hand and arm allows the surgeon to execute and control the intraoperative orientation of a surgical instrument in eight degrees of freedom (DOF), thus permitting optimal movement and positioning of the instrument for effective tissue-instrument interaction during open surgery. Compared with open surgery, laparoendoscopic surgery has significant disadvantages. The performance of delicate surgery through so small an access

imposes restrictions on instrument manipulation [2, 3, 6]. The intracorporeal function of standard instruments does not translate the actions of the hand as effectively in laparoendoscopy as in open surgery. The invariant point of insertion (IPI) of instruments through the abdominal wall creates a virtual ball and socket joint that allows only four degrees of movement of the functional tip of the instruments within the operative field (see Table 1) (Fig. 1). In endoluminal endoscopy, distal angulation is essential for both diagnosis and therapy. Gastroscopy, colonoscopy, or broncoscopy is impossible without deflection of the distal part of the endoscope. However, the principle of distal angulation of flexible endoscopes cannot simply be transferred to a laparoscopic forceps. The design and construction of deflectable endoscopic instrumentation is constrained by spatial restrictions and the need for circular shafts to enable operations in a low-pressure gas environment. Early prototypes were developed and preliminary animal applications performed in a collaboration between Tu bingen University, Germany, and the Karlsruhe Research Center, Germany [10, 14]. Based on the handling technology developed for nuclear research [5], a simple mechanical manipulator was developed for endoscopic surgery. The instrument permits axial deflection of the tip at +/120 and axial rotation of the tip during angulation. Although we have documented that steerable dexterous instruments facilitate laparo-endoscopic manipulations, problems such as tip deflection, jaw rotation, mechanical stiffness, dissassembly, and cleaning still needed to be solved. This paper describes the development and initial results of experimental and clinical tests of an improved deflectable endoscopic instruments system. Materials and methods
The development process follows a standardized protocol that has been evaluated by Melzer et al. [11]. First, the surgical problem and the related technological issues are defined. An analysis of currently available instrumentation dictates how the development should proceed. Ideas for improvements or innovations are discussed. Based on sketches, a simple wooden or wire model is constructed to establish the feasibility of the idea. A list of general and specific requirements completes the first phase.

Correspondence to: A. Melzer, MRI, University Witten/Herdecke, Schulstrasse 10, 45468 Mu lheim/Ruhr, Germany

1046 Table 1. Intraoperative degrees of freedom of endoscopic instruments Straight instruments axial translation along the insertion line axial rotation of the jaws element pivoting in two directions within the insertion point Axial curved instruments axial translation along the insertion line axial rotation of curved end without rotation of the jaws pivoting in two axes within the insertion point Articulated instruments axial translation along the insertion line axial rotation of the shaft pivoting in two axes within the insertion point distal angulation of the functional Tip +/120 axial rotation of the jaws (effector) +/360

Defining the technological issues Improvement of intraoperative manipulation can be obtained by increasing the degrees of freedom of the functional end of the instrument via a design that permits two additional movements: angulation of the functional end by 120 and 360 axial rotation of the jaws element independent of the distal angulation. Maximal stiffness of the distal section independent of the angulation and intuitive singlehanded operation of the angulation and jaw rotation are also required.

Analysis of current instruments

Fig. 1. The invariant point of insertion (IPI) of laparo-endoscopic instruments decreases their intraoperative maneuverability.

In the second phase, modifications of available instruments or simple prototypes are tested during animal experiments. The design process of a near to product system is started, and technologies such as CAD/CAM/ CNC (computer-aided design, manufacturing, and numerical control), laser cutting and welding, EDM (electric discharge machining) calculation, graphical computer simulation, biocompatibility tests, etc., are used if necessary. Most instruments do not require a specific animal test because the phantom setup with animal tissue in a lap trainer can reproduce the operative environment. Prior to any clinical use, the modified instruments are tested in trainers in order to document their reliable and precise function and to expose any limitations that may require modification of the original specifications. Instruments for clinical use are manufactured according to ISO 9000, 9001, and EN 46001. They are constructed from medical-grade materials such as stainless steel, Nitinol, POM, Teflon, and PEEK. During clinical testing, a protocol is followed that includes workup and sterilization manual as well as instructions for the scrub nurse and assistant surgeons. The surgical application is assessed in different steps. These steps include insertion through the cannula; study of the instrument function without tissue contact, for its intended purpose, and with tissue contact; documentation of the results and of any technical mishaps; and institution of any required improvements. Cleaning and maintenance are assessed separately.

Current instrument design is constrained by the necessity for insertion through a trocar or cannula. Probes, graspers, forceps, and needle drivers all require straight, circular, long shafts. Straight, rigid instruments permit only four degrees of freedom [7, 10, 13].

Curved instruments Although curved instruments do not feature additional degrees of freedom, they allow increased working space (Buess 1984 and Cuschieri 1992) [1, 2, 4, 6]. Axial distal curvature of the functional end in a radius of 25 mm and in 45 up to 90 angulation results in a significant increase of the intraoperative working field of the instrument (Fig. 2). Tubular organs such as the esophagus, intestine, ducts, and vessels can be encircled to facilitate the passage of slings and ligatures. The jaws cannot be rotated axially and are thus not optimally orientated. The insertion of curved instruments requires flexible cannulae [16].

Technical results: Phase 1

Variable curvature instruments The Dundee variable curvature instruments, offered in a reusable form by Storz (Tuttlingen, Germany) and a disposable form by USSC (Norwalk, CT, USA), incorporate a curved preshaped blade that is withdrawn and constrained in an outer sleeve. When the blade is advanced, it gradually recovers the circular shape, thus providing adjustment of the degree of curvature (Fig. 3). The instrument improves the retraction and mobilization of tubular organs [4]. This con-

Defining the surgical problem In laparoscopic and thoracoscopic surgery, the manipulation of internal organs and dissection of tissue are significantly hampered by the limited degrees of freedom of the rigid instruments and the restriction of movement due to the invariant point of insertion.

1047

Fig. 2. Axially curved instruments increase the intraoperative working space but not the degrees of freedom of movement. Fig. 3. The Cuschieri variable curvature spatula facilitates the handling of tubular organs. Fig. 4. Endoflex segmented instruments feature variable angulation between 0 and 90 but no rotation of the axial jaws.

cept has been taken further by Yarvis (Raychem, Menlo Park, CA, USA) to a variable curvature forceps. The Roticulator (USSC, Norwalk, CT, USA) comprises a precurved PTFE-covered (polytetraflourethylene, Teflon) Nitinol tube [12]. The curvature can be adjusted by movement of the outer sleeve to and fro. Although adjustable angulation is achieved, the relatively low degree of stiffness of the thin Nitinol tube limits precise handling. Axial rotation of the jaws is not provided, and the factor of its disposability adds cost to each procedure.

Segmented shaft variable curvature instruments Retractors and forceps with a distal segmented section permit the deployment of a predetermined curved shape within the peritoneal cavity (Surgical Innovations, Leeds, England). The curvable section of the instrument is constructed of six to 15 toothed cylindrical segments aligned by three cables. The two lateral cables are operated by an attached wheel at the handle. Once the instrument is put under tension, the segments are approximated and firmly locked by

1048 Table 2. Angulation elements Angulation element Ball and socket joint Bolt/bore-hinged tube segments Elastic link of tube segments Four-bar links Two-gear connection Toothed rings Internal diameter + + 0 Stiffness/ stability 0 + + + 0 Gradual angulation + ++ + 0 Manufacturing characteristics + 0 0

+ +, very good; +, good; 0, acceptable; , poor; , very poor

the toothed surface of the segments. Different curvatures can be created by the manufacturer by varying the design of each single segment and the angle and position of the teeth. In a collaboration between our group and Surgical Innovations, an additional outer sleeve has been incorporated into the original design. This sleeve can be moved relative to the shaft, thus exposing more or fewer segments and allowing an adjustment of the degree of the distal curvature. The segmented variable curvature retractors are excellent for tissue manipulation because they are relatively stiff and the retraction surface is atraumatic (Fig. 4). Drawbacks include the lack of axial rotation of the jaws and the difficulty in cleaning the segments and handle. Technical results: Phase 2 It follows from the preceding analysis that the solution to the surgical problem is the use of steerable distal deflecting instruments with adjustable axial rotation of the jaws. This type of instrument has the advantage of variable orientation of the distal end and jaws, so that the surgeon can grasp and move the tissues in the most appropriate direction. Important technological issues are the angulation elements and external control mechanism. We have designed and tested a variety of articulating elements and handles. Articulating elements The articulating elements have to feature specific characteristics, such as continued stiffness independent of the deflecting angle, continuous and gradual angulation, and large internal space to carry the push rod for the jaws. A brief technical assessment is presented in Table 2. The realization of deflectable articulated instruments involves a wide variety of linkages, such as ball and socket joints, bolt-hinged tube sections, elastic linkages of tube sections, etc. Deflectable instruments made of ball and socket joints are comprimised by insufficient linkage and stiffness of the elements. Joints made of tube sections connected by bolts are already incorporated in conventional flexible scopes. Although this type of link provides maximum inner diameter, it suffers from wear and tear and does not angulate gradually. Spatial mechanisms shaped like elephant trunks with highly sophisticated four-bar linkages offer optimum control and gradual angulation. Highprecision manufacturing is required for the elements [13]. The drawback is the small inner diameter. Unfortunately, all of these linkages have small gaps that preclude adequate cleaning, which is a mandatory precon-

dition for disinfecting and sterilization. In order to obviate those problems, we have developed designs based on elastically connected angulation segments that rely on the use of superelastic Nitinol. This design incorporates the benefits of gradual angulation, stiffness, and ease of cleaning. Based on these elastic links, the first deflectable instruments for clinical use have been developed and produced in collaboration with EPflex, Dettingen, Germany (Fig. 5). The Deflectable Endoscopic Instrument System (DENIS) employs a modular concept to allow the exchange of different functional tips and total disassembly. HF hook, dissecting forceps, and scissors can be supplied in a wide range of sizes and shapes. The diameter of the current prototype fits in a 7-mm trocar or cannula. External control mechanism and handle The concept of using direct mechanical linkages of movements in contaminated areas is familiar from the development of manipulators in nuclear research [5]. Every pivot of the handle results in an equivalent deflection of the distal end (Fig. 6). This design was discarded because it involves an inergonomic wrist position and the handle can cause interference with the abdominal region. Control of the deflection via separate finger rings and triggers constitutes a better solution. For the final design of the handle, we decided to control the deflection and jaw rotation with two separate wheels (Fig. 7). The jaws are operated by a trigger. The transmission of rotation and deflection is formed by three concentric tubes. This approach provides sealing against gas leak and easy disassembling for cleaning and maintenance purposes. The deflecting element can be exchanged intraoperatively without taking the tubes apart. The central rod is isolated; it is connected to the jaws and transmits the axial rotation during deflection of the distal part. A release bottom allows simple extraction and exchange of the jaws and rod. Experimental and clinical evaluation Preliminary clinical applications of the DENIS system have been performed in phantom trainers in order to establish reliable and precise functions and confirm the preliminary medico-technical specifications. The instruments were then used during laparoscopic cholecystectomy (n 12), appendectomy (n 4), hernia repair (n 6), colonic resection (n 4), and thoracoscopic procedures (n 3). Tests of ease of cleaning and sterilization were done to meet the required hygenical standards. Our clinical experience with a

1049

Fig. 5. a: The Deflectable Endoscopic Instruments System (DENIS) (EPflex, Dettingen, Germany) provides variable angulation between 0 and 120 and axial rotation of the jaws of 360. b: The instrument can be completely disassembled for cleaning and maintenance. Fig. 6. Direct mechanical reflection of the angulation and rotation of the handle provides intuitive control, but it may interfere with the abdominal wall (instrument prototype by FZK, Karlsruhe, Germany). Fig. 7. Angulation and jaw rotation of the deflectable DENIS instrument are controlled via two wheels in the axial handle.

deflectable high-frequency preparation hook, scissors, and dissecting forceps has surpassed expectations. The adjustability of the tip according to the anatomic position, as in open surgery, facilitates dissection procedures and has the potential of enabling a significantly better outcome in less time when the instrument is used in advanced laparoscopic procedures and difficult cholecystectomies. Dexterous instruments generally require more training than simply curved instruments. After 1 h on a phantom trainer, a laparoscopically experienced surgeon can become acquainted with the DENIS system and perform surgical maneuvers correctly.

Future developments An extension of the concept of distal articulation to a miniaturized manipulator system has been realized in collaboration with Daum (Schwerin, Germany). The so-called Endohand and the smaller brother Endofinger are mechanical models based on the human kinematic lower armhandfinger system (Fig. 8). The preliminary experimental applications carried out by B. Helms (Rostock, Germany) and our group are promising. In this application, the hand is placed in an ergonomic position in a special glove that mechanically transmits the surgeons finger and wrist movement through wires and cables to a distal miniaturized three-fingered hand. Thus, the capture and manipulation of tissue is similar to direct hand action. Although the effect is somewhat reduced through the mechanic transmission, the

tactile feedback and kinematic response transmit considerable information on the anatomic structure that is palpated. In the Intelligent Surgical Instrument Systems (ISIS) [9] and the endoscopic manipulator system (ARTEMIS) [15], the deflectable instrument represents the intracorporeal manipulator of the operating system. It is connected to an external manipulator arm with four degrees of freedom that conducts the converted movement of the instrument into the IPI in the abdominal wall. Electric motors at the interface of the external manipulator and the internal section drive the intracorporeal actions. Sensors detect the force, speed, and position of the manipulator arm (Fig. 9). The surgeon operates the system from a man-machine interface where kinematic response and tactile feedback are provided. This manipulator system is currently under development by the Karlsruhe-Tu bingen group. Considerable further research is required before its clinical application becomes a reality.

Conclusions Deflectable endoscopic instruments constructed out of elastically linked elements have been shown to be a potential solution to the problems of effective intraoperative tissue manipulation. These problems emanate from our attempts to replace the direct manual handling of tissue with remote mechanical grasping using long, rigid instruments. Handle displacement in a downward direction leads to a movement

1050 Table 3. Results of clinical applications Surgical procedure Laparoscopic cholecystectomy Appendectomy Hernia repair Colonic resection Thoracoscopic procedures Number 12 4 6 4 3 Techniques Dissection cystic artery, duct, and gallbladder Grasping, dissection, and mobilization of appendix Dissection of spermatic cord, placement of mesh graft Dissection, adhesiolysis, mobilizing and encircling Dissection, adhesiolysis Complications 0 0 0 0 0 Assessment Improved technique Improved technique; instrument too long Improved technique; problems with handle Improved technique; handling difficulties Improved technique; instrument too long

Fig. 8. Endohand (Daum, Schwerin, Germany) transmits the movements of the surgeons hand to a distal miniaturized version of a mechanical hand with three fingers.

ment in 30 laparo-endoscopic surgical procedures indicate that the distal angulation of the instrument, combined with axial rotation of the jaws, greatly facilitates dissecting procedures and mobilization of organs. The angle of the jaw opening and the tissue margins can be correlated to a large extent. The deflectable instruments mimic the actions of the surgeons hand much better than a conventional straight instrument. There are still problems in handling the instrument and operating the deflecting and axial rotation. These shortcomings are related to the design of the handle, which will be modified in the next-generation prototype. During the experimental and clinical tests, we questioned whether electrically driven deflection and axial rotation would lead to improved intuitive handling. However, this refinement would involve additional costs and safety problems. The subject of handle design will be considered at the next evaluation cycle. Variable distal curvatures do not always improve tissue handling. As more geometric positions and angulations of the functional end become possible, the conversion of movements also becomes more difficult, as so does the orientation of the instrument. Thus, special training is mandatory prior to clinical use of deflectable probes and forceps. The next generation of deflectable instruments will have a 5-mm diameter. It will also incorporate an improved handle design with better intuitive operation of distal angulation and axial jaw rotation. Comparative studies with conventional straight and axial curved rigid implements will be performed in our final assessment of the clinical benefit of these new and improved articulated instruments.
Acknowledgment. We are grateful to Bernhard Uihlein for his excellent technical support and the production of the Endodeflector prototypes. We also thank Daum GmbH for providing the Endohand and Endofinger for clinical tests, and we acknowledge Axel Winkel for his technical support.

Fig. 9. The future application of deflectable instruments in an endoscopic manipulator system.

References of the functional end in the opposite direction and vice versa. Although curved instruments increase the working field, the degrees of possible distal movements are still restricted to translation, rotation, and pivoting within the access port. We found the lack of axial orientation of the jaws still not optimal for grasping of tissue borders or sutures, though it certainly represented an improvement over straight instruments. Our initial clinical tests of a 7-mm deflectable instru1. Buess G, Thei R, Hutterer F (1984) Endoskopische Operationen in der Rektumho hle. In: Bue G, Unz F, Pichelmaier H (eds) Endosko rzte Verlag, Ko pische Techniken. Deutscher A ln, pp 132141 2. Cuschieri A (1992) General principles of laparoscopic surgery. In: Cuschieri A, Buess G, Pe rissat J (eds) Operative manual of endoscopic surgery. Springer, Heidelberg, Berlin pp 172175 3. Cuschieri A, Nathanson LK, Shimi SM (1992) Laparoscopic antireflux surgery. In: Cuschieri A, Buess G, Pe rissat J (eds) Operative manual of endoscopic surgery. Springer, Heidelberg, Berlin pp 300 318 4. Cuschieri A, Shimi SM, Banting S, Van Velpen G, Dunkley P (1993)

1051 Coaxial curved instrumentation for minimal access surgery. Endosc Surg Allied Techn 1: 303305 Ko hler GW (1985) Historical survey of manual operated instruments. Robotersysteme I. Springer, Heidelberg Lirici MM, Melzer A, Reuthebuch O, Buess G (1993) Experimental development in colorectal surgery. Endosc Surg Allied Techn 1: 2025 Loeffler M, Trispel S (1993) Technological principles of curved instruments and flexible cannulae. Endocs Surg Allied Techn 1: 36570 Melzer A (1996) Endoscopic instrumentsconventional and intelligent. In: Touli, Hunter J, Gossot F (eds) Endosurgery. Churchill Livingston, London, pp 6995 Melzer A, Schurr MO, Kunert W, Buess G, Voges U, Meyer Y-U et al (1993) Intelligent Surgical Instrument System. Concept and preliminary experimental application of components and prototypes. Endosc Surg Allied Techn 1: 165170. Melzer A, Schurr MO, Dautzenberg P, Trapp R, Buess G (1992) Erho hung der Freiheitsgrade bei Instrumenten fu r die minimal invasive Chirurgie. Langenbecks Arch Chir (Suppl) 377: 279285 11. Melzer A, Buess G, Cuschieri A (1994) Instruments and allied technologies of ES. In: Cuschieri A, Buess G, Pe rrisat J (eds) Operative manual of endoscopic surgery II. Springer, Heidelberg, pp 14 12. Melzer A, Stoeckel D (1995) Using shape memory alloys. Med Dev Techn 5: 1623 13. Mueglitz J, Kunad G, Dautzenberg P, Neisius B, Trapp R (1993) Kinematic problems of manipulators for minimal invasive surgery. End Surg 1: 160164 14. Schurr MO, Melzer A, Dautzenberg P, Trapp R, Buess G (1993) Development of steerable instruments for minimally invasive surgery in modular conception. Acta Chir Belg 93: 7377 15. Schurr MO, Melzer A, Kunert W, Dautzenberg P, Neisius B, Breitwieser W, Buess G (1994) Experimental evaluation of components of an endoscopic manipulator system Artemis. Minimal Invasive Ther (Suppl) 13: 28 16. Shimi SM (1994) Access with flexible cannulae. End Surg 3: 5154

5. 6. 7. 8. 9.

10.

Surg Endosc (1997) 11: 10101012

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

Delayed diagnosis of malignant tumors missed at laparoscopic cholecystectomy

W. Junger,1 W. G. Junger,2 J. Hutter,1 K. Miller,1 E. Moritz1
1 2

Second Surgical Department, Landeskrankenanstalten Salzburg, Mu llner Hauptstrasse 48, 5020 Salzburg, Austria Division of Trauma, Department of Surgery 8236, University of California San Diego Medical Center, 200 West Arbor Drive, San Diego, California 92103, USA Received: 6 October 1996/Accepted: 19 February 1997

Abstract Background: The aim of this study was to compare the significance of routine examinations prior to laparoscopic cholecystectomy (LC) with intraoperative abdominal investigation. Preoperative evaluation becomes increasingly important when laparoscopic procedures are performed for the removal of gallstones because other intraabdominal diseases may coexist in these patients, mimicking biliary tract disease. Methods: Over the last 6 years, we treated 816 patients with symptomatic cholecystolithiasis using LC. Prior to surgery, routine tests such as upper abdominal ultrasonography, chest radiography, and standard laboratory blood tests were carried out. Results: Despite these routine tests, coexisting colonic cancers escaped detection in four out of 816 cases. This indicates a risk of more missed pathologies during the course of laparoscopic operations compared to standard laparotomy. Conclusion: The risk of missing coexisting diseases during laparoscopic operations has to be minimized by placing additional emphasis on careful evaluation of anamnesis. Physical examination and additional laboratory testssuch as analysis of tumor markers and blood in the stool combined with complete abdominal ultrasonography, gastroscopy, and/or complete colonoscopy should be performed prior to LC. Key words: Laparoscopic cholecystectomy Colon cancer Colonoscopy

thiasis. Standards and methods of patient examination prior to cholecystectomy have not changed much since the introduction of LC. Typical symptoms with pain and colic in the right upper quadrant often extending back to the right shoulder indicate the need for further exams including abdominal ultrasonography, various laboratory tests, and chest x-rays. Additional tests are being performed only when the results from the initial tests diverge from our clinical observations. The necessity for manual exploration of abdominal organs in the course of laparotomy is generally recognized [12, 16] and such manual examination would provide an effective and inexpensive alternative to detect tumors. Laparoscopy, however, offers only a two-dimensional picture, and laparoscopic surgery is a nontactile procedure that offers limited access to the abdominal cavity. The surgeon is unable to perform manual examination of the gut during laparoscopic surgery. Thus, abdominal exploration is limited compared to that of open procedures. For LC, surgical techniques similar to those used for conventional surgery have been suggested. In their papers introducing LC, Dubois [7, 8] and Pier [15] described intraoperative circulatory inspection of the abdominal cavity to detect possible coexisting diseases. Siewert [19] and Ko [13] recommended that any changes in traditional surgical techniques and methods should be minimized, as these methods have been deemed dependable and have been used for over 100 years. We think that either intraoperative laparoscopic ultrasound investigations such as recently reported by Cavina [4] and/or meticulous preoperative examinations including endoscopy are minimal requirements to minimize the risk of undetected abdominal cancer during LC.

Laparoscopic cholecystectomy (LC) has become the standard treatment for patients with symptomatic cholecystoli-

Case reports
From June 1990 to March 1996, we performed 816 LCs in our department. All patients presented with typical symptoms of biliary tract disease episodes of colic and pain or tenderness in the right upper quadrant. At our

Correspondence to: W. Junger

1011 Table 1. Patients who had concomitant malignancy Patient no. 67 234 345 526
a b

Age 63 64 71 50

Sex F F M F

Laparotomya 2 weeks 2 months 2 months 15 months

Malignancy Carcinoma of ascending colon,b T3N1M0 Carcinoma of appendix, T4N1M0 Cancer of transverse colon,b T3N0M0 Cancer of transverse colon,b T4NxM1

Time of laparotomy after LC Laparotomy for acute bowel obstruction

department, we perform LC only in patients who suffer from symptomatic cholecystolithiasis. Prior to surgery, the patients history was taken, and routine tests including upper abdominal ultrasonography, radiograph of the chest, and standard laboratory blood tests (red and white blood counts, erythrocyte sedimentation rate, blood clotting, serum electrolytes, and liver function tests) were carried out. In all cases ultrasonographic evaluation revealed stones in the gallbladder. During LC, circulatory inspection of visible intraabdominal organs was routinely performed and intraoperative cholangiography was attempted in all cases.

