International Journal of Obstetric Anesthesia (2001) 10, 192197 2001 Harcourt Publishers Ltd doi: 10.1054/ijoa.2000.

.0833, available online at http://www.idealibrary.com on

CASE REPORT

Anaesthesia for caesarean section in a patient with Ehlers-Danlos syndrome and mitral valve prolapse
P. Dill-Russell, L. St. John Jones*
Anaesthetic Department, St. Georges Hospital, Tooting, London, UK, *Crawley Hospital, Crawley, West Sussex, UK SUMMARY. We report the successful use of single-shot spinal anaesthesia for elective caesarean section in a multigravid woman with Ehlers-Danlos syndrome, complicated by mitral valve prolapse. This is a rare inherited collagen disorder that has been reported to have a high pregnancy-related morbidity and mortality, particularly from uterine rupture and uncontrolled bleeding from friable vessels, coagulopathy and major vessel rupture. The literature concerning obstetric anaesthetic management for different types of Ehlers-Danlos syndrome and the management of mitral valve prolapse in pregnancy are reviewed. 2001 Harcourt Publishers Ltd

INTRODUCTION The problems of Ehlers-Danlos syndrome in relation to pregnancy and anaesthesia have previously been reported in the literature, but have usually referred to type IV Ehlers-Danlos syndrome, which is relatively uncommon. Due to the rarity of the condition and probable under-reporting of non-complicated cases, it has not been possible hitherto to quantify risk in relation to specic types of Ehlers-Danlos syndrome. By reviewing the literature and discussing the problems that are encountered by these patients in pregnancy, we attempt to outline the differences between the types of Ehlers-Danlos syndrome that are relevant to anaesthetists. It is important that all those involved with the care of these patients understand the potential dangers in some types of the condition. As far as the authors are aware, this is the rst reported case of spinal anaesthesia for caesarean section in a patient with Ehlers-Danlos syndrome complicated by mitral valve prolapse.

CASE REPORT A 27-year-old woman suffered from palpitations during her rst pregnancy in 1995, which led her to seek advice from a cardiologist. A 24-h Holter monitor at that time showed frequent ventricular ectopic beats and runs of bigeminy and one brief run of ventricular tachycardia. A transthoracic echocardiogram showed mitral valve prolapse but was otherwise normal with good biventricular function. She was noted by the cardiologist to be 185 cm tall and to have an arm span of 190 cm, with evidence of joint hypermobility and to have a mild pectus deformity of her chest but no scoliosis. There was no history of ocular problems and her palate was normal. There was no family history of sudden death due to aortic rupture. The cardiologist was suspicious that she had a collagen disorder such as Marfans or Ehlers-Danlos syndrome and referred her to a geneticist. During the remainder of her pregnancy she was kept under surveillance by both obstetricians and cardiologists, but her palpitations settled and there were no further problems. No further cardiological investigations were done and she had a spontaneous vaginal delivery of a healthy infant at full term. Subsequently, she saw a geneticist who diagnosed Ehlers-Danlos syndrome. However, she declined to undergo typing. Over the course of the next 2 years she had a total of three spontaneous pneumothoraces, one of which required a chest drain and hospital admission for 2 days. She then booked at 16 weeks for her second
192

Accepted September 2000 P. Dill-Russell FRCA (Specialist Registrar), Anaesthetic Department, St Georges Hospital, Blackshaw Road, Tooting, London SW17 0QT, UK. L. St. John Jones FRCA (Consultant), Crawley Hospital, West Green Drive, Crawley, West Sussex, RH11 7DH, UK. Correspondence to: P. Dill-Russell, Tel: +44 (0) 20 8672 1255, ext: 3316/7; E-mail: pdilr@yahoo.co.uk