Case 4
A woman 50 years of age was admitted with episodes of cramps and abdominal pain in the right quadrant 24 h prior to admission. With ultrasound we noted acute hydrops of the gallbladder and the common bile duct was found to be dilated up to 15 mm in diameter without calculi. Laboratory values showed a moderately elevated white blood cell count. An endoscopic retrograde cholangiopancreatography (ERCP) was carried out. A severe hydrops of gallbladder obstructed the distal common bile duct; no stones could be detected. On the next day, the patient underwent LC without complications and was discharged 4 days later. Fifteen months later she was returned to our unit because of nausea and severe vomiting. Clinical and radiographic examination of abdomen showed complete small-bowel obstruction. She underwent laparotomy. The entire abdominal cavity was covered with peritoneal metastases. No resection or palliative enteral anastomosis was performed and the patient died 1 day later. Autopsy revealed cancer of the transverse colon with extensive peritoneal seeding. Histology of the tumor: Adenocarcinoma (T4N M1).

Case 1
A 63-year-old woman with a history of cholecystolithiasis and repeated pain in the right upper quadrant during the last 1 year was admitted to our hospital. Defecation was reported to occur without problem. Laboratory values including CEA (carcinoembryogenic antigen) were found to be normal. Ultrasonographic examination revealed gallstones and bilateral renal cysts. LC was without complication; however, after 3 days, clinical and radiograph examination showed small-bowel paresis. Conservative therapy allowed the discharge of the patient on day 15 after LC. Two weeks later, the patient was readmitted with symptoms of complete bowel obstruction. Laparotomy revealed obstruction of ascending colon by colon cancer. Right colectomy was performed. Histology: Adenocarcinoma (T3N1M0).

Discussion Few reports have been published on additional findings by routine abdominal exploration during LC [1, 3, 5]. Little consideration has been given to the possibility of missing additional intraabdominal pathologies during the course of LC. In recent years sporadic reports have described cases of delayed diagnosis of malignant tumors that evaded detection during LC [2, 6, 11, 14, 18]. Most of these malignancies were colonic cancers. Almost every surgeon with some years of experience can remember cases of colonic cancer detected by routine manual palpation of abdominal organs while performing open cholecystectomy. Based on his extensive study including 24,000 autopsies, Schmauss [17] demanded that gallstone patients scheduled for laparoscopic cholecystectomy who exceed an age of 50 years should be examined in order to be able to exclude the presence of coexisting colonic cancer. This patient group was found to have an increased tendency to develop malignant tumors located preferentially in the right hemicolon. Similar findings have been reported by Gafa ` [9] and Jergensen [10]. The role of secondary bile acids and their influence on the incidence of colon cancer are controversial [9, 10, 17]. In our hands, four abdominal cancers were not detected in the course of LC (Table 1). These lesions may have been detected during manual examination in a conventional surgical procedure. Retrospectively, cases 1 and 2 would probably not have led to the diagnosis of the cancer through a routine examination prior to LC. In both cases the symptoms were consistent with cholecystolithiasis and ultrasound examination revealed the presence of gallstones. There is the possibility that the symptoms interpreted with cholecystolithiasis were actually caused by colon carcinoma, but it seems more likely that both diseases coexisted.

Case 2
A woman 64 years of age exhibited a history of cholecystolithiasis, repeated pain, and biliary colic in the last 3 months before admission. Defecation was reported to be normal. Normal laboratory values were found. Ultrasonography showed numerous small stones in the gallbladder. She underwent uncomplicated LC and was discharged home 3 days later. After 2 months she was referred to our unit with severe abdominal pains. Immediate laparotomy revealed a ruptured cancer of appendix. A right colectomy was performed. Histology: Adenocarcinoma (T4N1M0).

Case 3
A man, 71 years old, with pain in the right upper abdominal quadrant and repeated colic with pain extending to the right shoulder was admitted to our unit. This patient reported no problems with defecation. Standard laboratory values were normal; ultrasound examination of the upper abdomen showed many small stones in the gallbladder. During LC, cholangiography revealed many small calculi in the common bile duct; these were removed by laparoscopic choledochotomy without complications. He did well in the postoperative period and was discharged on day 8 after removal of T-tube drain. Eight weeks later he was referred to our unit because of unclear anemia. Gastroscopy was found to be normal, but with colonoscopy we detected an obstructive tumor in transverse colon. Laparotomy with resection of transverse colon followed. The postoperative period was without problems and the patient was discharged. Histology of the tumor: Adenocarcinoma (T3N0M0).

1012

In cases 3 and 4, diagnosis of cancer was even more unlikely because clear symptoms with cholecystolithiasis were evident (hydrops and gallstones with jaundice). In these patients it can be assumed with great certainty that both suffered from a combination of cholecystolithiasis and colon cancer. However, these four patients presented significant intraabdominal conditions that most certainly were present when the LC was done. The resection at this time may have been curable. LC is an elegant new approach that minimizes the invasive nature of the operation needed to treat cholecystolithiasis. This is of considerable benefit to the patients, greatly reducing the length of stay in the hospital. However, with the advent of such convenient new operation techniques, less comprehensive preoperative evaluations are made. Both the comparatively low level of invasiveness and the enthusiasm inspired by that new procedure may result in failure to consider the possibility of other diseases such as abdominal carcinoma. This and the loss of access to the abdominal cavity for manual examination during laparoscopic surgery constitute a disadvantage compared to open surgery. We strongly urge careful examination of LC patients with preoperative endoscopy, echolaparoscopic examinations [4], and/or other available techniques in order to counter the risk of overlooking additional coexisting diseases in these patients. The laparoscopic procedures, especially LC, should not lose their well-accepted positions in todays surgery. References
1. Atabek U, Spence RK, Manigat Y, Barse F, Leese KH, Davies R, Vilanueva D, Camishion RC (1995) Documentation of abdominal exploration during laparoscopic cholecystectomy. Surg Endosc 9: 2224 2. Becker M (1994) Nicht erkanntes Kolonkarzinom bei endoskopischer Cholezystektomie. Minimal Invasive Chirurg 1: 3032 3. Bonatsos G, Leandros E, Dourakis N, Birbas C, Delibaltadakis G,

4. 5. 6. 7. 8. 9.

10.

11.

12.

13.

14.

15.

16. 17. 18.

19.

Golematis B (1995) Laparoscopic cholecystectomy. Intraoperative findings and postoperative complications. Surg Endosc 9: 889893 Cavina E, Goletti O, Buccianti P (1994) Echolaparoscopy: an indispensable procedure for laparoscopic surgery. Endosc Surg Allied Technol 2: 143148 Collet D, Edye M, Pe rissat J (1993) Conversions and complications of laparoscopic cholecystectomy. Surg Endosc 7: 334338 Denning DA, Lipshy KA (1995) Missed pathology following laparoscopic cholecystectomy: a cause for concern? Am Surg 61: 117120 Dubois F (1990) Coelioscopic cholecystectomy. 330 cases. Endosk Heute 1: 3032 Dubois F, Icard P, Berthelot G, Levard H (1990) Coelioscopic cholecystectomy. Preliminary report of 36 cases. Ann Surg 211: 6062 Gafa ` M, Sarli L, Sansebastiano G, Longinotti E, Carreras F, Pietra N, Perracchia A (1987) Prevention of colorectal cancer. Role of association between gallstones and colorectal cancer. Dis Colon Rectum 30: 692696 Jergensen T, Rafaelsen S (1992) Gallstones and colorectal cancer there is a relationship, but it is hardly due to cholecystectomy. Dis Colon Rectum 35: 2428 Junger W, Hutter J, Miller K, Moritz E (1995) Significance of preoperative colonoscopy in laparoscopic cholecystectomy. Endoscopy 27: 138139 Junginger T (1992) Gallenblase und Gallenwege einschlielich intraoperative Endoskopie und Folgeverletzungen nach endoskopischen Manahmen. In: A. Encke (ed) Breitner Chirurgische Operationslehre Band V Chirurgie des Abdomens 3. Urban & Schwarzenberg, Mu nchen, p 45 Ko ST, Airan MC (1991) Review of 300 consecutive laparoscopic cholecystectomies: development, evolution and results. Surg Endosc 5: 103108 Morgan AR, Jackson S, Mason MC (1995) Delayed diagnosis of malignant tumours missed at laparoscopic cholecystectomy. Br J Surg 82: 569 Pier A, Thevissen P, Ablamaier B (1991) Die Technik der laparoskopischen Cholecystektomie. Erfahrungen und Ergebnisse bei 200 Eingriffen. Chirurg 62: 323331 Saegesser M (1963) Spezielle Chirurgische Therapie, Medizinischer Verlag Hans Huber, Bern und Stuttgart 7. Auflage, p 494 Schmauss AK, Ehrhardt U (1983) CholelithiasisCholezystektomie und Kolonkarzinom. Zentralbl Chir 108: 449456 Sharp EJ, Sprigall RG, Theodorou NA (1994) Delayed diagnosis of malignant tumours missed at laparoscopic cholecystectomy. Br J Surg 81: 1650 Siewert JR, Feussner H, Scherer MA, Brune IB (1993) Fehler und Gefahren der laparoskopischen Cholecystektomie. Chirurg 64: 221 229

Surg Endosc (1997) 11: 982985

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

MR cholangiography (MRC) in the evaluation of CBD stones before laparoscopic cholecystectomy

P. Pavone,1 A. Laghi,1 D. Lomanto,2 F. Fiocca,2 V. Panebianco,1 C. Catalano,1 P. Mazzocchi,2 R. Passariello1
1 2

Department of RadiologyII Chair, University of Rome La Sapienza, Policlinico Umberto I, Viale regina Elena 324, 00161Rome, Italy Department of Surgery, University of Rome La Sapienza, Policlinico Umberto I, Viale regina Elena 324, 00161Rome, Italy

Received: 9 October 1996/Accepted: 10 February 1997

Abstract Background: The aim of our work was to evaluate the predictive value of MR cholangiography (MRC) in detecting CBD stones before laparoscopic surgical treatment. Methods: MRC was performed as a unique preoperative imaging modality in 45 selected patients (16 male; 29 female; age range: 2872; mean age: 54.4) before laparoscopic cholecystectomy. MRC imaging was obtained with a 3D Turbo Spin-Echo sequence (TR 3000 ms, TE 700 ms, Echo Train Length128) with an acquisition time of 5 min 48 s. Diagnostic confirmation was obtained in all the cases at i.o. cholangiography. When a stone was detected it was removed by transcystic or transcholedochal approach. Results: Eighteen of the 45 patients (40%) had CBD stones. MRC correctly evaluated 16 out of 18 stones, with a resulting sensitivity of 88.9%, specificity 100%, positive predictive value 100%, negative predictive value 90%, and accuracy 95.6%. Conclusions: Despite the good results of MRC, it cannot be proposed as a screening technique to be performed in all patients submitted to laparoscopic cholecystectomy due to high cost and the limited amount of MR equipment. In conclusion, only selected patients should be submitted to MRC before laparoscopic cholecystectomy. Key words: CBD stones Laparoscopic cholecystectomy MR Cholangiography

fact, the laparoscopic approach is more acceptable thanks to lower morbidity, reduced postoperative pain, lower cost, and earlier return to work. The new therapeutic approach has brought new diagnostic needs related to the evaluation of the associated common bile duct (CBD) stones [1]. In preoperative stage, different approaches have been proposed for the evaluation of choledocholithiasis: abdominal US examination, CT, i.v. cholangiography, ERCP [23]. To date there is no consensus about the use of these techniques; in fact, poor results have been obtained by US, CT, and i.v. cholangiography, whereas ERCP has to be considered an invasive procedure with related morbidity and mortality. The aim of our work was to evaluate the predictive value of MRC in detecting CBD stones prior to the surgical treatment.

Materials and methods

Between September 1995 and March 1996, MR cholangiography was performed as a unique preoperative imaging modality in 45 patients (16 male; 29 female; age range: 2872; mean age: 54.4) admitted to our hospital for surgical treatment of laparoscopic cholecystectomy. Besides known gallbladder stones patients presented with one or more of the following criteria: recurrent episodes of jaundice, pain, elevation in bilirubin, ALT, AST; alkaline phosphatase, GT, ultrasonographic finding (US) of dilated bile ducts (>6 mm). All patients were studied with a 0.5-Tesla superconductive magnet (Gyroscan T5; Philips, The Netherlands). The body coil was used for both excitation and signal reception. A T2-weighted TSE sequence (TR 3,000, TE 120, NEX 4) on the axial plane was performed to localize the biliary ducts. For MRCP imaging, a 3D Turbo Spin-Echo (TSE) sequence was acquired with the following imaging parameters: TR 3,000 ms, TE 700 ms, Echo Train Length (ETL) 128. Images were obtained on the coronal planes with a 3-mm section thickness and no interslice gap; image matrix size, 83 128; and eight signals were averaged for all images. The resulting acquisition time was 5 min 48 s. Respiratory compensation was used for all images. Coronal images were postprocessed with the MIP algorithm to produce 12 projections rotated about 15 on the coronal axis. MRCP images were contextually evaluated by two experienced radiologists by means of consensus both on MIP-reconstructed and source

Since 1987, when Mouret [15] performed the first laparoscopic cholecystectomy, there has been a remarkable change in the surgical treatment of gallstones. In most centers laparoscopic cholecystectomy has become the procedure of choice for symptomatic cholelithiasis [6, 13, 24]. In

Correspondence to: P. Pavone

983

Fig. 1. A On MIP reconstructed image a filling defect at the level of the distal tract of the CBD is observed. The gallbladder is distended. B The 3-mm stone is better appreciated on the single slice. Fig. 2. A On MIP reconstructed image the CBD appears to be 1 cm in caliber, with a stenosis at the level of the papilla. Just above the stricture a filling defect is clearly evident. Please note the low insertion of the cysticductadditional useful information for the surgeon before laparoscopic cholecystectomy. B The evaluation of the single slice (tomographic-like) better shows the 7-mm stone. Fig. 3. A In this patient with a 4-mm stone in the distal tract of a normal-caliber (<5 mm) CBD, multiple stones in the gallbladder are visualized. B The evaluation of the single slice is very important because it shows the presence of at least two stones inside the cystic duct (arrow). images. In difficult cases, additional rotations of the images on the coronal plane were available. Diagnostic confirmation was obtained in all the cases at i.o. cholangiography. When a stone was detected it was removed by transcystic or transcholedochal approach.

Results MRC images of optimal quality were acquired in all the patients, although in one case the examination was compromised by the poor clinical condition of the patient; however, the study was considered of sufficient diagnostic quality. Eighteen of the 45 patients (40%) had CBD stones (single in 14 cases and multiple in four) ranging in diameter between 3 and 13 mm (mean: 7 mm), as confirmed at i.o. cholangiography. Stones were detected as filling defects inside the bile ductsthat is, as oval or round-shaped le-

sions of lower signal intensity compared to the high signal intensity of the bile (Figs. 13). MRC correctly evaluated 16 out of 18 stones. Two falsenegative examinations were due to the learning curve of the radiologist observing MRC images. In fact, a retrospective evaluation allowed the stones to be correctly identified. No false-positive examinations occurred. The resulting sensitivity and specificity were, respectively, 88.9% and 100% with a positive predictive value (PPV) of 100%, a negative predictive value (NPV) of 90%, and an overall accuracy of 95.6%. Discussion To date 95% of patients with gallstones are treated by laparoscopic cholecystectomy due to the several advantages and the few contraindications (acute peritonitis, severe por-

984

tal hypertension, carcinoma of the gallbladder, coagulation disorders, pregnancy) of this surgical technique [16, 21]. The new therapeutic approach has brought new diagnostic needs related to the evaluation of the associated common bile duct stones, usually demonstrated in 333% of patients [1]. Several papers have been published about the problem related to choledocholithiasis, but there is no consensus regarding diagnostic or therapeutic aspects. In fact, on the therapeutic point of view, both preoperative endoscopic procedure and laparoscopic exploration of the CBD can be performed [4]. Whichever therapy is performed, a correct evaluation as to the presence of CBD stones is of course mandatory. In preoperative stage, different techniques have been proposed for the evaluation of choledocholithiasis: US, CT, i.v. cholangiography, and ERCP. Abdominal US has an overall sensitivity of 55% in detecting CBD stones, but if the diameter of the CBD is less than 6 mm, in up to 86% of patients, stones cannot be detected [17]. CT has been widely evaluated in the past as a noninvasive screening method in diagnosing choledochal stones, with reported sensitivities ranging between 83% and 90% [3, 11]. But more recent studies [2] reassessed the criteria for a CT diagnosis and found lower sensitivity (75%). Preoperative i.v. cholangiography has the main limitation of a lack of opacification of the biliary system, which occurs in 3040% of cases. In a prospective study i.v. cholangiography was useful in detecting choledocholithiasis only in 1.5% of cases [18]. Moreover, risks related to adverse reaction of contrast agent, the cost of the procedure, and the poor anatomical definition of the intra- and extrahepatic bile ducts should be considered [22]. The use of ERCP before laparoscopic cholecystectomy is controversial. Good results have been obtained by combining ERCP and endoscopic sphincterotomy in the diagnosis and management of suspected CBD stones. But patient selection is very important because 10% of patients who undergo this procedure develop some type of complication with a mortality of 0.371.0% [5]. Even if ERCP cannot be considered as a screening procedure in preoperative evaluation of patients with suspected CBD stones, it could be performed if a high clinical suspicion for the presence of stones is evident. Several authors use a series of clinical, laboratory, and radiologic indicators to indicate the presence of stones [10, 20]. A direct correlation between the number of positive criteria and the presence of stones has been established. But 410% of patients with bile duct stones are not detected due to lack of symptoms. The second approach is to perform an intraoperative study by i.o. cholangiography and ultrasonography. I.o. cholangiography performed during a laparoscopic procedure is time-consuming and not harmless, and a bile duct injury can occur [7]. Therefore i.o. cholangiography would not appear to have merit unless demonstration of stones in the duct leads to immediate exploration of the CBD, which still remains a technically difficult procedure. Laparoscopic ultrasonography has some limitations due to technical problems and can be considered as an ancillary technique to i.o. cholangiography. The rationale of our work was to evaluate the predictive

value of a new, noninvasive imaging modality, MR cholangiography, in detecting CBD stones prior to the surgical treatment. Thanks to developing new equipment and sequences, dedicated techniques for the study of the biliary system have been evaluated [9, 14]. In our work we used a 3D Turbo Spin-Echo sequence to acquire heavily T2weighted images in order to maximize nonflowing fluid and to cancel the signal intensity of the background both of the liver parenchyma and the flowing blood [19]. Images are acquired on the coronal planes and they are subsequently reconstructed by dedicated algorithm on the console, in order to obtain 2D and 3D projection images. Recent papers have stated the high reliability of MRC in detecting stones in the main bile duct. In the largest series a sensitivity of 88% and a specificity of 98% were described, with a PPV of 91% and a NPV of 97% and an overall diagnostic accuracy of 96% [8]. Our results, even if obtained on a smaller series of patients, are in agreement with the literature [12]. MRC has several advantages over the other imaging modalities. In fact, it is a noninvasive procedure, requiring neither any biliary intervention nor any contrast medium administration. Moreover, it provides a cholangiographic view of the intra- and extrahepatic ducts, similar to ERCP and PTC. For what concerns stones they are depicted as filling defects inside the CBD, with images similar to conventional techniques, but based on different physical background. Stones are better evaluated on the single slices rather than on the reconstructed image, because an MIP algorithm, similarly to overfilling of contrast medium during ERCP, can cancel small stones inside the CBD (Figs. 1 and 3). No complications or contraindications are related to this imaging technique except for the common contraindications to an MRI examination (pacemaker, intracranial vascular clips). In our series where patients were selected on the basis of positive criteria for CBD stones, in 60% of cases no stones were detected; in these particular cases the utility of MRC was to avoid useless ERCP or surgical exploration of the CBD. On the other side it is well known that almost 4 to 10% of asymptomatic patients have a CBD stone; as a consequence the conclusion should be that MRC has to be performed in all the patients to be submitted to laparoscopic cholecystectomy. This is not possible because of the high cost and limited amount of equipment with adequate software to perform MRC. Therefore, in conclusion, MRC cannot be used as a screening technique in all patients to be submitted to laparoscopic cholecystectomy; it should be confined to selected cases in which a preoperative US examination could not be of diagnostic value.

References
1. Arnold DJ (1970) 28.621 cholecystectomy in Ohio: results of a survey in Ohio hospital by the Gallbladder Survey Committee, Ohio Chapter, American College of Surgeons. Am J Surg 119: 714717 2. Baron RL (1987) Common bile duct stones. Reassessment of criteria for CT diagnosis. Radiology 162: 419 3. Baron RL, Stanley RJ, Lee JKT (1982) Prospective comparison of the evaluation of biliary obstruction using computed tomography and ultrasonography. Radiology 145: 91

985 4. Berci G (1991) Cholangiography and choledochoscopy during laparoscopic cholecystectomy, its place and value. Dig Surg 8: 9296 5. Cotton PB, Lehman G, Vennes J, Geenen JE, Russell RCG, et al. (1991) Endoscopic sphincterotomy complications and their management: an attempt at consensus. Gastrointest Endosc 37: 383393 6. Franceschi D, Brandt C, Margolin D, Szopa B, Ponsky J, Priebe P, Stellato T, Eckhauser ML (1993) The management of common bile duct stones in patients undergoing laparoscopic cholecystectomy. Am Surg 59: 525532 7. Grogono JL, Woods WGA (1986) Selective use of operative cholangiography. World J Surg 10: 10091013 8. Guibaud L, Bret PM, Reinhold C, Coll E (1994) Diagnosis of choledocolithiasis: value of MR cholangiography. AJR Am J Roentgenol 163: 847850 9. Hall-Craggs MA, Allen CM, Owens CM, et al. (1993) MR cholangiography: clinical evaluation in 40 cases. Radiology 189: 423427 10. Hauer-Jensen M, Karesen R, Nygaard K, Solheim K, Amlie E, et al. (1985) Predictive ability of choledocholithiasis indicators: a prospective evaluation. Ann Surg 202: 6468 11. Jeffrey RB, Federle MP, Laing FC, Wall S, Rego J, Moss AA (1983) Computed tomography in choledocolithiasis. AJR Am J Roentgenol 140: 1179 12. Laghi A, Pavone P, Catalano C, Broglia L, Messina A, Passariello R (1995) Choledocholithiasis: diagnostic accuracy of MR cholangiography. Radiology 197(P): 312 13. McSherry CK (1989) Cholecystectomy: the gold standard. Am J Surg 158: 174185 14. Morimoto K, Shimoi M, Shirakawa T, Coll E (1992) Biliary obstruction: evaluation with three-dimensional MR-cholangiography. Radiology 183: 578580 15. Mouret P (1991) From the first laparoscopic cholecystectomy to the frontiers of laparoscopic surgery: the future prospectives. Dig Surg 8: 124125 16. National Institute of Health Consensus Development Conference Statement on Gallstones and Laparoscopic Cholecystectomy (1983) Am J Surg 165: 390398 17. Pasanen P, Partanen K, Pikkarainen P, Alhava E, Pirinen A, Janatuinen E (1992) Ultrasonography, CT and ERCP in the diagnosis of choledocal stones. Acta Radiol 33: 5356 18. Patel JC, McInnes GC, Bagley JS, Needham G, Krukowski ZH (1993) The role of intravenous cholangiography in pre-operative assessment for laparoscopic cholecystectomy. Br J Radiol 66: 11251127 19. Pavone P, Laghi A, Catalano C, Broglia L, Messina A, Scipioni A, Di Girolamo M, Passariello R (1996) MR cholangiopancreatography (MRCP) at 0.5T: technique optimisation and preliminary results. Eur Radiol 6: 147152 20. Reis R, Deutsch AA, Nudelman I, Kott I (1984) Statistical value of various clinical parameters in predicting the presence of choledochal stones. Surg Gynecol Obstet 159: 273276 21. Schirmer BD, Edge SB, Dix J, Hyser MJ, Hanks JB, Jones RS (1991) Laparoscopic cholecystectomy: treatment of choice for symptomatic cholelithiasis. Ann Surg 6: 665677 22. Shehadi WH, Toniolo G (1980) Adverse reaction to contrast media. Diagn Radiol 137: 299302 23. Thomas MJ, Pellegrini CA, Way LW (1982) Usefulness of diagnostic tests for biliary obstruction. Am J Surg 114: 102108 24. Zucker KA, Bailey RW, Gadacz TR, Imbembo AL (1991) Laparoscopic guided cholecystectomy. Am J Surg 161: 3644

Surg Endosc (1997) 11: 10131016

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

Perioperative tumor localization for laparoscopic colorectal surgery

S. H. Kim, J. W. Milsom, J. M. Church, K. A. Ludwig, A. Garcia-Ruiz, J. Okuda, V. W. Fazio
Department of Colorectal Surgery, A 111, The Cleveland Clinic Foundation, 9500 Euclid Avenue, Cleveland, Ohio 44195, USA Received: 18 July 1996/Accepted: 10 March 1997

Abstract Background: Because of the inability to palpate colonic tumors during laparoscopy, their location must be precisely identified before resection is undertaken. Method: A retrospective study was performed of 58 patients in order to be able to describe our methods of tumor localization for laparoscopic colorectal operations and to review their effectiveness. Results: In all patients, the entire colon was examined preoperatively by colonoscopy. In one patient, preoperative colonoscopic localization was inaccurate. In 31 patients, tumors were easily detectable at surgery. In five patients with the tumor in the right colon, even though the lesion was not detectable at surgery, right colectomy was performed without marking because preoperative colonoscopy reliably identified the lesion adjacent to the ileocecal valve. Twentytwo patients required some type of procedure to localize the tumor. The procedures and their problems were as follows: preoperative tattoo (five)tattoo not visualized (one); intraoperative colonoscopy alone (six), combined with intraoperative tattoo (four) or clip (three)poor operative exposure due to bowel distension (nine), hard to see the clip (three), dislodged clip (two), inadequate resection margin (one); intraoperative proctoscopy alone (two), combined with laparoscopic stitch (two)no problems. In no patient was tumor present at a resection line and in no patient was the wrong segment resected. Conclusions: Reliable preoperative identification of the tumor adjacent to the ileocecal valve can permit right colectomy without marking. Lesions in the upper rectum can be approached via intraoperative proctoscopy suture placement. If the surgeon anticipates intraoperative localization may be difficult, lesions other than rectal or cecal ones should probably be marked by preoperative tattooing. Further studies regarding the technique of tattooing are warranted.

Key words: Tumor localization Laparoscopic surgery Colon tumors Rectal tumors

The need for tumor localization in laparoscopic compared to conventional colorectal tumor surgery may be heightened since palpation with the hand is not possible. Preoperative colonoscopy alone may be unreliable in delineating the tumor localization with enough accuracy for surgery. This has led to reports of laparoscopic resection of the wrong segment of the colon [6, 12]. Certain techniques must therefore be developed to ensure that the tumor-bearing segment of bowel is removed. The purpose of this retrospective study is to describe the methods we have used to precisely localize tumors where we anticipated intraoperative localization might be difficult at laparoscopic colorectal operations, and to review their effectiveness.

Methods
Between December 1992 and July 1996, 58 patients underwent laparoscopic surgery for colorectal tumors. Patients who had familial adenomatous polyposis or underwent stoma creation alone for palliative purposes were excluded from the study. Patients were operated on for both premalignant and malignant diseases of the colon and rectum (14 adenomas, 44 adenocarcinomas). Resection of a segment of bowel was performed in 56 patients (25 right colectomy, two left colectomy, 20 proctosigmoidectomy, seven abdominoperineal resection, and two subtotal colectomy), and a colotomy and removal of polyp in two patients, respectively. All cancer patients who underwent a curative resection were involved in an ongoing prospective randomized study in our department comparing laparoscopic to conventional colorectal cancer surgery.

Results
The abstract of this manuscript was selected for poster presentation for the Scientific Session of the SAGES Annual Meeting 1922 March 1997, San Diego, CA Correspondence to: J. W. Milsom

In all 58 patients, the entire colon was examined preoperatively by colonoscopy. In one patient (16), preoperative colonoscopic localization was inaccurate. Preoperative colonoscopy reported a 5 5 cm adenoma at the splenic

1014 Table 1. Summary of perioperative procedures to precisely localize the lesiona Patient 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22
a

Location ASC SIG CEC ASC SIG SIG SIG ASC ASC REC REC SIG SIG SIG REC TRN TRN SPF REC ASC SIG REC ASC

Pathology Adenoma Cancer Cancer Adenoma Cancer Adenoma Cancer Cancer Cancer Cancer Cancer Cancer Cancer Cancer Cancer Adenoma Adenoma Cancer Cancer Cancer Cancer Cancer Cancer

Perioperative procedure Intraoperative colonoscopy/clip Intraoperative colonoscopy/clip Intraoperative colonoscopy Intraoperative colonoscopy/clip Intraoperative tattoo Intraoperative tattoo Intraoperative colonoscopy Preoperative tattoo Intraoperative proctoscopy Intraoperative proctoscopy Intraoperative colonoscopy Intraoperative colonoscopy Intraoperative tattoo Intraoperative tattoo Intraoperative colonoscopy Intraoperative colonoscopy Preoperative tattoo Intraoperative proctoscopy/stitch Preoperative tattoo Preoperative tattoo Intraoperative proctoscopy/stitch Preoperative tattoo

Operation C&P PS RC PS PS PS RC RC PS PS PS PS PS PS C&P RC STC PS RC PS PS RC

Comments

Multiple adenomas in ASC

3-cm sessile polyp Erroneous localization by preoperative colonoscopy Conversion due to nonvisualization of tattoos

CEC: cecum, ASC: ascending colon, TRN: transverse colon, SPF: splenic flexure, SIG: sigmoid colon, REC: rectum, RC: right colectomy, PS: proctosigmoidectomy, STC: subtotal colectomy, C & P: colotomy and polypectomy

flexure but intraoperative colonoscopy revealed the tumor to be in the proximal transverse colon. Laparoscopic right colectomy was performed but inspection of the specimen demonstrated a close distal resection margin (1.5 cm). Because the possibility of cancer could not be excluded, an additional 3 cm of the distal segment was resected through a widened port site. Pathologic examination revealed a villous adenoma. In 31 patients, tumors were easily detectable at surgery. In five patients with the tumor in the cecum or ascending colon, even though the lesion was not detectable at surgery, right colectomy was performed without any marking because preoperative colonoscopy reliably identified the lesion adjacent to the ileocecal valve. Twenty-two patients required some type of procedure (other than preoperative colonoscopy alone) to precisely localize the tumor before or during laparoscopy. In five patients with colon cancer, India ink was injected preoperatively through a colonoscope using a sclerotherapy needle (Flextip Sclerotherapy Needle, Bard Interventional Products, C.R. Bard, Inc., Tewksbury, MA, U.S.A.). The injection was performed more than 24 h preoperatively to avoid distension of the bowel and 2 cm distal to the lesion to avoid direct injection into the tumor. One patient (17) who had a cancer at the splenic flexure was converted to an open procedure due to a problem associated with tumor localization. Even though India ink was injected in four quadrants 3 days before surgery, the tattoos were not seen at laparoscopic examination. Because the small bowel was slightly distended and the mesentery was too fatty, intraoperative colonoscopy was not performed and the patient was converted to an open procedure and underwent subtotal colectomy with ileorectal anastomosis. At surgery, it was revealed that two tattoos were in the retroperitoneal portion of

the colonic wall, one was within the thick mesentery side of the colonic wall, and one was covered by the omentum. In six patients, the tumor was localized by intraoperative colonoscopy alone. In one case (patient 16), an inadequate resection margin was obtained and an additional resection was needed. In four cases with rectosigmoid cancers, methylene blue was injected through a colonoscope during surgery. In two cases with sigmoid cancer and one case with asending colonic adenoma, the tumor was marked by clips placed laparoscopically during colonoscopy. Clips were hard to see in all three cases and dislodged from the serosa in two cases. Bowel distension by air insufflated during endoscopic examination interfered with operative exposure in nine of the total 13 patients with intraoperative colonoscopy. In four patients with rectal cancer, intraoperative proctoscopy was performed. In two of them, the distal margin of the tumor was marked by a laparoscopically placed stitch after tumor localization using intraoperative proctoscopy. The procedures and their problems are summarized in Table 1 and Table 2, respectively. In no patient was tumor present at a resection line and in no patient was the wrong segment of bowel resected.

Discussion Laparoscopic resection of the wrong segment of the colon, requiring conversion to a standard laparotomy and an additional resection, was first described by Larach et al. [6]. In a recent survey of the members of the American Society of Colon and Rectal Surgeons [12], 18 of 278 responders (6.5%) reported the removal of the wrong segment of the

1015 Table 2. Problems of various procedures for tumor localization Procedure Preoperative tattoo (5) Intraoperative colonoscopy (13) Colonoscopy alone (6) Combined with intraoperative tattoo (4) Combined with laparoscopic clipping (3) Intraoperative proctoscopy (4) Proctoscopy alone (2) Combined with laparoscopic stitching (2) Problem Tattoo not visualized (1) Poor operative exposure due to bowel distention (9) Inadequate resection margin (1) Hard to see the clip (3), dislodged clip (2) No problems

colon during laparoscopic colorectal surgery. Monson et al. [9] reported on a patient who was converted to an open subtotal colectomy with ileorectal anastomosis because a laparoscopically resected right hemicolectomy specimen revealed previously unsuspected multiple adenomatous polyps. Vara-Thorbeck et al. [10] reported on a patient who had a relapse 9 months after laparoscopic sigmoidectomy. The second operation by standard operation demonstrated that the relapse was localized 10 cm from the laparoscopic anastomosis, which was healthy. The author presumed this was a case of a synchronous primary tumor which was present in the first operation and suggested simultaneous colonoscopy in determining the localization of the lesion and extension of the resection is important during laparoscopic surgery. McDermott et al. [8] reported a case of nearly obstructing sigmoid colon cancer resected using the laparoscopic technique, in which postoperative bowel obstruction occurred 2 weeks after the initial operation due to an unrecognized synchronous cecal cancer. In our series of 58 patients, we have not experienced any case of either wrong segment removal or missed synchronous cancers. However, in one case (1.8%), a minor problem with tumor localization occurred due to incorrect preoperative localization. The referring physician, an experienced endoscopist who performed preoperative colonoscopy, diagnosed the patient with a 5 5 cm adenoma at the splenic flexure, but intraoperative colonoscopy showed the tumor to be located in the proximal transverse colon. This shows that the judgment of even an experienced colonoscopist may sometimes be in error. A series by Vignati et al. [11] reported that endoscopic localization was correct in 86% of 320 colonoscopic examinations. Our experience serves as a warning not to resect bowel based only on preoperative colonoscopy. (This applies to conventional as well as laparoscopic surgery.) If there is any doubt about the precise location of the tumor at surgery, intraoperative colonoscopy is mandatory before resection. Precise tumor localization is not a prerequisite just for laparoscopic surgery. Even in conventional surgery, intraoperative localization of small nonpalpable tumors or polypectomy sites has been reported to be a difficult problem. Frager et al. [3] reported six patients in whom errors of diagnosis and therapy occurred in conventional surgery because of reliance on preoperative colonoscopic tumor localization. Three patients required a second laparotomy for surgical resection of a tumor that was missed at the first exploration. The authors insisted on a preoperative barium enema for precise localization of tumors, and preoperative endoscopic tattoo or intraoperative colonoscopy for cases in which a polyp had already been removed and a segmental

resection was planned because of invasive malignancy found in the resected polyp. Espiner et al. [2] also discussed the hazards of relying only on barium enema in reporting their experience of 12 patients with radiologically proven lesions of the colon. Colonoscopy to the cecum was performed on unopened bowel during laparotomy in all cases. Additional polyps were found in five patients, and in four of these patients the polyp was not really palpable at operation. Several authors [1, 4, 5] advocate the use of colonoscopic tattoo injection for precise tumor localization. Botoman et al. [1] tangentially injected 0.51.0 ml of 1:1 diluted India ink. The tattoos were easily seen at laparotomy in 11 of 14 patients. Most of the cases in which no ink was seen occurred early in their series when smaller volumes of ink and more superficial injections were used. The author advocated staining with India ink as permanent and clearly visible even after preoperative radiation. In an experimental study in dogs, Hammond et al. [5] examined the staining characteristics of several tattooing agents. Water-soluble dyes such as methylene blue, toluidine blue, and lymphazurine stained the serosal surface of the bowel quite well; however, by 24 h the dyes had completely diffused away. They observed that India ink and indocyanine green remained visible on the serosal surface of the bowel for up to 7 days. They also showed, in a clinical study of 12 patients [4], that endoscopically injected dye (1% indocyanine green) was easily visualized on the serosal surface of the colon at surgery in all patients. In our series, preoperative colonoscopic tattooing was used in five patients with colon cancer. India ink was injected more than 24 h preoperatively to avoid the problem of distended bowel at surgery. In one case, the patient (17) was converted to a standard laparotomy partly because the tattoos were not visualized at laparoscopic examination. The operative finding showed the dyes were injected into the retroperitoneal portion of the colonic wall and into the thick mesenteric side of the colonic wall. Additionally, the serosal surface of the bowel successfully injected by the dye was covered by the omentum. Further studies with regard to the technique of injection and the amount of dye are warranted, especially for obese patients. Preoperative colonoscopic mucosal clipping using metallic clips and intraoperative fluoroscopic visualization may be an alternative for these patients [7]. Intraoperative colonoscopy was performed in 13 patients. In nine cases, operative exposure was compromised by bowel distension with air insufflated during colonoscopy. In three patients, laparoscopic clips were applied to the serosal surface of the bowel under the guidance of in-

1016

traoperative colonoscopy. Even though all lesions were successfully resected with this method, clip marking was less attractive than tattooing because clips tended to dislodge from the serosa and were harder to see. In eight patients, the tumor was localized by intraoperative colono- or proctoscopy alone, without dyes or clips. In seven cases, the resection was successful with good margins. In one case (patient 16), an additional resection was needed due to what we believed was an inadequate distal margin (1.5 cm). Recently, intraoperative proctoscopy with a laparoscopic stitch applied to the serosa just distal to the lesion of upper rectal cancer was used. This technique is attractive to localize the tumor and to get a safe distal margin for a rectal lesion in that it may be faster and more precise than endoscopic tattooing. Conclusion Reliable preoperative identification by an endoscopist of a tumor adjacent to the ileocecal valve can permit a laparoscopic right colectomy without marking. Lesions that are in the upper rectum can be approached via intraoperative proctoscopy with or without suture placement distal to the lesion. If the surgeon anticipates intraoperative localization may be difficult, lesions outside of these areas of the colon and rectum should probably be marked by preoperative tattooing. Further studies with regard to the technique of tattooing are warranted.

References
1. Botoman VA, Pietro M, Thirlby RC (1994) Localization of colonic lesions with endoscopic tattoo. Dis Colon Rectum 37: 775776 2. Espiner HJ, Salmon PR, Teague RH, Read AE (1973) Operative colonoscopy. Br Med J 1: 453454 3. Frager DH, Frager JD, Wolf EL, Beneventano TC (1987) Problems in the colonoscopic localization of tumors: continued value of the barium enema. Gastrointest Radiol 12: 343346 4. Hammond DC, Lane FR, Mackeigan JM, Passinault WJ (1993) Endoscopic tattooing of the colon: clinical experience. Am Surg 59: 205210 5. Hammond DC, Lane FR, Welk RA, Madura MJ, Borreson DK, Passinault WJ (1989) Endoscopic tattooing of the colon: an experimental study. Am Surg 55: 457461 6. Larach SW, Salomon MC, Williamson PR, Goldstein E (1993) Laparoscopic assisted colectomy: experience during the learning curve. Coloproctology 1: 3841 7. Lehman GA, Maveety PR, OConnor KW (1985) Mucosal clippingutility and safety testing in the colon. Gastrointest Endosc 31: 273276 8. McDermott JP, Devereaux DA, Caushaj PF (1994) Pitfall of laparoscopic colectomy: an unrecognized synchronous cancer. Dis Colon Rectum 37: 602603 9. Monson JRT, Darzi A, Carey PD, Guillou PJ (1990) Prospective evaluation of laparoscopic-assisted colectomy in an unresected group of patients. Lancet 340: 831833 10. Vara-Thorbeck C, Garcia-Caballero M, Salvi M, Gutstein D, Toscano R, Gomez A, Vara-Thorbeck R (1994) Indications and advantages of laparoscopy-assisted colon resection for carcinoma in elderly patients. Surg Laparosc Endosc 4: 110118 11. Vignati P, Welch JP, Cohen JL (1994) Endoscopic localization of colon cancers. Surg Endosc 8: 10851087 12. Wexner SD, Cohen SM, Ulrich A, Reissman P (1995) Laparoscopic colorectal surgeryare we being honest with our patients? Dis Colon Rectum 38: 723727

Surg Endosc (1997) 11: 10011005

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

Telesurgical mentoring
Initial clinical experience
P. G. Schulam,1 S. G. Docimo,1 W. Saleh,2 C. Breitenbach,2 R. G. Moore,1 L. Kavoussi1
The Applied Physics Laboratory and The Brady Urological Institute, Johns Hopkins University, 4940 Eastern Avenue, Baltimore, MD 21224, U.S.A. Received: 12 August 1996/Accepted: 4 March 1997

Abstract Background: Minimally invasive surgical techniques yield significant individual, economic, and social benefits when performed by experienced surgeons. Unfortunately, many of these techniques, such as laparoscopy, are associated with steep learning curves, and the incidence of complications has clearly been shown to be inversely related to experience. The initial high complication rate and the dearth of experienced endoscopic surgeons have raised concerns over training, granting of hospital privileges, and most importantly patient safety. The goal of this study was to employ current telecommunications technology in a system for the mentoring of relatively inexperienced surgeons. Therefore, we created a telesurgical system that would allow an endoscopic specialist at a central site to offer guidance and assistance to a surgeon during a laparoscopic procedure. Methods: We developed a system that connected a central site and an operative site, a distance of approximately 3.5 miles, via a single T1 (1.54 Mbs) point-to-point communications link. The system provided real-time video display from either the laparoscope or an externally mounted camera located in the operating room, full duplex audio, telestration over live video, control of a robotic arm that manipulated the laparoscope, and access to electrocautery for tissue cutting or hemostasis. Results: Seven patients underwent laparoscopic procedures using the telesurgical consultation system over the communications link. In all cases, the primary surgeon had limited experience with the laparoscopic approach but still had the basic skills required to obtain intraperitoneal access. All seven cases were completed successfully without complications. Conclusion: These initial studies have demonstrated the feasibility, effectiveness, and safety of telementoring. Telesurgical applications have the potential to greatly improve surgical education, credentialing, and patient care by offering patients and their surgeons global access to surgical specialists.

Key words: TelesurgeryLaparoscopyTelemedicine

Correspondence to: L. Kavoussi

Laparoscopy is the most familiar of the therapeutic approaches collectively referred to as minimally invasive surgery. Laparoscopic approaches have many advantages over conventional open surgery, including less postoperative discomfort, shorter hospitalization, reduced convalescence, and an improved cosmetic result [18, 21, 24, 25]. Patient demand has resulted in its rapid application in both the academic and private sectors. The application of laparoscopy to cholecystectomy was in such demand that prospective randomized trials comparing open and laparoscopic cholecystectomy were not initially performed [12, 26]. As the popularity of laparoscopic cholecystectomy grew, it became apparent that major complications occurred more commonly during a surgeons initial experience with this new technique. At some low-threshold number of cases, the risk of major complications neared that for open cholecystectomy [1, 8]. The ability of surgeons to perform laparoscopic cholecystectomy improves significantly as they gain experience with the procedure. This phenomenon is known as the learning curve. Unfortunately, a large number of patients were initially exposed to the learning curve due to deficiencies in postgraduate hands-on surgical training. Thus, the surgical community must find a way to avoid the complications associated with relative inexperience and provide the safest environment for the patient. Telemedicine offers an economical, effective means of creating a uniform standard of care for laparoscopy and other new surgical approaches. Telemedicine is a rapidly developing field that takes advantage of the information highway to provide physicians and patients with global access to health care [19]. A telesurgical system would allow an endoscopic specialist at a central site to offer guidance and assistance to a surgeon during a laparoscopic procedure. We initially tested the premise for telesurgical mentoring by establishing a central

1002 Glossary T1: DSU\CSU: A high-bandwidth communications link that allows transmission of data at a rate of 1.544 Mbps. A device comprised of a digital service unit (DSU) and a channel service unit (CSU). The DSU converts terminal signals to a format usable over digital networks. A CSU actually interfaces to the T1 line and provides the appropriate termination. COder-DECoder converts input analog signals (video and voice) to a compressed digital format, and\or the compressed digital format to analog format. An international video compression standard that specifies compression formats and methodologies.

CODEC: H.261:

Fig. 1. An experienced laparoscopic mentor surgeon at a central site can provide guidance and assistance for the physicians actually performing the procedure at a remote site.

site (Fig. 1) within the same hospital as the operating room (1000 away) [4, 13]. All of the central components were directly wired to their sources in the operating room. This preliminary system provided real-time video display from either the laparoscope or an externally mounted camera located in the operating room. The central surgical consultant communicated with the operating surgeon via duplex audio and telestration. In addition, the central surgeon had control of a robotic arm that manipulated the laparoscope and a remote switch that activated the electrocautery for tissue cutting and hemostasis. Using this pilot system, 32 procedures were performed in a controlled environment. During these studies, there was no increase in the incidence of complications; nor did we find any statistically significant difference in the operating time between telementored and conventionally mentored surgical procedures [13]. This preliminary work showed that telesurgical consultation was effective and safe. Herein we will describe the development of a telesurgical system that offers the same functions and is packaged in a desktop personal computer requiring only a single highbandwidth telecommunications link. We believe that this system can be used to guide surgeons through the final stages of the learning curve without requiring any travel by the mentor or the trainee. This report documents our initial experience with telesurgical consultation performed over a public telecommunications line between two institutions separated by a distance of 3.5 miles. Methods System design
The central site was located at the Johns Hopkins Hospital (JHH) in the James Buchanan Brady Urological Institute. The remote site was 3.5 miles across town in the operating room at the Bayview Campus of the Johns Hopkins Medical Institutions (JHBMC).

COder/DECoder (CODEC) board and communications board comprised the core of the telesurgical workstation. The Windows 3.x compatible software (ICE, Inc., Reston, VA, U.S.A.) had been custom developed for this particular application. The inputs to the workstation included a balanced microphone, video input from the endoscope (Stryker Endoscopy, San Jose, CA, U.S.A.), and composite video input from the external room camera (Canon, Rochester, NY, U.S.A.). The remote site received audio, camera control, telestration data, and a shared video screen from the central site. The video input was coded under the H.261 (see Glossary) video compression coding standard, which resulted in smooth motion video of 30 frames per second at a resolution of 176 144 noninterlaced pixels. The audio input to the CODEC board conformed to established audio standards and resulted in high-quality full duplex voice communications. The data output from the CODEC was fed into a V.35 communications board (Zydacron, Manchester, NH, U.S.A.), which processed and formated the information to interface with a data switch controller (SLI, Ijamsville, MD, U.S.A.). The data switch controller contained a CSU\DSU (channel service unit\digital service unit) that provided the terminus for the T1 link (Bell Atlantic, Baltimore, MD, USA). This procedure ensured rapid and reliable data transfer between the remote and central workstations.

Central site
The central site was equipped with a workstation identical to the one in the remote site. A standard teleconferencing camera (Canon) and microphone were available for routine communications. A pen and pad assembly provided a user-friendly interface for menu selection and telestration. The custom software allowed the central physician to control the pan, tilt, zoom, and focus capabilities of the external camera in the operating room. The software had a switch that toggled the video source between the external view and the endoscopic view. The pen and pad assembly allowed the physician to annotate and draw freehand, multi-color figures over the full-motion video screen. These annotations appeared on both the remote and central monitors in <1 second. A software echo cancellation button enabled the audio echo cancellation feature of the CODEC board and resulted in high-quality audio. A surgical robot (AESOP 1000TS; Computer Motion, Goleta, CA, U.S.A.) was available at the remote site for manipulating the endoscopic camera. The robot, in its normal mode of operation, responds to commands issued to it by a foot controller located in the operating room. For the telesurgical system, the central surgeon was capable of driving the robot via a hand control located at the central site. Remote control of the robot had been integrated into the system, and the data was transmitted via the T1 line. The surgeon at the remote site was able to override the remote control of the robot with a foot control. The central surgeon also had the ability to activate the electrocautery device at the remote site. The electrocautery had both cutting and coagulation functions. A command generated at the central site and transmitted via the T1 link caused a normally open relay to close in a switch box (Dataprobe, Paramus, NJ, U.S.A.) at the remote site. These relays were connected in parallel to the electrocautery switch and activated the unit upon closure. For safety purposes, the unit was designed to be overridden at the remote site with a foot pedal activated by the operating surgeon.