Ehlers-Danlos syndrome, complicated by mitral valve prolapse pregnancy and was seen by the consultant obstetrician who, after reviewing literature regarding spontaneous pneumothoraces occurring during pregnancy and discussing with the patient and a consultant anaesthetist the potential dangers of a further pneumothorax occurring in labour, advised caesarean section. During the second pregnancy the patient did not complain of palpitations and was not reviewed or investigated further by the cardiologist. Discussion between obstetrician and anaesthetist took place early in pregnancy and the patient was rst seen by the consultant anaesthetist at 28 weeks gestation. Decisions about mode of delivery and, if by caesarean section whether by regional or general anaesthesia, were complicated by the fact that her Ehlers-Danlos syndrome had not been typed. Type IV Ehlers-Danlos syndrome is associated with an increased morbidity and mortality compared with other types. After a full discussion the patient elected for planned caesarean section under spinal anaesthesia. Anaesthetic management Blood was taken for grouping at the booking clinic at 16 weeks gestation and 4 units of blood were kept cross-matched during the third trimester. It had been decided that, should she go into labour before the planned date of her caesarean section, then consultant obstetric and anaesthetic staff were to be notied and she would have an emergency section under spinal anaesthesia. The patient was premedicated with oral ranitidine 150 mg and metoclopramide 10 mg 2 h before planned surgery. She was then transferred to the operating theatre where she was given 0.3 M oral sodium citrate 30 mL. Haemoglobin, platelets and clotting screen were normal. A 14-gauge cannula was inserted and she was preloaded with 1000 mL of warmed Ringers lactate solution. Pulse oximetry, non-invasive blood pressure and ECG monitoring were instituted and the subarachnoid block was sited in the sitting position with a 24-gauge Sprotte needle, using 0.5% heavy bupivacaine 2.5 mL with fentanyl 12.5 g. She was then placed in the supine position with left lateral tilt. Loss of sensation to cold was obtained to T4 and the sacral block was complete on testing with ice. Loss of sensation to light touch was obtained to T6. Caesarean section proceeded uneventfully, with blood loss of approximately 400 mL. Good uterine contraction was obtained with a 10-unit i.v. bolus of oxytocin followed by an infusion of 40 units in 500 mL of normal saline, which was the surgeons normal practice. Oxygen was given by Hudson mask at 4 L/min until delivery. A further 1000 mL of warmed Ringers

193

lactate solution was given over the course of surgery. Ephedrine to a total dose of 15 mg was given over the intra-operative period. Following local guidelines, amoxycillin 1.2 g and gentamicin 120 mg were given before surgery as prophylaxis against infective endocarditis. Systemic arterial pressure remained within 20% of baseline values throughout surgery and the postoperative period. Postoperative analgesia was provided by i.m. morphine injections, which although considered suboptimal, was at that time the hospitals usual practice. The postoperative period was uncomplicated except for some mild breathlessness on the fourth day, but the chest X-ray was clear and the symptoms resolved spontaneously. A healthy infant was delivered with Apgar scores of 8 at 1 min and 9 at 5 min. The infant was clinically tested as a matter of routine for congenital dislocation of the hip which was found to be negative. There has been a suggestion that infants born to mothers with Ehlers-Danlos syndrome are more likely to have congenital dislocation of the hip.

DISCUSSION Currently at least nine types of Ehlers-Danlos syndrome are described. It is an inherited disorder of collagen production which results in a number of common clinical features namely hyperelasticity of skin, bruising and joint hypermobility. In addition, there are a number of features of the condition that may pose particular problems for the anaesthetist. These include spontaneous rupture of major vessels, aortic dissection, uterine rupture, spontaneous pneumothorax, mitral valve prolapse, cardiac arrhythmias and uncontrolled bleeding. The latter is due to both abnormal platelet and clotting function as well as difculty tying off friable blood vessels. In addition, pregnancy poses its own set of hazards to patients with Ehlers-Danlos syndrome, namely premature rupture of membranes, preterm delivery (23.1% in one survey), excessive bleeding and spontaneous abortion (28.9%).1 Morbidity and mortality are dependent on the type of Ehlers-Danlos syndrome. Type IV is associated with the highest complication rate, with pregnancy-related mortality as high as 25% from uterine rupture or massive bleeding from vessel rupture in the postoperative period.2 However, other types are relatively benign with premature labour as the most common complication. Reported complications are summarised in Table 1.314 It is important to note that type IV, the vascular type, contributes fewer than 10% of pregnancies. Type I represents 30%, type II 35%, type III 1030% and the others less than 10%. Therefore the high risk type IV,

194

International Journal of Obstetric Anesthesia

Table 1. Obstetric and anaesthetic complications or features in Ehlers-Danlos syndrome types IIX Subtype Type I Type II Inheritance AD AD Complications/Features Premature labour (35%) Scoliosis Intra-uterine growth retardation Atonic uterus at caesarean section Premature rupture of membranes Type III Type IV AD AD Preterm labour Post-partum haemorrhage Arterial rupture Difcult vessel suturing Uterine rupture High maternal mortality (11.5%25%) Spontaneous pneumothorax Tissue fragility Scoliosis Scoliosis Short stature No anaesthetic relevance Bowed long bones References Beighton (1970)3 Beighton (1969)26 Ploeckinger (1997)4 Kiilholma (1984)5 Sorokin (1994)1 Taylor (1981)6 Sorokin (1994)1 Ainsworth (1993)7 Kiilholma (1984)5 Atalla (1998)8 Sorokin (1994)1 Rochelson (1991)9 Lauwers (1997)10 Pepin (2000)11 Cikrit (1987)12 Pepin (2000)11 Pepin (2000)11 Rudd (1983)2 Pope (1988)13 Bruno (1997)14 Beighton (1969)26 Not reported Not reported Not reported Not reported