Remote site
The operating room at the JHBMC was installed with a Trunk-1 (T1) line operating at 1.544 Mbps. A 120-MHz Pentium computer with a video

Clinical trial
Seven patients underwent laparoscopic procedures using the teleoperative system. In all cases, the primary surgeon was less experienced in the

1003 Table 1. Telementored laparoscopic urological procedures Patient no. 1 2 3 4 5 6 7 Case Orchiopexy Varix ligation vasectomy PLND Renal biopsy Radical nephrectomy PLND Radical nephrectomy Remote surgeon Ped Urol Ped Urol Endo Endo Endo Endo Endo OR time (h) 3 2.1 4.25 2.75 5.5 2.83 7.7 EBL (ml) 0 0 50 200 150 150 300 Complications None None Ileus None None None None Hospital stay (days) 0 0 7 5a 3 1 5

PLND, Pelvic Lymph node dissection; Ped Utol, Pediatric Urologist; Endo, Endourologist The patients hospital stay was on the medical service for management of renal failure. Most patients undergoing laparoscopic renal biopsy are discharged home the day of the procedure.

laparoscopic approach but had basic laparoscopic skills to obtain intraperitoneal access. The experienced surgeon at the central site had previously performed 15 of each telementored procedure. An experienced laparoscopic surgeon was available in the operating room at all times in case of surgical difficulty.

Results All operative procedures were successfully completed with the aid of teleoperative consultation (Table 1). There were no intraoperative complications. A single patient had a postoperative ileus, which prolonged his hospital stay. The primary surgeon varied depending on the complexity of the procedure. All cases were telementored by one experienced endourologist (L.K.) except for the pediatric cases, which were telementored by a pediatric urologist (S.G.D.). The subjective impression of the central surgeons was that the system was adequate for telesurgical consultation. There was a nominal delay (<200 ms) within the system due primarily to the compression-decompression process and a less appreciable delay due to propagation of the data over the T1 communication link. The video received at the central site was an adequate rendition of the input analog video stream. The current system design yielded 176 144 noninterlaced pixels of resolution. Generally, the endoscopic field of view is roughly 4 4 cm2. Based on 176 144 pixels of resolution, the sytem was capable of resolving 0.2 0.3 mm2. This order of resolution was acceptable to the central surgeon in all cases. The initial testing of the audio capability was performed using a speaker phone connected to the CODECs phone line interface (RJ-11). During the initial trials, we found this level of audio communication inadequate for telesurgery because simultaneous input from both sites resulted in fragmented audio output. The initial audio link also provided no audio feedback for other occurrences in the operating room. The goal was to submerge the central surgeon into the operating room environment. In order to accomplish that goal, it was necessary for the surgeons to communicate simultaneously, as well as for the central surgeon to receive some of the background noise from the operating room. To meet these goals, we employed the balanced microphone input and amplified speaker output of the CODEC board and utilized an omnidirectional microphone, which allowed

transmission of background operating room noise. In addition, we employed echo cancellation, which provided60 dB of isolation to the audio channel. The current audio system provides full duplex mode, high-fidelity audio. Telestration on the full motion video screen was an integral component of the system and significantly enhanced communication between the central and operating surgeons. Initially, the telestration data was not being reliably received at the remote site, leading to incomplete figures from the central surgeon. Overflow of the receiving buffer resulted in incomplete data transmission. The solution incorporated both an error check for corrupted data and an increase in the size of the write buffers. In addition, color options for telestration were increased to enhance visibility of the figures. The modified telestration option was fully functional and reliable. The initial control of the robot and electrocautery was accomplished by transmitting the data for these units over the Public Switched Telephone Network (PSTN) using analog modems. In order to integrate all system components over a single communications link, the data switch controller was used to channelize the T1 link, enabling the robot and electrocautery data to be sent over the same communications link.

Discussion Although laparoscopic techniques have been adopted by the gynecologic community for intrapelvic surgery and by the general surgical community for cholecystectomy and appendectomy, more complicated extirpative and reconstructive operations have been slow to receive physician acceptance [7]. Given the advantages for the patient that have already been proven, it is desirable to broaden the spectrum of minimally invasive techniques. As compared to open surgery, laparoscopic techniques result in less pain, shorter hospitalization, faster return to the work force [18, 21, 24, 25], a lower incidence of intraabdominal adhesions [15, 16], and even some improvement in surgical results [3, 5]. These advantages are indisputable when comparing large series of laparoscopic to open cholecystectomies, and they are equally apparent with smaller series of urological procedures, including pelvic lymph node dissection, nephrec-

1004

tomy, and bladder neck suspension [13, 17]. Nor can we discount the psychological benefit of the vastly improved cosmetic result seen with laparoscopic rather than open abdominal procedures [6]. Some surgeons express a reluctance to perform complex laparoscopic procedures due to the following factors: prolonged operating times required for initial attempts, unfamiliarity with laparoscopic anatomy or advanced techniques, and a steep learning curve associated with higher complication rates for less-experienced surgeons. In 1991, the Southern Surgeons Club reported a 2.2% incidence of bile duct injury during laparoscopic cholecystectomy for the first 13 patients operated on by each surgical group, as compared to 0.1% for subsequent patients [2]. They further analyzed 8839 laparoscopic cholecystectomies performed by 55 surgeons and found that 90% of bile duct injuries occurred during the first 30 cases performed by an individual surgeon [14]. For pediatric laparoscopy performed by urologists, the reported complication rate is 8.3% for those with <20 cases and 2.8% for those with >100 cases (p < 0.0001) [20]. The majority of practicing surgeons have received little or no exposure to advanced laparoscopy during their formal training. Therefore, most surgeons learn laparoscopic techniques either through brief training courses or by participating with colleagues in laparoscopic procedures. Some reports have suggested that training courses alone are not adequate. When See et al. surveyed participants in a laparoscopic training course 3 and 12 months after completion, they found that at 3 months surgeons who performed laparoscopy without further training were 3.39 times more likely to have at least one complication than those who sought additional training [22]. Hunter et al. attempted to quantify the learning curve for surgeons taking a 3-day hands-on course on laparoscopic cholecystectomy. They reported that by the end of the 3rd day on average there was significant improvement in only one of the eight steps for the procedure, and despite their improvement with that step, it was generally considered not to have been mastered [9]. These reports suggest that there is a very different standard of learning for laparoscopy than traditionally has been accepted for standard open surgery. When surgical training is complete, the surgeon is expected to be fully competent to perform the usual operations of the specialty. But it appears that the novice laparoscopic surgeon only develops competence while actually practicing the art on his or her initial patients. Thus, some patients may be exposed to undue surgical risk during the early laparoscopic experience of their surgeons, while others may be deprived of the benefits of minimally invasive surgery due to their surgeons lack of training in complicated laparoscopic procedures. In 1990 [23], the Society of American Gastrointestinal Endoscopic Surgeons (SAGES) issued its guidelines for credentialing laparoscopic surgeons. Suggested requirements included completion of a surgical residency or fellowship incorporating laparoscopic training. For surgeons without that experience, the society suggested completion of a didactic course combining clinical experience and laboratory practice, followed by observation by a surgeon experienced in these procedures. As a final safety measure, the society recommended that applicants be observed by a proctor prior to granting them hospital privileges. There are no

data to show adherence to and success of the guidelines. However, the New York State Department of Health requires that inexperienced surgeons perform 1015 laparoscopic cholecystectomies under supervision [17]. The telerobotics research group at our institution was asked to combine engineering research and teleconferencing standards with the medical knowledge of laparoscopic surgeons in order to create a telesurgical training system. The results obtained over 2 years using two different systems clearly demonstrate the technical feasibility and clinical safety of telesurgical training. The system we describe may help to enhance current methods of laparoscopic training and realize the potential advantages of minimally invasive surgical techniques more fully. This system allows the surgical mentor to see both the activity in the operating room and the laparoscopic video image in real time. The resolution of the laparoscopic video image at the central site was 176 144 pixels, which yields resolvable features on the order of 0.2 0.3 mm2. This compares well with the NTSC signal viewed in the operating room. The number of active lines per frame of NTSC video is 488 lines per frame. Since NTSC is presented in an interlaced fashion, the perceived resolution is reduced by a factor of 0.3, which yields about 342 lines per frame [11]. This is approximately a factor of two higher vertical resolution than the telesurgical workstation. In the horizontal direction, NTSC has a perceived resolution of 452 lines, or a factor of three higher resolution than the telesurgical workstation. Based on these numbers, the overall subjective quality of the video stream is comparable to VCR-tape quality. Although the resolution of the CODECs output was less than standard video, it was considered adequate for teleconsulting. Continual two-way audio communication combined with the ability to draw on the video screen allows the mentor to advise the operating surgeon accurately on points of technique and laparoscopic anatomy. The mentor controls the visual field of the operating surgeon through the use of the robotic arm. Just as a teaching surgeon in the room would dictate the order of a surgical dissection, the telementor can direct the dissention by controlling the visual field. The mentor also controls the electrocautery, which we found necessary to keep the mentor actively involved in the surgery. If the role of the central surgeon is to be less a mentor and more a consultant or advisor, the electrocautery and/or the robotic arm could be directly controlled by the surgeon in the room. By only using the video image and one-way audio communication, the current system can easily be converted to a proctoring system whereby a central surgeon acts as an observer for credentialing purposes. The studies done to date prove that this type of surgical teaching is technically feasible and safe for the surgical patient. It is important to note that all of these operations were performed by trainees or associates of the physicians acting as telementor. It remains to be seen whether these techniques will be as effective when there is a less well-established relationship between the telementor and the operating surgeon. See et al. determined that the long-term risk factors for increased laparoscopic complication rates include attending training alone, solo practice, and a variable operative assistant [22]. They concluded that an association with other

1005

surgeons skilled in laparoscopy lessened the likelihood of risk due to discourse, feedback, and direction provided by a skilled laparoscopist. Telemedicine may provide a cost-effective association between the limited number of laparoscopic specialists and those in their final stages of learning. The availability of an affordable telesurgery system for laparoscopy or other endoscopic techniques would have several advantages for the surgical patient. It would allow an inexpensive and efficient way for hospitals to credential surgeons for laparoscopic techniques, thus reducing concerns about complications due to lack of experience. The availability of the system would allow surgeons to seek intraoperative consultation in particular techniques from experts when difficulties arise. Any consulting surgeon who could be contacted by phone could also be available to teleconsult. The patient would benefit from advice given by a participant with more intensive experience in the procedure being performed. Implementation of the complete system would also provide a venue for remote surgical teaching, which would encourage laparoscopic surgeons competent in basic techniques to broaden their experience under the guidance of a laparoscopic expert. This system would not only benefit the individual patient undergoing a minimally invasive technique that might have been previously unavailable, it also has the potential for assuring an acceptable uniform standard of care. Conclusion Because of the paucity of experience with complicated minimally invasive procedures of most surgeons, their patients are being deprived of the potential benefits of these techniques. The dearth of standardized training programs and the steep learning curve associated with minimally invasive procedures lead to higher risks of complications for patients of inexperienced laparoscopic surgeons. An inexpensive telesurgical system is one means of addressing this problem. Teleproctoring could lead to more standardized laparoscopic credentialing, while teleconsultation would allow more expert surgeons to participate in procedures being performed by those with limited experience. With telementoring, surgeons could receive training in advanced techniques from experts anywhere in the world without leaving home. Finally, telerobotic surgical teaching would increase the general availability of minimally invasive surgery while decreasing the likelihood of complications due to inexperience with these new techniques.
Acknowledgment. We thank Lorenz J. Happel for his technical expertise, David T. Kingsbury for his support, Pat Peldo and Bell Atlantic for supplying the T1 service, and ICE communications for their contribution of equipment and support. Addendum: Recently this system was employed in two international telesurgical mentoring procedures. The central site was at Johns Hopkins Hospital and the remote sites were in Bangkok, Thailand (Laparoscopic Varix ligation) and Innsbruk, Austria (laparoscopic adrenalectomy). Both cases were completed without complications.

References
1. Brunt LM, Soper NJ (1993) Laparoscopic cholecystectomy: early results and complications. Comp Surg 25. 2. Club SS (1991) A prospective analysis of 1518 laparoscopic cholecystectomies. N Engl J Med 324: 10731078 3. Docimo S (1995) The results of surgical therapy for cryptorchidism: a literature review and analysis. J Urol 154: 11481152 4. Docimo SG, Moore RG, Adams J, Ben Chaim J, Kavoussi LR (1996) Early experience with telerobotic surgery in children. J Telemed Telecare 2: 4850 5. Docimo SG, Moore RG, Adams J, Kavoussi LR (1995) Laparoscopic orchidopexy for the high palpable undescended testicle: preliminary experience. J Urol 154: 15131515 6. Docimo SG, Moore RG, Adams J, Kovoussi LR (1995) Laparoscopic bladder augmentation using stomach. Urology 46: 565569 7. Duckett JW (1994) Pediatric laparoscopy: prudence, please [editorial]. J Urol 151: 742743 8. Hawasli A, Lloyd LR (1991) Laparoscopic cholecystectomy. The learning curve: report of 50 patients. Am Surg 57: 542545 9. Hunter JG, Sackier JM, Berci G (1994) Training in laparoscopic cholecystectomy: quantifying the learning curve. Surg Endosc 8: 2831 10. Kavoussi LR, Kerbl K, Capeluto CC, McDougal EM, Clayman RV (1993) Laparoscopic nephrectomy for renal neoplasms. Urology 42: 603609 11. Kenyon ND, Nightingale C (1992) Audiovisual telecommunications. 1st ed. BT Telecommunications series. Chapman & Hall 12. Miller-Catchpole R (1991) Laparoscopic cholecystectomy. JAMA 265: 15851587 13. Moore RG, Adams JB, Partin AW, Docimo SG, Kavoussi LR (1996) Telementoring of laparoscopic procedures: initial clinical experience. Surg Endosc 10: 107110 14. Moore MJ, Bennett CL (1995) The learning curve for laparoscopic cholecystectomy. Am Surg 170: 5559 15. Moore RG, Kavoussi LR, Bloom DA, Bogaert GA, Jordan GH, Kogan BA, Peters CA (1995) Adhesion formation after urologic laparoscopy in the pediatric population. J Urol 153: 792795 16. Moore RG, Partin AW, Adams JB, Kavoussi LR (1995) Adhesion formation after transperitoneal nephrectomy: laparoscopic vs. open approach. J Endourol 9: 277280 17. Laparoscopic Surgery. Albany: New York State Department of Health; June 12 1992. New York State Department of Health Memorandum, series 92-20. 18. Parra RO, Andrus C, Boullier J (1992) Staging laparoscopic pelvic lymph node dissection: comparison of results with open pelvic lymphadenectomy. J Urol 147: 875878 19. Perednia DA, Allen A (1995) Telemedicine technology and clinical applications. JAMA 273: 483488 20. Peters CA (1996) Complications in pediatric urological laparoscopy: results of a survey. J Urol 155: 10701073 21. Polascik TJ, Moore RG, Rosenburg MT, Kavoussi LR (1995) Comparison of laparoscopic and open retropubic urethropexy for treatment of stress urinary incontinence. Urology 45: 647652 22. See WA, Cooper CS, Fischer RJ (1993) Predictors of laparoscopic complications after formal training in laparoscopic surgery. JAMA 270: 26892692 23. Society of American Gastrointestinal Endoscopic Surgeons (1991) Granting of privileges for laparoscopic general surgery. Am J Surg 161: 324325 24. Soper WJ, Barteau JA, Clayman RV, Ashley SW, Dunnegan DL (1992) Comparison of early postoperative results for laparoscopic versus standard open cholecystectomy. Surg Gynecol Obstet 174: 114 118 25. Vogt DM, Curet, MJ, Pitcher DE, Martin DT, Zucker KA (1995) Preliminary results of a prospective randomized trial of laparoscopic versus conventional inguinal herniorrhaphy. Am J Surg 169: 8490 26. Wolfe BM, Gardiner BN, Frey CF (1991) Laparoscopic cholecystectomy: a remarkable development. JAMA 265:15731574

EndoScope: world literature reviews

Surg Endosc (1997) 11: 10521054

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

Original articles from a wide range of international surgical journals are selected by our editors and presented here as a structured summary and critical review. EndoScope serves as a quick and comprehensive survey of the expansive endoscopic literature from all the corners of the globe.

Laparoscopic aortic replacement in the porcine model: a feasibility study in preparation for laparoscopically assisted abdominal aortic aneurysm repair in humans

Right colonic arterial anatomy: implications for laparoscopic surgery

Garcia-Ruiz A, Milson JW, Ludwig KA, Marchesa P Dis Colon Rectum (1996) 39:906911 To investigate previous reports and clinical observation that the right colic artery arises infrequently from the superior mesenteric artery (SMA), detailed dissections of the SMA in 56 human cadavers were performed. Only in six cases (10.7%) out of the 56 subjects did the authors find the right colic emanating directly from SMA. Blood supply to the ascending colon was mostly from a secondary branch emanating from the ileocolic artery (66%) or the middle colic artery (23.3%). This is in contradistinction to most surgical textbook decriptions of a normal blood supply to the right colon consisting of three independent branches (ileocolic a., right colic a., and middle colic a.), all originating from the SMA. Of the 56 adult cadavers, 29 were male and 27 were female; none of the subjects were removed from the study due to previous surgery or pathology. All corpses were either bequested or considered unclaimed; 50 dissections were performed in Cleveland, OH, while six came from Mexico City, Mexicoall within a 1-year period. No radiologic or perfusion studies were performed. Combined with review of other anatomic studies of the past, the authors concluded that in the vast majority of cases there were only two independent branches arising from the SMAs that supply the ascending colon, the ileocolic, and the middle colic arteries. The right colic artery directly arising from SMA is unusual. In the case in which the origin of that vessel was not the SMA, the right colic artery was considered absent, and in its place a right colic branch was noted coming from the ileocolic or middle colic artery. The presence of the middle colic artery was recorded in 98.2% of the cadavers with the absence of right colic artery. The middle colic artery was absent in only one of the dissections. The intermesenteric arcade of Riolan between the marginal artery of the middle colic artery and the ascending branch of the left colic artery of the IMA was not present in any of the subjects, although the usual reported incidence is up to 12%. The limited exposure currently provided in laparoscopic intestinal resection demands a precise knowledge of mesenteric vascular anatomy to avoid such complications and to expedite the procedure. This is a practical low-tech study

Chen MH, Murphy EA, Levison J, Cohen JR J Am Coll Surg (1996) 183:126132 As the title states, porcine models were utilized for feasiblity assessment of laparoscopic aortic replacement; 21 subjects underwent the transabdominal approach while two subjects underwent the retroperitoneal approach for the laparoscopic dissection of the aorta. The infrarenal aorta was then cross-clamped and a custom-designed polytetrafluoroethylene intraluminal graft was inserted throught a 4-cm linear aortotomy incision. The graft was then secured in place with umbilical tapes. After restoring flow, distal runoff was assessed by witnessing blood return through a hollow-bore needle probe inserted distal to the graft. A total of 15 functioning aortic grafts were able to be placed in 21 subjects undergoing the transabdominal approach. Only two attempts were made for the retroperitoneal approach, of which one graft was successfully placed. Complications consisted of bladder (n 2), ureter (n 1), renal vein (n 1), inferior vena cava (n 1), aorta (n 1), and lumbar vessel (n 1) injuries. The ureter and the left renal vein injuries occurred during the retroperitoneal dissection only. Toward the end of the series, the operative time was reduced from 6 h to less than 2 h (cross-clamp time from 60 min to less than 15 min). The blood loss also decreased from about 1 l to less than 150 ml. The purpose of this study was to determine the feasibility of laparoscopic dissection and replacement of aorta in an animal model. As the authors correctly point out, the major limitations of this study include the use of noncalcified, nonaneurysmal aorta, external control of lumbar vessels and a customized endoluminal graft tied into place with only umbilical tapes. Indeed, the method of choice for the anastomosis between the graft and the aorta has yet to be determined under laparoscopic control. The study, however, does address the approach and setup required for a laparoscopic dissection of the aorta, which is the intent of the paper. The authors will no doubt furnish more results in the near future regarding laparoscopic aortic replacement.

1053

which will act as an invaluable guide to high-tech procedures.

Endosonography for preoperative staging of specific nodal groups associated with esophageal cancer

Chandawarkar RY, Kakegawa T, Fujita H, Yamana H, Toh Y, Fujitoh H World J Surg (1996) 20:700702 Preoperative endoscopic ultrasonography (EUS) in 74 patients with nonobstructing esophageal carcinoma was compared with postoperative histopathology evaluation of the lymph node specimen from cervical, mediastinal, and abdominal esophageal regions. Overall accuracy, specificity, and sensitivity were 87%, 90%, and 37%, respectively. EUS has an accuracy of more than 80% for detecting metastatic nodes in the cervical paraesophageal, supraclavicular, right recurrent laryngeal, left paratracheal, upper and lower paraesophageal, infraaortic, infracarinal, and lower posterior mediastinal regions. Its sensitivity is highest for cervical and upper thoracic paraesophageal, infracarinal, left paratracheal, and recurrent laryngeal nodes. Accuracy is maximum for periesophageal nodes and varies inversely with the axial distance of the nodes from the esophageal axis. Therefore, the paper concludes with the recommendation that EUS be used routinely for preoperative evaluation of cervical and mediastinal lymph nodes for metastatic involvement. Although EUS provides superior evaluation of esophageal lymph node groups, as shown by this study, the authors still advocate supplementing the examination with CT scan. Since esophophageal resection can be considered to be a palliative measure at best, perhaps EUS alone can be used in the preoperative evaluation of esophageal carcinoma. If EUS can provide esophageal evaluation comparable to that of CT scan, EUS may supplant CT as the mode of evaluation for esophageal carcinoma. The impact of EUS on early detection of esophageal carcinoma, of course, still remains speculative.

toms suggestive of GORD and seven asymptomatic volunteers underwent this prospective trial. None of the subjects had esophageal erosions or other peptic lesions as evaluated by endoscopy. Esophageal pinch biopsy specimens were taken 2 cm and 5 cm above the GE junction for histological evaluation; 24-h pH monitoring was performed on all subjects and used as the gold standard for assessment of true positive reflux patients. Infiltration with neutrophils and eosinophils and basal zone thickening could be shown only in few reflux patients and paradoxically in few asymptomatic volunteers. The sensitivity proved to be very low (46%). Nonerosive GORD produced no significant or consistent changes histologically sufficient to warrant the blind biopsy technique to fulfill the criteria of becoming a diagnostic tool. Thus, the authors contend, pinch biopsies can not be recommended for the diagnosis of nonerosive GORD. Although expensive and inconvenient, 24-h pH monitoring still remains the most sensitive and specific test for GORD diagnosis.

Factors related to the failure of endoscopic injection therapy for bleeding gastric ulcer
Brullet E, Campo R, Calvet X, Coroleu D, Rivero E, Simo-Deu J Gut (1996) 39:155158 This paper attempted to identify factors that may predict endoscopic injection failure leading to recurrent bleeding from gastric ulcer erosion. Over a 3-year period, 1,661 patients admitted for upper GI bleeding underwent emergency endoscopy within 12 h of admission. In 178 patients, a gastric ulcer with either a bleeding or a nonbleeding visible vessel was identified. Except for the three patients who underwent surgery due to endoscopic inaccessibility of the bleeding source, the remaining 175 lesions were injected with 815 ml of adrenaline (1/10,000) followed by 310 ml of polidocanol (1%). Permanent hemostasis was achieved in 134 of 175 injected cases; 41 developed further bleeding; 31 required reinjection therapy; five required emergency surgery and another five received palliation due to their advanced malignancy status. Of the reinjection group, 21 out of the 31 patients achieved permanent hemostasis while the other 10 patients underwent surgery from repeated injection therapy failure. Overall success rate of injection therapy in this study group was 87% (155 of 178). Rebleeding rate after first injection was 23.4% (41 of 175) and 32.2% (10 of 31) after the second treatment. The overall rebleeding rate was 12% (23 of 175). The overall mortality rate was 10.6% (19 of 178)7.7% (12 of 155) if definitive hemostasis by injection therapy was achieved and 30.4% (7 of 23) if injection therapy failed. The univariate analysis of clinical and endoscopic variables with logistic regression to adjust simultaneously for multiple covariates revealed certain predictive factors for injection therapy failure. Hypovolemic shock, active bleeding, ulcer location high on the lesser gastric curvature, and

Diagnostic value of histology in nonerosive gastro-esophageal reflux disease

Schindlbeck NE, Wiebecke B, Klauser AG, Voderholzer WA, Muller-Lissner SA Gut (1996) 39:151154 This study examined whether the diagnosis of nonerosive gastro-esophageal reflux disease (GORD) could be made by histological examination during endoscopy as an alternative to the long-term pH monitoring test; 24 patients with symp-

1054

ulcer size larger than 2 cm were all independent variables significantly associated with endoscopic therapy failure. The strongest predictive factor for failure was ulcer size >2 cm, which was associated with a 3.6 times higher risk for rebleeding. Interestingly enough, neither NSAID use nor multiple transfusions nor hemoglobin levels revealed statistical significance in multivariate analysis. The authors conclude that injection therapy is useful in

treating bleeding from high-risk gastric ulcer as permanent hemostasis is obtained in 87% of the cases. The presence of significant risk factor(s) for injection failure may require the addition of thermal therapy and or early operation. Indeed, for bleeding gastric ulcers excluding any variceal component, the method of choice for initial management appears to be endoscopic injection therapy as supported by this commendable study.

Letters to the editor

Surg Endosc (1997) 11: 10551056

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

Hiding the scars of an appendectomy

New method of port placement
It has been more than two decades since Semm [6] removed the first appendix via laparoscopic surgery. In recent years, the laparoscopic appendectomy has evolved from being perceived as a novelty, performed only occasionally, into a primary treatment for patients with the clinical diagnosis of appendicitis. There have been numerous retrospective and prospective studies [1, 2, 5, 7, 8] praising the advantages of the laparoscopic method, which include shorter hospital stay, decreased pain, and superior cosmetic outcome, while affording the surgeon the ability to inspect other intraabdominal and pelvic organs when the appendix is found to be normal. As surgeons have become more comfortable with the execution of laparoscopic surgery and new equipment becomes part of the routine [3, 4] technical modifications evolve to increase the ease of performance combined with an excellent clinical result. We have found a method, utilizing three strategically placed trocars, which is a technically simple approach and leaves virtually no visible scars. Discussion Appendicitis is commonly the disease of younger active individuals, and, as in all other surgical procedures, we should strive for the best cosmetic results and least disruption of their lives while providing the best clinical outcome. The advantages of laparoscopy have already greatly improved a number of abdominal procedures. In our institution, removal of the appendix laparoscopically has become the mainstay of treatment, and over the few years that it has been practiced, the exact method has been changed. Many different port locations have been attempted. The placement of ports in the positions described above allows the surgeon optimal access to the appendix, whichever anatomic position it is found in, ease of bimanual dissection without instrumentation difficulties, and scarless surgery. The umbilical port scar is easily hidden in the skin folds of the umbilicus and the suprapubic and left-sided ports are covered with the regrowth of the shaved pubic hair. References Method
Laparoscopy is begun after general endotracheal anesthesia, nasogastric, and Foley catheters are in place and the pubic area has been shaved. (See Fig. 1 for port placement.) A 1-cm intraumbilical curvilinear incision is made and carried down to the peritoneum. A Hassan trocar is introduced under direct vision and a pneumoperitoneum is established. A camera is place in the umbilical port and the abdomen is visualized. The two other ports are placed in areas which hide the surgical scars. The first is a 1012-mm trocar at the left superiolateral margin of the pubic hairline. The diameter of the second port is determined by the findings upon visualization of the peritoneal cavity. If perforation is found or the appendix is difficult to visualize, than the suprapubic port is 10 mm. If the aforementioned conditions are not found, this port is 5 mm. After these two ports are place in parallel fashion, dissection is carried out bimanually. The appendix is isolated. Using the suprapubic port, the appendix is elevated and held in position and an automatic stapling device is fired across the mesoappendix. The appendix is then divided in the same fashion using intestinal staples. The specimen is brought out of the abdomen using an endobag through the lateral trocar site. Inspection for bleeding and irrigation is carried out, the fascia is closed with 2-0 Vicryl, and skin is reapproximated with subcuticular 4-0 Vicryl suture. It has been our experience that these patients will be discharged form the hospital between 24 to 135 h after the surgery depending on the absence or presence of perforation [7]. 1. Attwood SE, et al (1992) A prospective randomized trial of laparoscopic versus open appendectomy. Surgery 112(3): 497501 2. Cox MR, et al (1993) Laparoscopic appendectomy: a prospective analysis. Aust N Z J Surg 63(11): 840847

Correspondence to: M. McKenney

Fig. 1. Port placement for scarless appendectomy. A Hassan trochar, intraumbilically placed B Port placement for dissection and endoscopic stapling 1012 mm or 11.5 mm C Port placement for dissection; 5 mm or 10 mm.

1056 3. Daniell JF, et al (1991) The use of an automatic stapling device for laparoscopic appendectomy. Obstet Gynecol 78(4): 721723 4. Olsen D (1991) Laparoscopic appendectomy using a linear stapling device. Surg Rounds 973983 5. Richards W, et al (1993) A review of the results of laparoscopic versus open appendectomy. S G O 177(5): 473480 6. Semm K (1983) Endoscopic appendectomy. Endoscopy 15: 5964 7. Sosa JL, Sleeman D, McKenney M, et al (1993) A comparison of laparoscopic and traditional appendectomy. J Laparoendosc Surg (3)2: 129131 8. Tate JJ, et al (1993) Conventional versus laparoscopic surgery for acute appendicitis. Br J Surg 80(6):761764

D. Yarish,1 M. McKenney,2 D. Sleeman,2 L. Martin,2 U. Desai1

1

Department of Surgery, Jackson Memorial Medical Center, University of Miami School of Medicine, University of Miami, Miami, FL 33101, USA 2 Department of Surgery, University of Miami School of Medicine, University of Miami, P.O. Box 016960, Miami, FL 33101, USA

Surg Endosc (1997) 11: 9951000

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

Visual integration of data and basic motor skills under laparoscopy

Influence of 2-D and 3-D video-camera systems
Y.-M. Dion,1,2,3 F. Gaillard1,2
1 2

Department of Surgery, St-Franc ois dAssise Hospital, 10 de lEspinay, Que bec City Quebec, Canada, G1L 3L5 Laval University, Que bec City, Canada 3 Quebec Biomaterials Institute, Que bec City, Canada Received: 11 November 1996/Accepted: 2 April 1997

Abstract Background: The qualities of visual perception and of motor reaction to the visual stimulus have never been studied in reference to the type of video-camera system (2-D vs 3-D) used during laparoscopy. Methods: The study was designed in two parts. The first evaluated the ability of the eye to discriminate how objects are spaced relative to one another. The second investigated the motor reaction to the visual stimulus in an environment where depth was the preponderent cue. The tests were performed in a pelvi-trainer in which were inserted different modules built either for visual observation (Part 1) or for evaluation of motor ability (Part 2). Variables studied during Part 1 were the time required to do the test and the number of errors committed during its performance. The variable evaluated during Part 2 was the time needed to terminate the test. Each of these two parts of the study were completed alternating the 2-D and 3-D systems. A total of 304 observations were recorded. Statistics used were the paired t-test, the independent group t-test, and the NewmanKeuls multiple comparisons test. Results: Results of Part 1 of the study confirm that visual perception varies significantly among individuals (n 10) (p < 0.05) and that a true 3-D video-camera system facilitates visual perception when compared to a 2-D system (p < 0.001). Results of Part 2 of the study also show significant differences among participants (n 9) (p < 0.05). The true 3-D system allowed significantly faster motor performances than the 2-D system (p < 0.001). Conclusion: Our experiment shows that the 3-D system allowed significant improvements in the execution of the evaluated parameters. Also noted were significant differences among participants in term of visual and motor skills.

Key words: Laparoscopy In vitro tests threedimensional videocamera system Monocularity Binocularity

As minimal access surgery develops, complex surgery is now daily realized through small trocar sites. However, this type of surgery is more demanding to the surgeon than open surgery. Therefore, the purpose of the present investigation is to evaluate whether or not utilization of a threedimensional videocamera system which theoretically provides a more realistic view of the abdominal content when compared to a two-dimensional system would facilitate visual and basic motor performances. To our knowledge, this information is not available. Materials and methods
This in vitro study was designed in two parts, each representing a major determinant of the ability to perform laparoscopy. The first part evaluated the aptitude of the eyes to discriminate how objects are positioned in relation to each other. The second part placed emphasis on motor performance. Tests were conducted with a 2-D videocamera system (Striker Endoscopy, San Jose, CA, USA) in alternance with a 3-D videocamera system (Carl Zeiss, Inc, Thornwood, NY, USA) (Fig. 1). This particular system uses a single-rod lens system installed in the endoscope. The endoscope is attached to a device that contains two microchip cameras. Two beam-paths, one for the left and one for the right image, are guided through the single optical system. The separation of the two beam-paths takes place only at the proximal end of the camera device. The right beam-path is presented to the right camera and the left to the left. As do human eyes, the cameras view the object at different angles. The signals of the two cameras are directed to a processor, which transmits the two pictures alternately to the same video monitor with a frequency of 100 Hz. This procedure produces a flicker-free picture for the surgeon. Presently, the surgeon has to wear active-liquid-crystal shutter-glasses. The glasses obstruct the view of the left eye when the picture of the right beam is displayed on the monitor and vice versa. During one second, 60 pictures each are presented separately to the right and the left eyes of the observer. The convergence range extends from 1 cm to infinity from the distal tip of the endoscope [25]. For

Correspondence to: Y.-M. Dion

996

Fig. 1. Setup of the study. The pelvi-trainer is seen on the O.R. table. The participant sat behind the pelvi-trainer in which various test modules were inserted. Tests were performed alternately using the 3-D videocamera system seen on the left of the table or the 2-D system on its right. Inserted through the front port of the pelvi-trainer, the laparoscope was maintained in a fixed position during both parts of the study. Fig. 2. One observation module used in Part 1 of the study shows 12 wooden sticks distant by 1 cm from each other and placed in rows of five different depths. The participant was asked to evaluate, beginning from the left, on which row a given wooden stick belongs. During this part of the study, four different modules were used. These modules could be reversed to allow eight different presentations to the eye of the observer. Fig. 3. One module used in Part 2 of the study. Using a grasping forceps, inserted through the same port located on top of the pelvi-trainer, each rubber ring had to be placed around its corresponding wooden stick beginning from the left. Four different series of wooden sticks were available which could be reversed, making a total of eight different presentations to the operator. Fig. 4. Part 1. Mean times used by the 10 participants. The Newman-Keuls multiple comparisons test confirmed that some participants performed significantly differently than others. The profession of each of the participants is also indicated: A: anesthetist; B: blue collar worker; C: chest physiotherapist; N: nurse; R1: first-year surgical resident; R2: second-year surgical resident. Fig. 5. Part 1. The number of errors committed varied significantly among the participants whose professions were mentioned in Fig. 4. The Newman-Keuls multiple comparisons test shows that participants 3, 10, and 4 performed significantly better than the others. Fig. 6. Part 2. Significant differences (p < 0.05) in motor skills were noted among participants. Their professions were the following: S: surgeon; R1,2,3 or 4: surgical residents in their respective year of training.

997 the purpose of our study we used an endoscope with an outer diameter of 10 mm and a 0 viewing direction. The endoscope had a field of view of 60. Various test modules were inserted in the pelvi-trainer. For Part 1, four different modules were used to ensure that the subject would not recognize a particular sequence (Figs. 26). These modules were reversed to allow eight different presentations to the eyes of the 10 participants. These subjects were randomly recruited among the operating room personnel. There were four physicians (one surgeon, one anesthetist, two residents), four nurses, one chest physiotherapist and one blue collar worker (Fig. 4). Each of the modules consisted of 12 wooden sticks of different heights evenly spaced (1 cm) but placed on five different rows. Each one of the subjects was given 10 s to look at a given sequence before being asked for his answer. The laparoscope was fixed and did not move during the course of this experiment. Each subject was required to do in an alternate manner eight observations with the 2-D system and eight with the 3-D system. A total of 160 observations were available for this part of the study. Variables studied were (1) the time taken to read each sequence, measured with a chronometer, and (2) the number of erroneous answers obtained for a given sequence. For Part 2, four different series of wooden sticks were available which could be reversed, making a total of eight different presentations to the operator (Fig. 3). Each of the nine subjects who did not participate to Part 1 had to grasp with one pair of straight forceps the rubber rings seen in the foreground, one at a time, and, beginning from the left, place them around the corresponding stick. The laparoscope was fixed and did not move during the course of this experiment. Each participant (one surgeon, two first-year residents, two second-year residents, three third-year residents, one fourth-year resident, Fig. 6) performed 16 tests, half of them in 2-D and half in 3-D, in an alternate manner, for a total of 144 tests. The variable measured was the time to completion of each sequence. Statistics used were the paired t-test, the independent group t-test, and the Newman-Keuls multiple comparisons test. Table 1. Comparative scores of 2-D and 3-D systems Time (sec.) Part 1 2D 3D Part 2 2D 3D 22.3 8.3 20.9 8.1 116.7 51.4 97.1 32.2 p < 0.001 Probability (p) Number of errors 3.8 2.6 2.3 2.1 Probability (p)

p 0.16

p 0.001

Results In Part 1 of the study, the mean times (Fig. 4) needed by the participants to complete each one of the eight sequences varied from 11.41 3.1 to 31.88 10.8 seconds. The Newman-Keuls multiple comparisons test confirms a significant difference in term of time taken by the individuals to realize their task. The Newman-Keuls multiple comparisons test is a useful tool with which to compare certain outcomes among individuals. The following is a detailed analysis of the test shown in Fig. 4. There is no absolute value ascribed to the ordinate nor the abscissa of the graph. Small numbers are always placed to the left of the graph and they increase as one looks toward the right of the graph. When two lines do not touch each other in the vertical plane, a significant difference of at least p < 0.05 is reached. For instance, their is a significant difference between the mean times subjects 1 and 2 took to read their sequences. Their is no significant difference between the mean times subjects 27 took to perform their task since their results lie on the same horizontal line. There is no significant difference between subjects 8 and 9 who, however, took more time (p < 0.05) performing than the first seven subjects. There is also no difference between subjects 9 and 10 but there is a significant difference (p < 0.05) between subjects 8 and 10. Subject 10 took longer than the others to perform his task while subject 1 took the least time. There were variations among individuals concerning the number of errors they committed per sequence (Fig. 5). The Newman-Keuls multiple comparisons test showed that three subjects made significantly less errors (p < 0.05) reading the sequences than the seven others who were indistinguishable statistically from each other. There was no relationship be-

tween the time taken to read the sequences and the number of errors noted except for one subject who took the longest mean time and who had the second-best score in term of number of errors. No difference was noted between 2-D and 3-D systems as far as time spent by the individuals to read the sequences (p 0.16) (Table 1). However, a significant difference was noted favoring the 3-D system when the number of errors committed by the participants was evaluated (p < 0.001) (Table 1). Part 2 of the study showed that participants took a mean of 74.4 11.7 to 192.4 51.3 s to realize their sequence. This difference in basic skills among participants is demonstrated by the Newman-Keuls multiple comparisons test (Fig. 6). Comparison between the two videocamera systems showed that the participants performed better with the 3-D system than with the 2-D system (p < 0.001) (Table 1). Discussion A number of articles claim advantages for the 3-D video system over the 2-D system [13, 7, 11, 14, 15, 18, 21, 22, 25]. Some reports demonstrate that a 3-D system provides excellent depth perception, definition, and resolution [3, 25] and may improve skills as demonstrated by reduced procedural time in a clinical setting. However, none of these articles evaluated separately variables like visual performance and motor skills as a function of field depth. Most visual capabilities, including detection, discrimination, and recognition, are enhanced with two eyes compared with one [5, 8, 10, 16]. This improvement with binocular viewing has been attributed to the statistical advantage of having two independent sources of input rather than one, which increases the probability of veridical perception [19]. Binocular enhancement has also been explained as a process of neural summation wherein information from each eye is combined at a higher stage [20]. The information includes spurious components uncorrelated with the signal (noise) which decrease with sample size, making the signal more salient with two eyes compared with one [9]. In addition to probability and neural summation, other theories of binocular enhancement have been advanced [4, 6], and it is likely that the specific mechanism in play depends on the nature of the task [12]. Studies of the kinematics under binocular and monocular vision during prehension showed that removal of binocular information, even in a viewing environment that contains a rich array of monocular cues, can interfere with the

998

performance of a skilled prehension movement [24]. Thus, even though monocular cues such as motion parallax, relative position, accommodation, and possibly familiar size are available, performance is clearly disturbed by the removal of binocular information [24]. These monocular or 2-D cues consist at laparoscopy of cues such as light and shade, relative size of objects, object interposition, texture gradient, aerial perspective, and, motion parallax (moving the endoscope) [2, 3, 25]. Still, the surgeon has to find the position of instruments by touching the object to be manipulated and so determine their position before using them [3]. In our study, these monocular cues were reduced to a minimum so that absolute perception of depth would become the most important determinant of success in the evaluation of individual performance using the 2-D system in comparison with the 3-D system. Since the camera was fixed, there was no motion parallax. The inside of the pelvi-trainer was deliberately made with dark material so only the working area would be visible. This led to only a small area available to establish perspective. Shades were maintained to a minimum. Distance is underestimated under monocular vision [23]. The kinematics of the movement of the upper limb to reach an object has been studied [24]. Under monocular vision, the latency to begin the movement and the movement duration are longer than under binocular vision. In addition, the peak velocity and acceleration of the reach under monocular vision are reduced relative to binocular vision. Finally, the time spent decelerating is longer under monocular viewing, particularly during the period of lowvelocity movement at the very end of the reach [24]. This extended deceleration appears to be a consequence of two things. First, it reflects in part the need to adjust a trajectory that was programmed on the basis of an initial underestimate of object distance. Second, using monocular information to make those adjustments is clearly not as efficient as using binocular information [23]. These findings and the experiment by Cruz et al. [10] provide support to our results, indicating an improved performance when our subjects used an apparatus which provides three-dimensional perception. Cruz et al. [10] studied the effect of difference in binocularity on perceived absolute distance. They studied four groups of 10 individuals (Group 1: Binoculars; Group 2: True monoculars, i.e., anatomical or functional loss of one eye with normal visual acuity in the remaining eye; Group 3: Induced monocular subjects i.e., binocular patients whose nondominant eye was occluded at the time of experiment; Group 4: Cross-eyed observers (both eyes are functional but only one is used in the act of fixation). They found that the onset of binocular deficiency was quite different among groups. Indeed, cross-eyed patients lost binocularity early and therefore developed adaptive sensory mechanisms such as suppression and anomalous retinal correspondence in a slow and gradual manner. Thus, this class of patients acquires monofixer status through continuous interaction with space. In contrast, true monocular observers, having lost one eye, acquire a definitive deficiency in an abrupt manner. When asked whether they had difficulty in judging distances immediately after losing one eye, these patients answered positively and stated that they slowly adapted to the new conditions with time. Even though the judgment of

distance by the monocular group did not differ significantly from that of the others, the authors concluded that monocularity initially induces underestimation which is corrected cognitively in the direction of overestimation in egocentric judgments (absolute perceived distance) between 1 and 3 meters (m). Part 1 of our experiment showed that simple visual integration of data, before any coordinate reaction of the subject is required, varies from individual to individual. It also shows that individuals taking less time to make these observations do not make more errors than those taking more time. Of importance because of its direct impact on laparoscopy is the demonstration in our study that the 3-D videocamera system allows a more realistic perception of the relationship between objects than the 2-D system: The subjects made significantly less errors while using the 3-D system (Table 1). From the above discussion, it is clear that a video system which allows utilization of binocular vision during laparoscopic surgery provides the surgeon with cues similar to those he uses in his nonlaparoscopic surgical work. In support of the facilitating effect the 3-D technology may afford during laparoscopic procedures, some authors comment that working 3-D decreases mental fatigue [3, 13]. However, 3-D systems available today do not compare exactly with the binocularity afforded by the human eye. Stereopsis depends primarily on the angle at which the same image is received by the eyes or the 3-D system. This angle is a function of the interpupillary distance (i.d.) and of the distance to the object (o.d.). Human i.d. is approximately 60 mm with a minimal o.d. of about 25 cm. The Zeiss Endolive system i.d. is 5 mm with an o.d. of approximately 46 cm. This means that the Zeiss Endolive image provided is similar to that given with human eyes at a viewing distance of 5070 cm. However, because Zeiss Endolive has such a large focus of depth, it provides some similarities to normal viewing, which uses convergence and accommodation. These results can only be attained with a single-channel optic system. A double-channel optic system cannot provide this versatility because it has a fixed i.d. at the end of the telescope. As mentioned previously, a three-dimensional video image can be produced by providing each eye with its own series of images, which differ from one another such that the brain builds this into a single 3-D image. There are two different techniques of directing the correct images to its respective eye without the other eye seeing it. These are a with glasses technique and a without glasses technique. The with glasses technique uses either active or passive polarized glasses to permit the observers to appreciate the right eye image by the right eye and the left eye image by the left eye. Some [1] prefer to use passive polarized glasses working in conjunction with an active screen placed in front of the monitor. When passive polarized glasses are used, the screen size is limited because the cost of a larger active screen is at present almost prohibitive. Also the distance and range of viewing on polarizing screen are reduced compared with the active glasses system. The image is not as bright on this type of screen. At present, most 3-D systems use the rather expensive and heavy battery-powered active glasses which allow the best performance. A flicker-

999

ing effect may be obtained with these glasses as the operator moves his head away from the monitor [13]. This is due to the fact that room lights cycle at 50/60 Hz and consequently not in sequence with the active glasses (100120 Hz). This side effect can be reduced by dimming the operating room lights. The without glasses technique uses lenticular lenses in front of the screen so that the right eye sees only the right eye image and the left eye sees only the left eye image. However, with this latter system, the 3-D image can be seen only from a single focal point, at which the observers head must be placed. This focal point is small and there is little room for movement. This method is currently impractical in the operating room setting, but in the future, it may become an excellent way of displaying a three-dimensional image. Three-dimensional systems are divided into two categories: true and false 3-D [2]. The former is generated from a stereoscopic image fed into two separate CCD cameras with the generated images cycled at 100120 Hz. These images give true depth perception as the object being viewed is seen from two different angles using the three-dimensional depth cues of parallax and stereopsis. One false 3-D system uses a single image from a standard monocular laparoscope onto which a single camera is focused and the generated image is cycled at 100120 Hz with either a time lag or a phase shift onto a video monitor. Another false 3-D system focuses two cameras onto a single image generated by a standard monocular laparoscope. These images are cycled, but because the object is viewed through a monocular laparoscope, there is no effective stereopsis. Both these false 3-D systems generate a 3-D image but there is no true depth perception, despite an apparent 3-D image on the monitor [2]. Since depth perception is important for the accurate placement of instruments, these false 3-D systems are not ideal. However, they do make motion parallax easier to use. The reconfigured laparoscope (true 3-D laparoscope) provides an image which is more closely aligned with human vision; however, the cost of this system is greater than a false 3-D system and an entire system must be purchased [21]. Presently, only stereoscopic video technology (true 3-D laparoscope) with shutter glasses provides the observer with spatial information that is often of such decisive importance for minimal access surgery [17]. Part 2 of our study confirms the theoretical advantages of true 3-D systems over 2-D systems (Table 1). The subjects performed the motor task significantly faster and thus more efficiently (p < 0.001) in the 3-D environment compared to the 2-D milieu. This part of the study, limited in 2-D cues, required accurate appreciation of depth since the targets (the wooden sticks) were placed 6 inches further than the point of pick up of the rubber rings. Moreover, the lumen of the rubber ring was slightly larger than the diameter of the wooden stick, so precise placement was essential in order for the test to be completed. However, the surgeon was an exception and exhibited good skills, performing as well in 2-D as in 3-D. This may be due to the fact that, with experience, one comes to interpret the 2-D image as if it were 3-D, a process called learned 3-D [2, 10]. However, as previously discussed, this process may be taking place at the expense of increasing mental fatigue, which could occur during a long and difficult procedure since additional cere-

bral integration of data has to be performed in order to correct deficient kinematics of movements [23, 24]. In our study, the 3-D system was preferred by a majority of subjects. However, some commented that the eyewear was bothersome and one subject felt dizziness. The 3-D system had an angle of vision of 60 compared to 110 for the 2-D system. Consequently, in order to show the same image on both 2-D and 3-D monitors, the tip of the 3-D laparoscope was further away from the target by a distance of 6.5 cm. The newest generation of Zeiss 3-D scopes now has an angle of vision of 80. We measured the light emitted from the 2-D and 3-D monitors and although more than twice the light intensity of the 2-D system was noted from the 3-D monitor, the same quantity of light reached the eyes once the eyewear was activated. The size and weight of the third-generation 3-D videocamera system are now comparable to those of a 2-D system. We agree with Satava [21] that the promise of 3-D vision is that it brings a more accurate and realistic image to the surgeon, which will allow the performance of complex surgical procedures with greater precision; it is one of the essential enabling technologies that will propel surgery forward. As manifest by the application to laparoscopic surgery, telepresence surgery, diagnostic imaging, and virtual reality, 3-D vision will be a centerpiece in the surgery of the future. From our own perspective, we feel that laparoscopic surgery already benefits from available true 3-D technology and will even more in the future as development continues in this field. In conclusion, our experiment shows that significant differences in term of visual and motor skills were noted among participants in this study. The 3-D system proved superior to the 2-D system during these tests evaluating the quality of visual perception and motor reaction to the visual stimulus.

References
1. Birkett DH (1993) 3-D imaging in gastrointestinal laparoscopy. Surg Endosc 7: 556557 2. Birkett DH (1994) Three-dimensional laparoscopy in gastrointestinal surgery. Int Surg 79: 357360 3. Birkett DH, Josephs LG, Este-McDonald J (1994) A new 3-D laparoscope in gastrointestinal surgery. Surg Endosc 8: 14481451 4. Blake R, Fox R (1981) The psychophysical enquiry into binocular summation. Percept Psychophys 161185 5. Blake R, Levinson E (1977) Spatial properties of binocular neurons in the human visual system. Exp Brain Res 221232 6. Blake R, Sloane M, Fox R (1981) Further developments in binocular summation. Percept Psychophys 266276 7. Boeckmann W, Pichler CV, Effert P, Wolff JM, Rau G, Jakse G (1995) Comparison of 2D and 3D videoendoscopy in laparoscopic operations. Minimally Invasive Ther 4(Suppl 1): 39 (abstract) 8. Cagenello R, Arditti A, Halpern DL (1993) Binocular enhancement of visual acuity. J Opt Soc Am 10: 18411848 9. Campbell FW, Green DG (1965) Monocular versus binocular visual acuity. Nature 208: 191192 10. Cruz AAV, Fukusima SS, Schor P, Da-Silva JA (1989) Effect of differences in binocularity on perceived absolute distance. Braz J Med Biol Res 22: 13551359 11. Cushieri A (1991) Minimal access surgery and the future of interventional laparoscopy. Am J Surg 162: 404407 12. Frisen L, Lindblom B (1988) Binocular summation in humans: evidence for a hierarchic model. J Physiol 402: 773782

1000 13. Geis WP (1996) Head-mounted video monitor for global visual access in mini-invasive surgery. Surg Endosc 10: 768770 14. Geiss PW, Kim C, Brennan M (1995) The benefits of a head-mounted video-display during complex minimally invasive surgical procedures. Surg Endosc 9(5): 604 (abstract) 15. Kunert W, Van Bergen P, Schurr MO, Tijerina LO, Buess GF (1995) Comparative study of endoscopic 2D- and 3D-vision systems: further results. Surg Endosc 9(5): 604 (abstract) 16. Legge GE (1979) Spatial frequency masking in human vision: binocular interactions. J Opt Soc Am 69: 838847 17. Matouschek E, Becker H (1995) The value of 3D video endoscopy for surgical procedures. Bildgebung 62(3): 174178 18. Peitgen K, Waltz MV, Eigler FW (1995) Does 3-dimensional imaging improve minimally invasive surgery? Surg Endosc 9(5): 604 (abstract) 19. Pirenne MH (1943) Binocular and uniocular thresholds in vision. Nature 153: 608699 20. Rabin J (1995) Two eyes are better than one: binocular enhancement in the contrast domain. Ophthal Physiol Opt 15(1): 4548 21. Satava RM (1993) 3-D vision technology applied to advanced minimally invasive surgery systems. Surg Endosc 7: 429431 22. Schwaitzberg SD, Pankratov MM (1995) 3-D laparoscopy: side-byside comparison of monitor-based and head mounted display 3-D visualization system in the operating room. Minimally Invasive Ther 4(Suppl 1): 39 (abstract) 23. Servos P, Goodale MA (1994) Binocular vision and the on-line control of human prehension. Exp Brain Res 98: 119127 24. Servos P, Goodale MA, Jakobson LS (1992) The role of binocular vision in prehension: a kinetic analysis. Vision Res 32(8): 15131521 25. Wenzl R, Lehner R, Vry Uwe, Pateisky N, Sevelda P, Husslein P (1994) Three-dimensional video-endoscopy: clinical use in gynaecological laparoscopy. Lancet 344: 16211622

Surg Endosc (1997) 11: 10341035

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

An improved method of flexible endoscopic creation of tracheoesophageal fistula for voice restoration
H. Chan,1 T. Mesko,1 K. Fields,2 J. Barkin2
1 2

Section of Surgical Oncology, Comprehensive Cancer Center, Mount Sinai Medical Center, 4306 Alton Road, Miami Beach, FL 33140, USA Department of Gastroenterology, Mount Sinai Medical Center, University of Miami School of Medicine, 4300 Alton Road, Miami Beach, FL 33140, USA Received: 24 October 1996/Accepted: 24 January 1997

Abstract. We present improvements of a previously reported method of tracheoesophageal puncture for voice restoration in postlaryngectomy patients. Our method utilizes a flexible endoscope to enable the tracheoesophageal puncture to be made under direct visualization using only local anesthesia and intravenous sedation. After 3 days, the created tracheoesophageal fistula tract is mature enough to allow placement of a voice prosthesis in the office. This allows the entire procedure to be performed in an outpatient setting with minimal risk. Key words: Laryngectomy Tracheoesophageal fistula Blom-Singer prosthesis Endoscope Voice restoration

cumbersome to use. Therefore, a tracheoesophageal fistula was created for voice restoration with insertion of a Blom-Singer prosthesis [3].

Technique
Conscious sedation was performed utilizing intravenous benzodiazepam and narcotic analgesia along with Cetacaine topical anesthetic to the oropharynx. The flexible endoscope was inserted into the esophagus and the esophagus was insufflated with air. The tip of the endoscope was then directed against the anterior esophageal wall to increase its rigidity. The posterior superior aspect of the tracheostoma (12 oclock position) was punctured with a 16-gauge angiocatheter after injecting 0.5% lidocaine with epinephrine. The puncture was directed into the anterior esophageal wall. Its impression could be seen endoscopically and needle passage into the esophageal lumen was under direct visualization (Fig. 1). A 0.038-inch guide wire was then passed through the catheter, captured via a snare, and pulled upward through the mouth. The angiocatheter was then removed and Cook vascular dilators of inceasing caliber (10F, 12F, 14F) were sequentially passed over the guidewire. The 14F dilator was passed with a Cook peel-away sheath. The sheath was left in place when the dilator was removed. A 14F red rubber catheter was placed over the guidewire and passed through the sheath and into the esophagus. The sheath was then peeled apart, leaving the catheter in place. The patient was re-endoscoped and the red rubber catheter was advanced forward into the stomach. The red rubber catheter was left in place for 3 days (Fig. 2), which allowed for maturation of the tract (Fig. 3) prior to prosthesis placement.

We had previously reported the technique of using a flexible endoscope for creation of a tracheoesophageal fistula under conscious sedation in a patient post laryngectomy [1]. The purpose of this case presentation is to report a modification which has improved upon this procedure.

Case report
A 51-year-old female with an 80-pack-year history of smoking developed voice changes and increasing stridor. She was diagnosed with a T3 supraglottic laryngeal cancer. A tracheostomy was performed for respiratory compromise and the patient was begun on chemoradiation therapy. Although the patient had clinical regression of the tumor, she continued to have residual laryngeal edema and was unable to be extubated from her tracheostomy tube. An exam under anesthesia with biopsies showed persistent moderately to poorly differentiated squamous cell carcinoma and she underwent a total laryngectomy. She had an uneventful recovery and is without evidence of recurrence 1 year after surgery. Unfortunately she has been unable to learn esophageal speech and finds the electrolarynx too

Discussion Patients who have undergone total laryngectomy can achieve voice rehabilitation by one of three techniques: (1) esophageal speech, (2) use of an electrolarynx, (3) tracheoesophageal fistula with a valve prosthesis [2, 4]. Unfortunately, esophageal speech is difficult to master and the voice quality obtained with an electrolarynx is suboptimal. This paper describes a modification of a previously reported technique described by Barkin et al. [1] in which the tracheoesophageal fistula is created percutaneously with the assistance of a flexible endoscope. This technique can be performed using local anesthesia and intravenous sedation. Using the endoscope, the tracheoesophageal puncture can

Correspondence to: T. Mesko

1035

Fig. 1. Depiction of the tracheoesphageal puncture performed with the aid of the fiberoptic endoscope.

be performed under direct visualization. This described modification uses the Cook peel-away introducer set including various size dilators and the peel-away introducer sheath to aid in placement of the red rubber catheter once the tracheoesophageal tract has been dilated. There is also a prepackaged tracheoesophageal puncture kit made by International Healthcare Technologies which contains a needle catheter, wire, dilator, and tube. This kit has been described for use with a rigid endoscope but may be modified for use with the flexible endoscope as well. This technique is a quick and easy method of placing a voice prosthesis in an outpatient setting without using general anesthesia. Possible complications include infection, bleeding, and inadvertent dislodgement of the prosthesis with premature closure of the fistula. We recommend that this technique of voice restoration be utilized more frequently in laryngectomy patients.
Fig. 2. Tracheoesophageal fistula with indwelling red rubber catheter.

References
1. Barkin JS, Hartford JD, Mikalov A, Flescher LM (1991) Creation of tracheo-esophageal fistula for voice restoration using the flexible fiberoptic endoscope. Gastrointest Endosc 37(4): 469470 2. McNeil BJ, Weichselbaum E, Pauker SO (1981) Speech and survival: tradeoffs between quality and quantity of life in laryngeal cancer. NEJM 305: 982987

Fig. 3. Tracheoesophageal fistula after 3 days of tract maturation.

3. Singer MI, Blom ED (1981) Selective myotomy for voice restoration after total laryngectomy. Acta Otolaryngol 107: 670673 4. Singer MI, Blom ED (1990) Medical techniques for voice restoration after total laryngectomy. CA Cancer J Clin 40: 169

Surg Endosc (1997) 11: 10211025

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

Is laparoscopic surgery applicable to complicated colonic diverticular disease?

M. E. Franklin Jr.,1 J. P. Dorman,1 M. Jacobs,2 G. Plasencia2
1 2

Texas Endosurgery Institute, 4242 East Southcross Boulevard, Suite 1, San Antonio, TX 78222, USA Baptist Hospital of Miami, 8755 SW 94 Street, Suite 200, Miami, FL 33176, USA

Received: 26 August 1996/Accepted: 14 February 1997

Abstract Background: Expanding upon our experience with laparoscopic surgery for colonic benign and malignant processes and for bowel obstruction, we have reviewed our experience with minimal access laparoscopic surgery for complicated diverticular disease. We propose an approach of surgical care incorporating diagnostic laparoscopy in those not responding to medical therapy alone. Methods: Our study includes data from two different surgical teams working in separate hospital-and-patient environments. Our theory that laparoscopy could be widely applicable to this complex disease process is borne out by experience in both locations. One hundred forty-eight patients were managed by laparoscopic or laparoscopically assisted methods with 18 patients requiring drainage only without resection. Results: Our management of 148 of 164 patients (90%) by laparoscopic approach was successful, with a very acceptable morbidity of 5% in the elective cases and decreased ileus (20% of open vs 7% laparoscopic) in acute complicated cases. Elective resections required hospitalization of 45 days, demonstrating the benefits of incorporating laparoscopy in the care of these cases, particularly when compared to standard open procedures requiring 8 days hospitalization. Conclusions: We believe complications of diverticular disease including abscess, perforation, fistula, and bleeding can potentially be managed in this way by minimal access procedures, decreasing postoperative wound problems, decreasing length of hospitalization and overall morbidity, and improving patient care. Key words: Laparoscopic colectomy Complicated diverticulitis Bleeding

While following standard protocols of care for patients with complicated diverticulitis (bleeding, obstruction, abscess or fistula formation, perforation with or without peritonitis), we have introduced the use of laparoscopy in the approach to this disease, making the minimally invasive techniques a cornerstone of our practice. We feel this can be used to decrease the morbidity of the necessary surgical procedures when selectively applied to these cases. Drainage of abscess, necessary colonic resection, placement of colostomy, and examination and lavage of the peritoneal cavity can be achieved by making use of laparoscopic or laparoscopic assisted approaches. We report here our experience with 164 patients with colonic diverticulitis to emphasize the minimally invasive aspects of our approach to this problem. Our experiences with laparoscopy for colorectal carcinoma [5, 12] and in acute bowel obstruction [4] have led us to apply diagnostic laparoscopy to complicated diverticular disease. We emphasize in this report 43 patients treated for specific complications of acute diverticulitis. We propose an approach to surgical care of complicated diverticular disease based on appropriate preoperative diagnostic studies such as CT scan, colonoscopy, contrast enema, and ultrasound, followed by diagnostic laparoscopy in those cases not responding to standard medical therapy. In this way complications of diverticular disease including abscess, perforation, fistula, and bleeding can potentially be managed by minimal access procedures resulting in decreased postoperative wound problems, early diagnosis, rapid treatment, avoidance of delay in treatment, decreased morbidity, and decreased hospital stay and subsequent costs.

Materials and methods

Our experience with the laparoscopic approach to complicated colonic diverticulitis/diverticulosis (bleeding) from November 1991 until March 1996 is the basis of this study of 164 patients from two separate centers, involving two separate operating teams. The underlying medical problems in these patients were widely varied and often played a major role in management, as would be expected in this age group. Ages varied from 22

Correspondence to: J. P. Dorman Presented at the 4th International Congress of the European Association for Endoscopic surgery (E.A.E.S.), Trondheim, Norway, 2326 June, 1996

1022

Fig. 1. Laparoscopic treatment of complicated colonic diverticulitis in 58 patients. to 98 years. Average age was 62.1 years. There were 78 females and 86 males. Of the 164 patients, 15 underwent open procedures or were converted to open. Of the 149 laparoscopic/laparoscopic assisted patients, 106 patients had elective procedures after bowel preparation (included both intracorporeal laparoscopic and laparoscopic-assisted anastomosis). The 43 patients of the acute diverticulitis (specific complication) group included eight patients with colonic diverticular perforation (Hartmanns procedure in six, omental patch closure in two), 18 patients who underwent diagnostic laparoscopy with drainage and lavage only without resection, four patients with laparoscopic colonic resection for recurrent bleeding, six patients who underwent laparoscopic resection and reanastomosis for colonic fistula (three colovaginal, two colovesical, and one colocutaneous), and, finally, seven patients who underwent laparoscopic resection and colostomy for acute diverticulitis with obstructive phlegmon (Fig. 1). present in four of the seven cases and these underwent laparoscopic anterior resection of the inflammatory colon with diverting colostomy (Table 1). Complications were few but similar to those with open technique (pulmonary atelectasis in one, prolonged ileus of greater than 7 days in one, renal failure in one, urinary tract infection in one, and one with bleeding from the stomach requiring transfusion. Hospital stay ranged from 6 to 23 days (average 10.7 days). A single anastomotic stricture was responsive to dilatation.

Acute diverticulitis with perforationeight patients/laparoscopic lavage and drainage only18 patients
Additional findings at diagnostic laparoscopy revealed the primary complication of acute perforation of the sigmoid diverticular disease in eight patients. Hartmanns resection with descending colostomy was the mode of therapy in six cases; omental patch closure of the perforation and closed suction drainage/lavage achieved a satisfactory result in two cases. The latter two cases were treated similarly to the 18 patients who underwent no resection but had extensive laparoscopic lavage and suction drainage of the area of abscess/perforation. This group entered the hospital with acute abdominal symptoms, elevated temperature, elevated WBC, and clinical evidence of diverticulitis with obvious peritoneal signs CT scan, sonogram, and/or water-soluble contrast study per rectum showed the pathology in four of the patients with laparoscopy done for clinical signs in the remainder (14 patients). Follow-up has ranged from 4 to 34 months; three patients had subsequent resection of the diseased colonic segment performed or completed laparoscopically, primary anastomosis having totally avoided the need for ostomy. The remaining patients have done well with either surgical resection pending or avoided due to poor medical condition or refusal to have a procedure. Their diverticular disease is presently being well managed medically. Essential to this approach is the postoperative complete colonoscopy to rule out neoplasm. Complications in the patients with perforation (eight patients) and the laparoscopic lavage and drainage only (18 patients) have included bleeding requiring transfusion in two patients, ileus lasting more than 7 days in one, and urinary tract infection and pulmonary infection in one single patient. Another patient undergoing lavage and drainage had pneumonia and congestive heart failure. One patient suffered a postoperative cerebrovascular accident. Patients averaged 7.5 days postoperative hospitalization. Toxicity of the infection associated with their illness resolved on average in 4.5 days with continuation of antibiotics and bowel rest. (Temperature and WBC were used as criteria.)

Elective procedures/open cases

The elective procedures were performed on those patients who had diagnosed diverticular disease as a consequence of repeated attacks of symptoms (106 patients, 57 males, 49 females, ranging in age from 35 to 71 years [average 49.5 years]). All patients in this group had preoperative barium enemas and/or colonoscopy. Preoperative mechanical and antibiotic bowel preparations and intravenous antibiotics were given. Laparoscopic resection was accomplished (11 laparoscopic with intracorporeal stapled anastomosis, 95 laparoscopic assisted stapled anastomosis). Average length of hospital stay postoperatively was 4.5 days. One serious complication was noted with anastomotic stenosis over 30 days postoperatively, dealt with by open procedure and revision of anastomosis. The remainder of this study addresses those acute cases undergoing laparoscopy.

Open cases
Of the acutely complicated diverticulitis patients (58 cases), 15 patients were initially started laparoscopically but were opened for varying reasons including excessive adhesions, excessive purulence, inability to adequately visualize and mobilize the area of diverticular perforation, inability to visualize the ureter, friable bleeding tissue, and, in one case, circular stapler misfiring. Hospital stay varied from 8 to 17 days (average 13 days), with multiple complications noted (atelectasis in four, wound infection in one [6.6%], prolonged ileus greater than 7 days in three [20%]).

Laparoscopically managed acute complicated colonic diverticulitis (43 cases) acute diverticulitis with obstructive phlegmon (seven cases)
Seven patients underwent diagnostic laparoscopy after failure of medical management with obstructive diverticulitis phlegmon found. Abscess was

Acute diverticulitis with fistula

Diverticulitis-induced fistulas were dealt with in six patientsthree colovaginal in women with prior hysterectomy, two colovesical, and one with

1023 Table 1. Complicated resultslaparoscopic surgery for diverticular disease acutely complicated cases 58/elective cases 106 Protective ileostomy or colostomy Open or converted to open (15) Laparoscopic (43) 1) Acute inflammation phlegmon (7) 2) Acute and chronic fistulas (6) 3) Acute perforation (8) 4) Bleeding (4) 5) Laparoscopic lavage and drainage onlyabscess/localized peritonitis/sealed perforation (18) Elective cases (106) 8 4 4 6 0 0 0 Primary stapled reanastomosis 1 5 5 0 4 0 106 Table 2. Complicated diverticulitis: postoperative morbidity Open (15 cases) Bleeding (transfusion) Pneumonia/atelectasis Urinary tract infection Acute renal failure/azotemia Postoperative ileus > 7 days Wound infection Ureteral injury Trocar hernia Mortality Cerebrovascular accident
a

Laparoscopic (43 cases) 2 2a 2a 3a 3 1 1 1 0 1

4 4 1 0 3 1 0 0 0 0

Same 2 patients

a colocutaneous fistula. Preoperative workup included colonic barium contrast x-ray, CT scan, and colonoscopy. Each of the six cases underwent laparoscopic low anterior resection of the sigmoid/rectosigmoid; in five cases, reanastomosis was accomplished. One patient had resection and reanastomosis and diverting colostomy which was closed 5 months postoperatively. Two of the cases with primary reanastomosis had protective ileostomy. All anastomoses were accomplished intracorporeally. Follow-up ranges from 5 to 37 months (mean 21 months) with one patient lost to follow-up. One patient had colonic resection and colostomy for colovaginal fistula some 9 months after lung transplantation. One patient with colocutaneous fistula and abscess had further peristomal abscess and acute renal failure which resolved postoperatively.

Diverticular disease with associated bleeding

Four patients had intermittent bleeding associated with diverticular disease (age range 5998). Colonoscopy, tagged red blood cell scan, CT scan, barium enema, and angiogram were among those tests used preoperatively. All four patients underwent mechanical and antibiotic bowel preparation preoperatively. Two patients had laparoscopic total colectomy with ileoproctostomy anastomosis performed intracorporeally. One patient underwent laparoscopic right hemicolectomy with reanastomosis. One 98-yearold patient with bleeding was taken to the operating room with angiogram catheter selectively placed showing the site of bleeding. Intraoperative on-table injection of methylene blue dye allowed for exact resection of a segment of the left colon containing the bleeding site. This patient had a 5-day postoperative hospitalization. Postoperative complications were confined to one patient who had pulmonary atelectasis, urinary tract infection, acute renal failure, and ileus of greater than 7 days. This patient was alive at 8 months without further bleeding after laparoscopic total colectomy with ileorectal anastomosis. Follow-up has been from 6 to 26 months without any evidence of rebleeding in this small series of four cases.

protective ileostomy, but admittedly these were the most difficult cases of our series (eight out of 15, or 53% of the open cases vs 14 of 43, 32.6% of the laparoscopically managed cases). Very effective lavage of the peritoneal cavity was accomplished in the lavage and drainage group as no subsequent abscesses resulted (Table 2). Wound infection occurred in one patient in the open procedure group (one of 15, or 6.6%) and in one patient who had laparoscopic management of diverticulitis with fistula (one of 43, or 2.3%) in whom peristomal abscess occurred. Interestingly, this patient had concomitant colocutaneous abscess drained at the time of the original laparoscopic surgery. Operating times averaged 120 min (range 90240 min) for elective procedures. Lavage and drainage procedures routinely required less than 60 min. Prolonged ileus lasting more than 7 days remains a potential problem, occurring in three of the 15 open cases (20%) and in three of the 43 (7%) laparoscopic cases. Pneumonia/atelectasis accounted for two complications in laparoscopic cases (4.6%) and four complications in the open cases (27%). Bleeding requiring transfusion was noted in four of the 15 open cases and in two of the laparoscopic cases. Additionally, one laparoscopic patient suffered a postoperative cerebrovascular accident.

Approach to management and operative technique in complicated colonic diverticulitis We routinely have used bowel rest and antibiotics in our practice in patients with suspected acute diverticulitis, reserving operative intervention to those failing these measures: This has been the standard of care in the past and has been demonstrated to avoid surgery in as many as 70% of patients. Use of water-soluble contrast enema and CT scanning along with ultrasound remain important diagnostic studies to make the diagnosis as certain as possible. Improvement in the patients condition is the usual outcome of the above approach, but with increasing nonresolving abdominal pain, diffuse abdominal tenderness, tachycardia, sepsis, rising white blood cell count, or any sign of deterioration, we feel that a more aggressive approach is necessary in 24 to 48 h after diagnosis is suspected and treatment started. Percutaneous drainage of abscess has been shown an effective way to deal with localized intra-

Compiled results Complicated colonic diverticulitis/diverticulosis Compiled results of laparoscopic/laparoscopic assisted colonic procedures are presented in Table 1. Diagnostic laparoscopy preceded subsequent laparoscopic management in 43 cases of complicated diverticulitis. Eighteen cases underwent resolution of their acute problems by laparoscopic lavage and drainage of abscess; two had omental patch closure of the acute perforation laparoscopically. No acute reoperation was required in this group of 20 cases, and it should be pointed out that no colostomy was necessary and no fistulas developed. The open cases had the highest rate of colostomy or

1024

abdominal abscess [6, 14, 15]. In patients who can tolerate general anesthesia and in instances where percutaneous approach is unsafe by virtue of overlying bowel loops, obesity, or multiple operations, diagnostic laparoscopy can afford several advantages. First, good visualization of the diseased area of colon is possible with the laparoscope introduced often in an alternative site in either upper quadrant or the midabdomen. Our preference is the use of the Veress needle for insufflation of CO2 to 1415 mmHg pressure. Use of the 5-mm cannula and 5-mm 0 laparoscope allows for determination of free areas in the abdominal wall for further placement of trocars (either 5 or 10 mm) and subsequent necessary dissection of abdominal wall adhesions. Our approach to the operative management of the patient with acutely complicated colonic diverticulitis: 1. Three to five trocar sites are used, placed in a general crescent pattern with the opening of the crescent toward the pathologythat is, usually the sigmoid colon (highly variable and thus no standard of trocar location can be given). 2. Laparoscopic bowel clamps (Glassman) are preferred for gentle bowel handling. 3. Avoid dissection of friable bowel/mesentery where possible (lavage and drainage) to accomplish definitive single-stage procedure after infection is resolved at a later time. 4. Remove laparoscopically resected bowel in a reservoir bag if any possibility of neoplastic process. 5. Use trocar site (enlarged lower quadrant or alternatively the trocar site in the left rectus sheath) to remove the specimen and then place the proposed diverting ostomy at that site. 6. Lavage the peritoneal cavity/abscess cavity copiously with dilute Betadine/normal saline solution. 7. Colonoscopic evaluation of the colon is essential to evaluate for neoplasm in the postoperative period (used intraoperatively where necessary). 8. In feculent peritonitis, conversion to open procedure may be advisable for adequate management/resection. 9. Closed suction drainage is used on all cases with abscess. 10. Postoperative broad-spectrum antibiotics are guided by intraoperatively obtained cultures. 11. Postoperative nutritional support (parenteral or oral) depends on return of bowel function.

Discussion Our successful laparoscopic management of complicated diverticulitis was accomplished in 43 of 58 patients or 74% with an overall laparoscopic success rate of 90.2% (148 of 164 patients) when elective cases are added. It appears from the early results of this brief experience that laparoscopy has a definite place in the treatment of acute complicated colonic diverticular diseaseand in our practice a very vital role. Complicated colonic diverticular disease has proven a continued challenge in an attempt to convert a given patient from an emergent or urgent status to an elective operative status as suggested by Rothenberger and Wiltz [13]. We present our approach to complicated

diverticular disease and have used the Hughes [9] classification in an attempt to delineate the severity of disease found at operation. While lavage and drainage of purulent peritonitis is effective, we strongly recommend to the less experienced surgeon the conversion to open procedure for patients with feculent peritonitis. Application of diagnostic laparoscopy has helped us to better evaluate the severity of the inflammatory response to allow drainage and lavage and to provide a means to resect disease while avoiding potential wound complications (one of 15 open cases6.6% vs one in 43 laparoscopic cases 2.3%) in acute complicated cases. No infections were noted in the elective laparoscopic group (zero of 106). The shift toward one-stage procedures for resection of colonic diverticular disease has made innovation, in the form of percutaneous drainage [6, 14, 15] of pericolonic abscesses, an appealing modality initially in preparation for later resective procedure. In those cases particularly where radiologically directed percutaneous drainage is not possible because of intervening bowel, where obstruction is part of the clinical picture, and when there is an uncertainty of diagnosis and lack of response to conservative bowel rest and antibiotics, we feel that diagnostic laparoscopy provides an excellent option and potentially can be used to treat some of the complications of diverticulitis without an open procedure. Essential to the laparoscopic lavage and drainage approach is the use of postoperative complete colonoscopy to exclude neoplasm. Management of acutely perforated left colon by primary resection with or without protective colostomy has been shown preferable to three-stage operation in reduction of morbidity and mortality [2]. We have shown in our experience that each of these options can be accomplished by minimal access means. Our experience parallels that of OSullivan et al. [11] in the laparoscopic management of generalized peritonitis secondary to perforated colonic diverticuli with similar hospitalization of 10.7 days, on average, postoperatively. In contrast to the lavage technique of OSullivan, we have chosen to place a closed suction drain in each of our cases. Management of colocutaneous, colovesical, and colovaginal fistulas has been shown best handled by single-stage technique [3, 7, 10, 16] where possible, and our experience has shown this approach is quite feasible with laparoscopic technique. Preoperative workups with barium enema, CT scan, and cystoscopy remain valuable aids to diagnosis in colon-related fistulas. In a series of 92 patients, Fazio et al. [3] had 80% having one- or two-stage resection and anastomosis with 77% of patients without fistula or stoma postoperatively. This can be accomplished laparoscopically, as noted by Hewett [8] and as seen in our small series. Comparing length of hospitalization is difficult to assess, but we feel our 10.7-day average for laparoscopic management of acute diverticulitis, 4.5 days for elective resection, and 13 days for our open resections compare very favorably with the open data of Alanis et al. [1]16.2 days for primary resection and anastomosis, 19.4 days for primary resection with Hartmann procedure, and 26 days for first admission of delayed resection using three-staged procedure. Postoperative hospitalization in our acute complicated 58 cases was reflective of age and its associated medi-

1025

cal problems: Those less than 70 years of age stayed 7.2 days and those greater than 70 years of age 14.2 days. We continue our experience with laparoscopic resection of colonic diverticular disease and feel our success rate justifies this approach. Our low morbidity for elective procedures (less than 5%) and lessened morbidity for the laparoscopic approach compared to open approach to acutely complicated diverticulitis make this our preferred approach. Summary Expanding upon our ongoing experience with laparoscopic surgery for colonic carcinoma and with bowel obstruction [4, 5, 12] we have reviewed our experience with minimal access surgery for complicated colonic diverticular disease. There were a total of 164 cases, 106 of which were laparoscopic elective resections. Fifteen required open or conversion to open and 43 were acute complications of diverticulitis managed laparoscopically or laparoscopically assisted. This experience bridges the time frame from November 1991 until March 1996, including the 43 patients treated by diagnostic laparoscopy and subsequent laparoscopic resection and/or lavage and drainage. Although this is not a randomized study, and we realize that there is difficulty in comparison, we feel application of diagnostic laparoscopy and of laparoscopic means for treatment in 43 of 58 patients led to improvement in postoperative patient status, decreasing the risk of wound infection, postoperative atelectasis, and the overall length of hospital stay. Diagnostic laparoscopy is the key new ingredient in our approach to this problem and resulted in primary laparoscopic resection and reanastomosis and in avoidance of ostomy in 29 of the 43 (67.4%) acute patients managed completely laparoscopically or laparoscopically assisted. No mortality was encountered. A select group of 18 patients were thus treated by diagnostic laparoscopy, lavage, and drainage only and have either been subsequently electively treated by laparoscopic resection and anastomosis or medically only, thus totally avoiding ostomy. We propose an approach of surgical care for complicated diverticular disease based on preoperative appropriate diagnostic studies such as CT scan, colonoscopy, contrast enema, and ultrasound, followed by diagnostic laparoscopy in those cases not responding to medical

therapy. In this way, complications of diverticular disease including abscess, perforation, fistula, and bleeding can potentially be managed by minimal access procedures, decreasing postoperative wound problems and overall morbidity. References
1. Alanis A, Papanicolaou GK, Tachos RR, Fielding P (1989) Primary resection and anastomosis for treatment of acute diverticulitis. Dis Colon Rectum 32: 933939 2. Auguste LJ, Wise L (1981) Surgical management of perforated diverticulitis. Am J Surg 141: 122127 3. Fazio VW, Church JM, Jagelman DG, Weakley FL, Lavery IC, Tarazi R, van Hillo M (1987) Colocutaneous fistulas complicating diverticulitis. Dis Colon Rectum 30: 8994 4. Franklin ME Jr, Dorman JP, Pharand D (1994) Laparoscopic surgery in acute small bowel obstruction. Surg Laparosc Endosc 4: 289296 5. Franklin ME Jr, Rosenthal D, Norem RF (1995) Prospective evaluation of laparoscopic colon resection versus open colon resection for adenocarcinoma. Surg Endosc 9: 811816 6. Greco RS, Kamath C, Nosher JL (1981) Percutaneous drainage of peridiverticular abscess followed by primary sigmoidectomy. Dis Colon Rectum 25: 5355 7. Grissom R, Snyder TE (1991) Colovaginal fistula secondary to diverticular disease. Dis Colon Rectum 34: 10431049 8. Hewett PJ, Stitz R (1995) The treatment of internal fistulae that complicate diverticular disease of the sigmoid colon by laparoscopically assisted colectomy. Surg Endosc 9: 411413 9. Hughes ESR, Cuthbertson AM, Carden ABG (1963) The surgical management of acute diverticulitis. Med J Aust 1: 780782 10. Mileski WJ, Joeh RJ, Rege RV, Nahrwold DC (1987) One-stage resection and anastomosis in the management of colovesical fistula. Am J Surg 30: 8994 11. OSullivan GC, Murphy D, OBrien MG (1996) Laparoscopic management of generalized peritonitis due to perforated colonic diverticula. Am J Surg 171: 432434 12. Phillips EH, Franklin ME Jr, Carroll BJ, Fallas MJ, Famos R, Rosenthal D (1992) Laparoscopic colectomy. Ann Surg 216: 703707 13. Rothenberger DA, Wiltz O (1993) Surgery for complicated diverticulitis. Surg Clin North Am 5: 975992 14. Saini S, Mueller PR, Wittenberg J, Butch RJ, Rodksy GV, Welch CE (1986) Percutaneous drainage of diverticular abscess. Arch Surg 121: 475478 15. Stabile BE, Puccio E, vanSonnenberg E, Neff C (1990) Preoperative percutaneous drainage of diverticular abscesses. Am J Surg 159: 99 105 16. Woods RJ, Lauery IC, Fazio VW, Jagelman DG, Weakley FL (1988) Internal fistulas in diverticular disease. Dis Colum Rectum 31: 591 596

New technology
Surg Endosc (1997) 11: 10361039

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

Prehensile atraumatic grasper with intuitive ergonomics

T. G. Frank, A. Cuschieri
Department of Surgery, Ninewells Hospital and Medical School, University of Dundee, Dundee, United Kingdom Received: 3 February 1997/Accepted: 28 February 1997

Abstract Background: Minimal access surgery (MAS) procedures that require tissue to be grasped are impeded by the design of current instruments. The use of graspers and forceps can result in tissue damage and is highly inappropriate when handling larger organs such as the bowel, liver, and spleen. In addition, current instruments have unnatural handling characteristics. A new type of tissue grasper is presented as a solution to these problems. Methods: The new grasper design was evolved through a process of setting basic requirements and proceeding through cycles of design, construction, evaluation, and redesign. Results: The main features of the new device are prehensile grasping by finger-like jaws, which retain tissue by capture, and a novel handle design with intuitive ergonomics. The jaws are interchangeable to suit differing surgical tasks and the handle and trigger mechanism are designed so that the surgeons forefinger movement mirrors the instrument jaw action. The grasper has been used in 32 MAS procedures with no indications of trauma. Conclusions: A grasper that functions by capture has been demonstrated to be an effective solution for atraumatic tissue handling during many MAS procedures. Key words: Atraumatic Laparoscopy Endoscopy Instrumentation Graspers

The manual instruments for minimal access surgery (MAS) include essential assisting devices for direct tissue manipulation (e.g., graspers and retractors). These are generally used in conjunction with other instruments for dissection, hemostasis, and reconstruction [3]. The functional design of all these instruments is constrained by the need to operate through small access ports. In addition, the necessary length of the shaft and the restraining effect of the port mean that there is not an easily assimilated correspondence between
Correspondence to: T. G. Frank

handle movement and movement of the functional instrument tip. As a result, many instruments, particularly those for tissue manipulation, have relatively poor functional and ergonomic characteristics. Accounts of the manual difficulties encountered when performing MAS procedures often dramatically contrast with the equivalent open procedures [1]. A prime example of this is tissue grasping, where direct use of the hand has been replaced by mechanical graspers. The contrast between the way that the hand holds tissue and the way that graspers hold tissue is striking (Fig. 1). For the larger organs particularly, the problems of using graspers are major due to the tendency to cause damage and the limited scope for translating or rotating the grasped tissue. In addition, tissue damage will be more severe when it is necessary to hold tissue for prolonged periods during the course of a lengthy operation. While a wide range of MAS manual handling instruments are now available, there is a conspicuous absence of devices designed specifically for manipulating larger organs without causing trauma. The grasper described here was designed to be prehensile (i.e., to grasp in a similar way to the hand) so as to avoid or minimize these problems. Where handling ergonomics are concerned, it is noticeable that where a hand function (e.g., grasping) is replaced by an instrument function, the action of the hand on the instrument does not mimic the instrument function or the hand function it replaces. The new instrument handle was designed to function intuitively by utilizing a grasping action from the forefinger to activate a compound (twojointed) input trigger. The latter was linked to the jaws so that the jaw movement replicated the finger movement.

Materials and methods

The design procedure for the new grasper was based on the paradigm of attempting to include all requirements from the beginning. Therefore, in addition to atraumatic grasping of large organs and intuitive handling ergonomics, the list of requirements included aspects of safety, reliability, ease of disassembly for cleaning, avoidance of traps for contaminating material, biocompatibility, tolerance of autoclaving, replacement of worn or damaged parts, and easy incorporation of future design changes. Added

1037

Fig. 1. Hand grasp, with circular capture, compared with instrument grasp.

to this was a general effort to ensure that the device would meet the likely demands of a manufacturer. A modular design was chosen at the outset by separating the design into three parts: grasping jaws, functional handle, and the transmission linking handle function to jaw movement. This has practical significance since the transmission may be a standard component of many instruments while the handle may be suitable for a more limited number of other instruments. The instrument jaws then take the form of an interchangeable part and could be supplied in a range of shapes and sizes. A modular approach also simplifies the development process. The final design was the result of continuous collaboration between endoscopic surgeon and design engineer. In general, development followed a cyclical process: concept engineering drawing prototype construction evaluation new or altered concept. Initial designs were evaluated in the laboratory and surgical skills training systems. Later designs, following tests for safety and reliability, were evaluated in theater. Initial prototypes were constructed of plastic and aluminum for ease of machining. Prototypes for clinical evaluation were constructed from medicalgrade stainless steel. The working principles and critical dimensions of the jaw and trigger mechanisms (described below) were established in two ways. Firstly, large-scale models using simple links and pins were used to establish the basic working arrangement of the mechanisms. Secondly, the designs were drawn using computer-aided design software (Autocad, Corel Draw, or Mastercam) and the exact dimensions were determined. In later prototypes, parts were made using computer numerical control (CNC) of milling and turning. The CNC approach permits storage of the cutting programs, which simplifies the implementation of minor changes and gives automatic manufacture of multiple copies.

Fig. 2. Action of prehensile grasping jaws.

Results Technical results The chosen jaw design consists of symmetrically opening main jaw segments with an additional finger joint on each side to give grasping tips (Fig. 2). This assembly is connected to a transmission system formed by three concentric tubes. The instrument jaw size was chosen to suit handling of the small and large intestine in the first instance and the instrument diameter was set at 10 mm because a smaller diameter would put too high a demand on material strength. Both jaw parts have a semicircular cross section and operating links connect the tips to the central shaft via a pivot connected to the outer tube. The jaws are opened by movement of the central tube relative to the outer tube. The main jaws pivot on the intermediate shaft and movement of this shaft relative to the central tube angulates the tips. This arrangement allows the jaw tips to be angulated independently of the main jaw segments. The use of concentric transmission shafts allows for easy sealing against ingress of fluids. A release button (Fig. 2) allows the jaw and shaft assembly to be extracted from the outer tube for cleaning and the shafts are further detach-

able from the jaw assemblies so that the latter have the minimum number of parts. The outer tube can be rotated in small steps by means of a collar (Fig. 3). This rotation is transmitted to the jaws because they are locked to the outer tube by the release button but are free to rotate by virtue of rotatable connections to the handle trigger mechanism. The handle requires two input triggers: one for the main jaw segments and one for the jaw tips. These are combined into a compound trigger which is functionally the same as the jaw mechanism. The trigger pivots are coaxial with the metacarpophalangeal and proximal interphalangeal joints of the forefinger, respectively (Fig. 4). This arrangement permits matching forefinger with instrument jaw movement. The handle includes release buttons for detaching the shafts and a knob for locking the jaws in any position (Fig. 3) so that the forefinger can be relaxed during prolonged grasping. Clinical evaluation The grasper has been used surgically for 12 months in a single operating theater on 32 patients who underwent laparoscopic surgery. The results of the different activities are summarized in Table 1. It can be seen that the most frequent use has been as a retractor and that no trauma was observed even after retraction periods lasting up to 1 h. Discussion For open surgery, there are many types of grasper (or forceps), reflecting the range of tissue types to be grasped [4].

1038 Table 1. Summary of clinical use Procedure Laparoscopic pancreatic necrosectomy Staging of pancreatic cancer Staging of gastric and esophageal cancer Adhesiolysis Fig. 3. Side view of handle. Colon resections 6 Number of cases 1 4 Activities Removal of slough Holding stomach and colon Holding stomach and duodenum Palpating the pancreas Holding stomach during entry into lesser sac and biopsy of coeliac nodes Holding small and large intestine with circular capture beyond the bowel Holding colon during dissection with circular capture beyond the bowel Holding mobilized colon Reduction of stomach from chest Observations No trauma or bleeding No trauma

No trauma

No trauma

No trauma

Laparoscopic rectopexy Paraesophageal hernia

2 3

No trauma No trauma

Fig. 4. Top view of trigger mechanism.

Some of these (e.g., Babcocks and Littlewoods) are potentially atraumatic because they hold by encircling. However, these devices are intrinsically too large for direct adaptation to MAS. The difficulty with conventional short-jawed graspers is that they retain tissue by friction and, since tis-

sues are generally very slippery, a high gripping force is required. The pinching effect can be reduced slightly by making the grasper jaws much longer. However, severe crushing can still occur where tissue is near to the jaw hinge [3]. A similar difficulty has been described in relation to endoscopic clamps [2], where the solution was a clamp with parallel jaw motion. While this could be a feasible design for a grasper, such a device would still rely on friction, particularly when grasping tissue from above. Consideration of these issues emphasized that the design should be based on encirclement or capture and suggested that this can only be achieved easily by integrating additional articulations into the jaws. The primary problem of finding a satisfactory design for a grasper that functions by capture has been solved in the present design but at the expense of resorting to a 10-mm instrument diameter. The subjective conclusion from preliminary clinical use is that the design does provide effective and atraumatic grasping of large organs, particularly for basic retraction purposes. In addition, simple mechanical testing and prolonged use demonstrated that the device is adequately robust. However, during the trial period, nursing staff found the disassembly release buttons difficult to operate and it was discovered that ultrasonic cleaning caused some fixing screws to loosen. During surgical use, questions arose as to the optimum size and design of the instrument jaws for both general and special purposes. The handle was also the subject of debate. The fact that the handle has to be built either right or left handed proved to be a problem and the facility for rotation was found to be important for comfort and in cases where a low instrument angle would otherwise cause the handle to impinge on the abdomen of the patient. These observations will be considered during the final design and evaluation stage. This will include comparative studies of different jaw shapes and sizes with the aim of finding the best general purpose design and optimal designs

1039

for tasks such as suturing and dissection. Studies will also compare the utility of using a left and right pair of handles with (1) a very simple two-input handle without the intuitive layout of the current design and (2) a new two-input handle of coaxial design which has the advantage of allowing rotation without the need for a rotatable joint in the instrument [4]. The latter design, which maintains the principle of intuitive ergonomics, has recently been completed in Dundee University. While completion of the final development stage should allow the device to pass to a manufacturer, research in Dundee relating to the grasper will continue. In particular, the device will be involved in the development of motorized instrument functions, articulating instrument

shafts, telerobotic control, and force and tactile sensing and feedback. References
1. Cushieri A (1995) Whither minimal access surgery: tribulations and expectations. Am J Surg 169: 919 2. Frank T, Willetts GJ, Cuschieri A (1995) Detachable clamps for minimal access surgery. Proc Inst Mech Eng 209(H2): 117120 3. Melzer A, Buess G, Cuschieri A (1994) Instrumentation and allied technology for endoscopic surgery. In: Cuschieri A, Buess G, Pe rissat J (eds) Operative manual of endoscopic surgery 2. Springer-Verlag, Berlin 4. Patkin M, Isabel L (1995) Ergonomics, engineering and surgery of endosurgical dissection. J R Coll Surg Edinb 40: 120132

Surg Endosc (1997) 11: 986990

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

The anatomical basis of anal endosonography

A study in postmortem specimens
B. Gerdes,1 H. H. Ko hler,2 A. Zielke,1 O. Kisker,1 P. J. Barth,2 B. Stinner1
1 2

Department of General Surgery, Philipps-University of Marburg, Baldingerstrae, 35033 Marburg, Germany Department of Pathology, Philipps-University of Marburg, Baldingerstrae, 35033 Marburg, Germany

Received: 28 January 1997/Accepted: 28 February 1997

Abstract Purpose: Anal endosonography is an imaging modality new to the diagnostic workup of incontinence. Interpretations even of normal endosonomorphologic findings now vary considerably. The conjoined longitudinal muscle (LM), a widely ignored structure, has until recently not been fully recognized by anal endosonography. The aim of this study, therefore, was to accurately determine the normal anatomy of the anal canal and correlate it with the findings obtained by anal endosonography. Methods: Eight postmortem specimens of the anal canal were examined by endosonography. The findings were correlated with macroscopical dissection and gross sectional histology of the same specimens. Results: The external echogenic ring is composed of two anatomical structures: the LM and the external anal sphincter (EAS). However, during anal endosonography the LM cannot always be differentiated from the EAS. Histologically, the relation of the diameters of the LM and the EAS ranged from 0.45:1 to 1.25:1. The narrow hyperechogenic ring between the inner hypoechoic layer and the external hyperechoic ring is an artificial finding that cannot be related to a distinct anatomical structure and most likely represents a sonographic interface. Conclusions: This study exactly outlines the relation of diameters of the conjoined longitudinal muscle and external anal sphincter for the first time. Until now, the LM has been underestimated in its dimensions. The role of such a thick muscular structure should be included in the conception of anal continence in the future. Especially in view of the fact that anal endosonography is increasingly used in the diagnostic workup of incontinence and fistula in ano, it is essential to understand the anatomical basis of endosonography. This study accurately delineates the sonomorphology of the anal muscles. When viewed in light findings reported

here, endosonographic findings in diseases of the anal canal are now based on a correct idea of the correlation between endosonomorphology and anal anatomy. Key words: Anal endosonography Longitudinal muscle External anal sphincter

Correspondence to: B. Gerdes

Endoluminal ultrasound is an imaging modality used increasingly in the diagnosis of fecal incontinence [46, 9]. Nonetheless studies correlating the sonomorphologic pattern with the anatomy of the anal canal are rare [8, 17, 18] and endosonographic findings offer conflicting interpretations [46, 8, 9, 11, 12, 15, 17, 18]. The outermost echogenic band is interpreted by most authors as the endosonographic equivalent of the external anal sphincter (EAS) [6, 15]. Some studies described a separate hyperechogenic ring inside this layer caused by the longitudinal muscle (LM) [11, 17]. Others have attributed the outer ring of the EAS to the LM [8] or have completely ignored this muscle [12, 18]. Moreover, a narrow hyperechogenic ring between the internal anal sphincter (IAS) and the outer hyperechoic band has also been suggested to represent the LM [17], whereas others have proposed this band to represent only an additional intersphincteric layer [8, 9, 12]. However, currently there are no data available that would allow one to clearly delineate the contribution of the LM to the sonomorphology of the anal canal. Especially in the diagnostic workup of incontinence and fistulas in ano, it is crucial to understand the anatomical basis of endosonography. Therefore, the purpose of this study was to obtain a precise correlation between the anatomic and ultrasonographic morphology of the anal canal. Postmortem studies were used to compare endosonographic findings with macroscopic dissection and gross-sectional histology following dissection and removal of each layer in

987

Fig. 1. In vivo (A) and postmortem (B) examinations of the anal canal in the region of the pectinate line reveal identical sonomorphogical findings: A central narrow echogenic band, an inner homogeneous hypoechoic ring, and a thick external echogenic band. The correlation of sonomorphology and anatomy is listed in Table 2. Fig. 2. Five corpses were examined stepwise (each sector of the image represents one step, respectively): Step Step Step Step Step 1: 2: 3: 4: 5: Postmortem anal endosonography with the corpse in a supine left lateral position (1), Endosonography of the freshly removed specimen of the anal canal in a water-bath (2), Endosonography of the same specimen after 3 days of fixation in 4% formalin with plastic cone inserted into the anal lumen (3), Section of the anal specimen in the plane of the pectinate line (4) and Dissection of a gross histological section in this region (5). In the histological sector the anoderm and subepidermal tissue (a) can be identified. In the sonographic sectors this inner layer is compressed by the endoprobe. b and c show the smooth muscles of both the internal anal sphincter and the longitudinal muscle. d shows the striated muscle of the external anal sphincter. c and d contribute to the outermost hyperechoic ring.

sequence. This allowed one to clarify the contribution of the LM to the anatomy of the anal canal and to correlate this to findings obtained by anal endosonography. In particular, this study determined the relation of diameters of the LM and the EAS. Materials and methods Ultrasound equipment
Endosonography of postmortem specimens was carried out using a Combison 311 ultrasound device (Kretztechnik, Zipf, Austria) with a rotating bifocal multiplane rectal transducer (ERW 7/10 AK) that produces a transverse sector scan display of 360 and a longitudinal of 135, respectively. The mechanical 7.5/10-MHz endoprobe was equipped with a sonolucent endopiece 22 mm wide for direct acoustic coupling. The ultrasound transducer obtained images with 7.5 MHz and 10 MHz, respectively. Depending on the frequency used it provides an axial/lateral dissolution of <0.5 mm/ <0.8 mm and <0.4 mm/<0.65 mm. The penetration depth ranges from 340 mm/230 mm.

Table 1. Diameters of anal muscles in five post mortem specimensa IAS: mm (range) In situ In water-bath Formalin-fixed Gross section Relation LM: EAS 2.46 (2.22.7) 2.66 (2.13.1) 2.54 (1.73.0) 2.66 (2.33.0) EAS and LM: mm (range) 5.68 (2.67.6) 6.02 (3.97.5) 6.20 (4.57.4) 5.92 (4.57.6) 0.87 (0.451.25)

a IAS internal anal muscle, EAS external anal muscle, LM longitudinal muscle, in situ endosonography of the corpses in situ, in waterbath endosonography in the freshly removed specimen, formalinfixed endosonography after fixation of the specimen in 4% formalin, gross section measurement in a histological gross section in the plane of the linea dentata

In vivo studies
In our clinical investigations we performed 200 endoanal sonographies in patients without anal diseases (e.g., during examination for rectal cancer). We compared the sonomorphology of these patients with postmortem studies (Fig. 1).

Studies of postmortem specimens

Informed consent was obtained from relatives for this investigation. Eight fresh postmortem specimens were obtained from five male and three female individuals (median: 68 years, range 3883) who had no previous history of any anal disease including incontinence. Before death, three patients experienced a period of relaxation and artificial ventilation. Anal endosonography was performed in a left lateral supine position. The interval between time of death and performance of endosonography of the respective corpses ranged from 9 to 24 h. Postmortem and in vitro scanning

was performed and the results were compared to the anal anatomy in two different ways. At first (Fig. 2), in a group of five specimens, a direct correlation of sonographic findings to histological sections of the anal canal was obtained by gross histological sections in the transversal plane of the anal canal. To evaluate whether fixation alters ultrasound depiction of the sphincters, repeat ultrasound was performed at each step of dissection and fixation. Following postmortem scan and extirpation of the anal canal and the lower third of the rectum anal endosonography was repeated in a water bath. Then, specimens were fixed in 4% buffered formalin solution for 3 days. During fixation, a plastic cone of the same diameter as the sonographic transducer (22 mm) was inserted into the anal canal in order to prevent shrinkage of the tissue. After fixation, the specimen was again transferred to a water bath and anal sonography was repeated. The endosonographic thickness of the internal hypoechoic ring was measured in all four quadrants of the anal canal, the outer ring was measured in the left and right lateral quadrants, and mean values were calculated. Following endosonography, multiple sections of the specimen in the transversal plane of the anal canal in the region of the pectinate line were obtained. In a region without artifacts due to the cutting procedure, measurement of the diameters of the respective muscle layers (IAS, LM, EAS) were taken. The diameters yielded by the different steps during fixation and preparation as well as of the respective ultrasound examinations were compared. Additionally, the muscular thickness of the LM and the EAS was quantified on histological sections and the relationship between these values was determined (Table 1).

988

Fig. 3. Stepwise dissection of the boundaries of the different anal layers with photographic documentation of each step (a), endosonographical imaging in a water-bath (b), and (c) histological section of the removed tissue. At first, the perianal fatty tissue was removed (I), followed by the external anal sphincter (EAS) (II), the longitudinal muscle (LM) (III), the internal anal sphincter (IV), and finally the subepithelial tissue (V) including the M. canalis ani. The outer hyperechoic ring was identified as the EAS and the LM, whereby the LM contributes to the innermost and the EAS to the outermost part of this ring. The narrow echogenic band in bIII is a borderline interface, i.e., ultrasonographic artifact without an anatomical correlate. No difference between steps bIV and bV was observed.

989 Table 2. Correlation of anal sonomorphology, macropathology, and histology in a transverse plane Sonomorphology 1 2 3 4 5 Inhomogeneous reflections Thick hyperechoic ring Narrow hyperechoic ring Thick hypoechoic ring Narrow echogenic ring Macropathology Perianal fatty tissue M. sphincter ani externus and longitudinal muscle No anatomical equivalent M. sphincter ani internus Anoderm and subepidermal tissue (fibrous tissue, M. canalis ani) Histology Fat and fibrous septa Striated muscle Smooth muscle Smooth muscle Anal mucosa, few bundles of smooth muscle, fibrous tissue

Secondly (Fig. 3), to allow an accurate delineation of the ultrasonographic contribution of the different anatomical boundaries, repeated scanning after dissection and removal of each anatomical layer in sequence was performed in three specimens. Specimens were dissected in a similar manner as described above. However, in this study, removal of each layer was performed from the outermost to the innermost layer of the anal canal, removing the perirectal fat first, then the EAS, the LM, the IAS, and finally the subepithelial layer. At each step ultrasound was repeated in a waterbath. After each step the removed tissue was examined histologically. This procedure was performed in the transverse plane in two specimens as well as in the longitudinal plane in the third one. All sections were cut at a thickness of about 5 m and stained HE and EVG.

with endoanal ultrasound and histological examination of the removed tissue rendered it possible to correlate endosonographical findings and anatomical boundaries exactly. The sonographic patterns produced by the different anatomical layers were as follows (Table 2): 1. The outermost echoinhomogeneous area during endosonography represents the perianal fat and is composed of fatty tissue and fibrotic septa at histology. 2. The outer hyperechoic ring was identified as the EAS and LM: the LM contributes to the innermost and the EAS to the outermost part of this ring. Only occasionally, these two muscles are separated by endosonography. 3. Between the outer hyperechoic ring and the inner hypoechoic ring, a small band of hyperechogenicity is encountered. However, at histology, no tissue equivalent was found. This hyperechoic ring therefore represents an ultrasonographic artifact, most likely an interface. 4. The homogeneous inner hypoechoic layer represents the IAS and usually is a well-demarcated structure. 5. The innermost hyperechoic layer represents the anoderm and subepithelial tissue. Discussion Endorectal ultrasonography is well established in the staging of rectal carcinoma [1, 14]. The anatomic substrate and the respective sonomorphology have been clearly outlined in this condition [2, 3]. However, anal endosonography is new in the diagnostic workup of incontinence and fistulain-ano [46, 9]. Interpretation even of the normal endosonomorphologic findings varies considerably [8, 17, 18]. Especially the conjoined longitudinal muscle (LM) has not been fully appreciated during anal endosonography. The aim of the present study, therefore, was to establish a precise correlation of the anatomic and ultrasonographic morphology of the anal canal with special reference to the LM. Early postmortem endosonographic depiction of the anal canal perfectly corresponded to the findings obtained during in vivo examinations [6] (Fig. 1). These are characterized by a small hyperechoic ring close to the transducer, followed by a thick inner hypoechoic ring that is surrounded by an outer hyperechoic ring. In the literature the thick inner hyperechoic ring is uniformly accepted to represent the IAS [6, 8, 17] and the outer hyperechogenic ring to represent the EAS [6, 8]. Between these two bands a small hyperechoic band can usually be demonstrated, which has been referred to as an additional intersphincteric layer [6, 8, 9] and has been suggested to represent the LM [11, 17]. However, this conception is wrong.

Results Anal endosonography performed in postmortem specimens in situ, as well as after removal and fixation of the anal canal, revealed a pattern of layers that corresponded to the findings obtained in patients in vivo (Fig. 1). The layers of the anal canal that were shown in the different steps of the sonographic examination did not change during dissection and fixation (Fig. 2). They were visible on the histological sections in the transverse plane, dissected in the region of the pectinate line. The innermost layer of the anal canal consists of the anoderm and the subepithelial layers, which are depicted as a narrow hyperechoic band endosonographically, compressed by the ultrasound probe. Outside the subepithelial tissues there is a well-demarcated hypoechoic band corresponding to the IAS. The thick hyperechoic layer surrounding the IAS corresponds to two anatomical structures: the LM and the EAS (Table 2). The inner border of the internal anal sphincter could be discriminated much better using a 10-MHz transducer compared to 7.5 MHz. Independently of the chosen transducer the two parts of the outer hyperechoic ringLM and EAScould hardly be distinguished due to their similar echogenicity in three of the five specimens. The diameters of the anal muscles are shown in Table 1. The fixation procedure only had a negligible influence on the muscular thickness, caused by formalin-induced shrinkage of the tissue. The relation between the diameters of the LM and the EAS ranged from 0.45:1 to 1.25:1 in the gross section. The highest value comes from a 38-year-old male who died after a 3-month period of relaxation and controlled ventilation. Marked atrophy of the striated skeletal muscles was evident. The assumption that the striated EAS would likewise be atrophic was confirmed by endosonography in situ as well as in the postmortem specimen and the histological section revealing a relation of the LM:EAS of 1.25:1. In the second part of this study, stepwise dissection of the anal layers from peripheral to central in combination

990

In the first part of our study we obtained a precise ultrasonographic depiction of the anal canal. In the second part we demonstrated these findings to be exactly correlated to the macropathological anatomy of the anal canal by correlating its ultrasonographic composition with the muscular boundaries of the anal canal. For this purpose, we removed each anatomic layer of the anal canal in sequence, from the outermost to the innermost structure encountered and correlated ultrasonography with histology at each step of dissection (Fig. 3). Dissection of the layers from the innermost to the outermost layer, as suggested by others [17], would have had a significant impact on the ultrasonographic depiction of the remaining layers due to unpredictable changes of focus and penetration depth. Thus, if a realistic appraisal of the echogenic patterns produced by the different layers of the anal canal were to be obtained, dissection from the outside to the inside is mandatory. This technique allowed a precise identification of the tissue boundaries and accurate correlation of macroscopic aspect, endosonography, and corresponding histological findings. As expected, fixation by 4% formalin induced only minor and insignificant shrinkage of tissue [16] and did not impair the ultrasonographic depiction of the respective layers. Based on the results of the present study, we are now able to deliver a precise description of the ultrasonographic morphology of the anal canal (Table 2, Figs. 2 and 3). The perianal fatty tissue is depicted as an echo-inhomogeneous area that is visualized peripherally and adjacent to the outermost hyperechoic ring. This outermost hyperechoic ring correlates to the EAS, which contributes to the peripheral aspects of this ring, and the LM, which contributes to the central aspects of the same hyperechoic ring. Only occasionally are these two muscles separable by endosonography. We also found that the LM can attain the same thickness as the EAS (Table 2). Thus, we have established that the LM is depicted within the centrally located aspects of the outer hyperechoic ring during endoanal ultrasonography. The contribution of the LM to the ultrasonographic morphology of the anal canal and its dimensions have clearly been underestimated until now [8, 11]. The LM is not well described in the surgical literature [10]. The LM is a continuation of the longitudinal muscle layer of the rectum and condensates at the height of the anal canal into a rather prominent muscular layer. Muscular extensions of the LM penetrate the IAS and result in the canalis ani muscle [7]. Muscle fibers of the LM also penetrate the pars subcutanea of the EAS and insert at the height of the perineal cutis as the corrugator ani muscle [10]. Thus, the LM represents a conjunction between the smooth and striated muscles of the sphincter complex (conjoint longitudinal muscle). Neither the corrugator ani muscle nor the canalis ani muscle can be demonstrated during endoanal ultrasound. However, the functional contribution of the LM as a part of the anal continence organ has not yet been outlined. Also, the widely accepted topographic classification of fistulas-in-ano by Parks et al. does not consider this muscle [13]. The thick inner hypoechoic ring represents the IAS and almost always is well delineated during ultrasonographic examination. The hyperechoic band between the IAS and the LM, referred to as an additional intersphincteric layer by Law and Bartram [8], does not represent the LM. We

have shown that this band has no anatomical substrate and therefore represents an ultrasonographic artifact, most likely an interface (Fig. 3, bIII). A small hyperechogenic band exits between the ultrasound transducer and the IAS, which histologically corresponds to the anoderm and the subepithelial tissue. This area is compressed by the ultrasound probe and current instrumentation does not allow for further resolution of this particular area [12]. In summary, we have established that during endoanal ultrasonography the LM is depicted within the innermost aspects of the outer hyperechoic ring. The dimensions of the LM have been underestimated until now since the anatomic substrate of the LM is at least similar to that of the EAS. Particularly with respect to the diagnostic workup of incontinence and fistulas-in-ano, it will be necessary to further elucidate the role of the LM in the pathogenesis of these diseases. The conceptual inclusion of the LM will be an important future task in anal surgery. References
1. Beynon J, Foy DMA, Roe AM, Temple LN, McMartensen NJ (1986) Endoluminal ultrasound in the assessment of local invasion in rectal cancer. Br J Surg 73: 474477 2. Beynon J, Foy DMA, Temple LN, Channer JL, Virjee J, McMartensen NJ (1986) The endosonic appearance of normal colon and rectum. Dis Colon Rectum 29: 810813 3. Beynon J, McMartensen NJ, Foy DMA, Channer JL, Virjee J, Godard P (1986) Endorectal sonography: laboratory and clinical experience in Bristol. Int J Colorectal Dis 1: 212215 4. Eckardt VF, Jung B, Fischer B, Lierse W (1994) Anal endosonography in healthy subjects and patients with idiopathic fecal incontinence. Dis Colon Rectum 37: 235242 5. Emblem R, Dhenens G, Ragnar S, Mrkrid L, Aasen AO, Bergan A (1994) The importance of anal endosonography in the evaluation of idiopathic fecal incontinence. Dis Colon Rectum 37: 4248 6. Gerdes B, Lausen M (1993) Endosonographie des Analkanals bei Funktionssto rungen des BeckenbodensErste Erfahrungen. Chirurg 64: 130133 7. Hansen HH (1976) Die Bedeutung des Musculus canalis ani fu r die Kontinenz und anorectale Erkrankungen. Langenbecks Arch Chir 341: 2337 8. Law PJ, Bartram CI (1989) Anal endosonography: technique and normal anatomy. Gastrointest Radiol 14: 349353 9. Law PJ, Kamm MA, Bartram CI (1991) Anal endosonography in the investigation of faecal incontinence. Br J Surg 78: 312316 10. Lunniss PJ, Phillips RKS (1992) Anatomy and function of the anal longitudinal muscle. Br J Surg 79: 882884 11. Nielsen MB, Hauge C, Rasmussen OO, Sorensen M, Pedersen JF, Christiansen J (1992) Anal sphincter size measured by endosonography in healthy volunteers. Acta Radiol 33: 453456 12. Papachrysostomou M, Pye SD, Wild SR, Smith AN (1993) Anal endosonography in asymptomatic subjects. Scand J Gastroenterol 28: 551556 13. Parks AG, Gordon PH, Hardcastle JD (1976) A classification of fistula in ano. Br J Surg 63: 112 14. Saitoh N, Okui K, Sarashina H, Suzuki M, Arai T, Nunomura M (1986) Evaluation of echographic diagnosis of rectal cancer using intrarectal ultrasonic examination. Dis Colon Rectum 29: 234242 15. Scha fer A, Enck P, Fu rst G, Kahn Th, Frieling T, Lu bke HJ (1994) Anatomy of the anal sphincters. Dis Colon Rectum 37: 777781 16. Sellner W (1989) Fixierung histologischer Pra parate. In: Romeis B (ed) Mikroskopische Technik. 17th ed. Urban & Schwarzenberg, Mu nchen, pp 69111 17. Sultan AH, Nicholls RJ, Kamm MA, Hudson CN, Beynon J, Bartram CI (1993) Anal endosonography and correlation with in vitro and in vivo anatomy. Br J Surg 80: 508511 18. Tjandra JJ, Milsom JW, Stolfi VM, Lavery I, Oakley J, Church J, Fazio V (1992) Endoluminal ultrasound defines anatomy of the anal canal and pelvic floor. Dis Colon Rectum 35: 465470

Surg Endosc (1997) 11: 10321033

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

Necrotizing fasciitis after laparoscopic surgery

P. M. Hewitt, K. H. Kwong, W. Y. Lau, S. C. S. Chung, A. K. C. Li
1

Department of Surgery, Prince of Wales Hospital, Chinese University of Hong Kong, Shatin, New Territories, Hong Kong

Received: 26 August 1996/Accepted: 31 December 1996

Abstract. We report two cases of necrotizing fasciitis following laparoscopic surgery and discuss contributing factors, as well as treatment of this rare complication. Key words: Necrotizing fasciitis Soft-tissue infections Laparoscopic surgery

nal pelvic fixator, he was confined to bed with opiate analgesia. Two weeks later, he developed fever and abdominal pain. Ultrasound and CT scan were consistent with acalculous cholecystitis, and at laparoscopic surgery, a gangrenous perforated gallbladder was removed. Postoperatively, local tenderness and serous discharge were noted at the right-lateral operating port site. Twenty-four hours later, he had rapidly spreading cellulitis of his right side. Wound exploration revealed necrotizing fasciitis. Several debridements and prolonged antibiotic therapy were necessary before the patient recovered. Recovery was based on isolation of klebsiella from the gallbladder pus, as well as culture of necrotic abdominal wall tissue which yielded enterococci and klebsiella species.

Necrotizing fasciitis is a rapidly progressive infection of soft tissues that is accompanied by high mortality [3, 4]. It may occur after injury, which is often relatively minor, or postoperatively, particularly in patients with peritonitis or perineal sepsis [2, 3]. Recently, we encountered this complication after laparoscopic surgery. Case reports Case 1
A healthy 65-year-old woman presented with a short history of abdominal pain and generalized peritonitis. At laparoscopy, a perforated duodenal ulcer was found. This was sealed using gelatin sponge and tissue-glue. After the operation, a hematoma was noted at an epigastric port site. A week later the patient developed fever, with tenderness and discoloration around the wound. She became progressively toxic until the 11th postoperative day, when wound exploration revealed extensive soft-tissue necrosis and pneumatosis. Surgical debridement was performed and broadspectrum antibiotic therapy was commenced. Microbiological cultures from the initial surgery were negative, but the necrotic tissue yielded a mixture of bowel-based organisms, including enterococci and acinetobacter. After a 2-week period of appropriate antibiotics and dressings to the wound, the large defect was skin-grafted and the patient recovered uneventfully.

Discussion Necrotizing fasciitis has been described after open operations, particularly where there is significant fecal contamination [3], but we are not aware of any reports after laparoscopic surgery. The condition is one of a spectrum of soft-tissue infections characterized by rapidly progressive necrosis of subcutaneous tissue and muscular fascia, with variable involvement of skin and muscle. It is most often due to polymicrobial infection which is probably synergistic in nature [24]. The predominant organisms isolated in polymicrobial infections are gram-negative enteric bacilli, enterococci, and streptococcal and staphylococcal species, while Streptococcus pyogenes is most often cultured where a single bacterium is involved [3]. Factors which predispose patients to this life-threatening complication include obesity, malnutrition, malignancy, chronic alcoholism, drug abuse, peripheral vascular disease, diabetes mellitus, and immunosuppressive therapy [24]. Both of our patients were, however, healthy. In case 1, onset of fasciitis was delayed and may have been due to colonization of the wound hematoma by bacteria from a distant source [5]. In the second patient, surgical approach was limited by an external pelvic fixator, which necessitated oblique insertion of laparoscopic ports. This resulted in opening up of tissue planes and creation of dead space. Another contributing factor may have been the aerosol effect of CO2 insufflation, which disseminates infected peritoneal fluid and forces it into the tissues, particularly on withdrawal of laparoscopic ports, although pneumoperitoneum per se does not cause systemic sepsis [1].

Case 2
A 38-year-old driver sustained multiple injuries in a road traffic accident. After resuscitation and stabilization of fractures, which included an exter-

Correspondence to: A. K. C. Li

1033

Delay in diagnosis and operative treatment of necrotizing fasciitis are the most important factors influencing mortality [3, 4]. The diagnosis is essentially clinical, although this may be difficult, because initial signs are often minimal, particularly where deep tissue sites are involved. Wound pain and discharge may be the only early features of infection. Skin discoloration or blistering may occur later, but these manifestations are merely the tip of the iceberg as the infection spreads extensively along fascial planes [2, 4]. The condition should be suspected in any patient with systemic symptoms that are out of proportion to local signs of inflammation [2]. Early soft-tissue and fascial exploration, with biopsy and frozen-section examination in doubtful cases, is the most expeditious route to diagnosis [3]. Radiological investigations are of limited diagnostic value. Occasionally air may be seen in the soft tissues [2, 3], but this is not helpful after laparoscopic surgery, where surgical emphysema may result from uncomplicated pneumoperitoneum. Perioperative fluid resuscitation, correction of systemic acidosis and electrolyte abnormalities, and cardiopulmonary and nutritional support are of major importance in management of the disease. However, surgery remains the most important aspect of treatment [2]. This should include wide debridement beyond the margins of necrotic fascia, followed by regular wound inspections and repeated debridement where necessary [3, 5]. Patients should also receive empiric broad-spectrum antibiotics initially, with subse-

quent therapy guided by intraoperative culture results [3, 4]. Hyperbaric oxygen has been used as an adjunctive therapy in some instances [4], but its value remains unproven [2, 4] and it should probably be avoided [3]. Although necrotizing fasciitis is rare after laparoscopic surgery, a high index of suspicion with early wound exploration in suspected cases are advisable. Careful technique, including adequate hemostasis and perpendicular insertion of ports, may prevent this catastrophic complication. Patients should also be instructed in wound management, particularly when they are discharged from hospital soon after laparoscopic operation.

References
1. Gurtner GC, Robertson CS, Chung SCS, Ling TKW, Ip SM, Li AKC (1995) Effect of carbon dioxide pneumoperitoneum on bacteraemia and endotoxaemia in an animal model of peritonitis. Br J Surg 82: 844848 2. Laucks SS (1994) Fourniers gangrene. Surg Clin North Am 74: 1339 1352 3. McHenry CR, Piotrowski JJ, Petrinic D, Malangoni MA (1995) Determinants of mortality for necrotizing soft-tissue infections. Ann Surg 221: 558565 4. Pessa ME, Howard RJ (1985) Necrotizing fasciitis. Surg Gynecol Obstet 161: 357361 5. Svensson LG, Brookstone AJ, Wellsted M (1985) Necrotizing fasciitis in contused areas. J Trauma 25: 260262