Type V Type VI Type VII Type VIII Type IX

X-linked rec AR AR AD X-linked rec

AD: Autosomal dominant; AR: Autosomal recessive; rec: recessive.

accounts for much of the literature regarding pregnancy and Ehlers-Danlos syndrome, in fact, represents only a small proportion of Ehlers-Danlos syndrome pregnancies.15 The diagnosis is conrmed by the demonstration that cultured broblasts synthesise abnormal procollagen molecules or by the identication of a mutation in the gene for type III procollagen (COL3A1). Typing and genetic counselling are recommended in those known to be affected. In one study of 51 patients with Ehlers-Danlos syndrome, only four (8%) had no history of a bleeding tendency or bruising.16 Nine patients (18%) had signicant haemostatic abnormalities, ranging from platelet problems to clotting factor deciencies. Importantly 24 patients (47%) had normal tests for haemostasis of whom 20 (39%) had bleeding diatheses. There was no pattern to the abnormalities according to type of Ehlers-Danlos syndrome. The high frequency of a bleeding tendency in Ehlers-Danlos syndrome patients with normal tests for haemostasis (83%) was thought to support the conventional explanation for this clinical feature, that defects in the structural integrity of skin and blood vessels lead to easy bruising, and this may lead to uncontrolled bleeding through defective primary haemostasis. Mitral valve prolapse in pregnancy Our patient suffered from mitral valve prolapse, which may have an increased incidence in Ehlers-Danlos

syndrome, although one study involving 33 patients using standard two-dimensional echocardiographic views from the parasternal and apical windows found mitral valve prolapse in only 6.1% of Ehlers-Danlos syndrome patients and 7% of controls.1719 Mitral valve prolapse is now considered a common cardiac valvular abnormality having previously been reported in up to 8% of healthy individuals, although strict echocardiographic criteria now estimate a prevalence of 2.4%.20 It is characterised by elongated chordae tendineae and redundant valve leaets, which prolapse back into the left atrium as the ventricle empties. The prolapse may or may not result in regurgitation. Palpitations occur in 5070% of patients, although there is poor correlation between palpitation and instances of dysrhythmia detected by the Holter monitor in these patients. The majority of dysrhythmias are of supraventricular origin, but the incidence of ventricular dysrhythmias is also signicant (5060%), and includes premature ventricular beats, ventricular tachycardia, and ventricular brillation.21 These ventricular dysrhythmias may explain the sudden death in people with mitral valve prolapse. Our patient also suffered from palpitations and had a recorded episode of non-sustained ventricular tachycardia. Antidysrhythmic therapy is generally recommended in patients with symptomatic dysrhythmia, patients with a history of ventricular tachycardia or brillation or supraventricular tachycardia, and patients with complex and repetitive premature ventricular contractions. Most investigators recommend

Ehlers-Danlos syndrome, complicated by mitral valve prolapse -blockers for treatment since experience with other drugs is limited. The indications for therapy, however, are controversial. Propranolol works directly as a blocker and also has a quinidine-like action. It should be used with caution during pregnancy because it has the potential to initiate premature labour through the -blocking action on the uterus. It causes an increase in uterine tone which can potentially lead to a small infarcted placenta and a low-birth-weight infant. Reported effects of propranolol on the neonate include respiratory depression, hypoglycaemia and bradycardia. However, the incidence of these neonatal side effects remains low and prospective randomised trials failed to show a signicantly higher incidence of these complications with -blockers than with placebo, suggesting that these observed effects may instead have occurred as a result of fetal distress in high-risk pregnancies.22 There is no evidence that atenolol or metoprolol produce growth retardation except one study that showed growth retardation in babies receiving atenolol in the rst trimester.23 Betablockers should therefore be avoided in the rst trimester and cardioselective 1-blockers such as atenolol are the preferred agents for the second and third trimesters, as they may interfere less with 2mediated peripheral vasodilatation and uterine relaxation.22 The severity of prolapse depends on factors that reduce ventricular volume, such as increased myocardial contractility, decreased preload, tachycardia, straining and excessive airway pressure. The smaller the ventricular volume, the greater the prolapse. The anaesthetic goals, therefore, are prevention of increased myocardial contractility and/or heart rate, maintenance of normovolaemia and avoidance of elevation in airway pressure. Regional anaesthesia is therefore, usually avoided because of the sympathetic denervation resulting in increased venous capacity and decreased peripheral resistance. This leads to a reduction of ventricular volume, which may increase the degree of prolapse. However, the safe use of epidural anaesthesia in a patient with mitral valve prolapse has been reported and given that 2.4% of parturients may be affected by the condition anyway, it is quite possible that the majority remain undiagnosed during pregnancy and undergo regional anaesthesia without any untoward haemodynamic side effects.24 Current American Heart Association guidelines stipulate that the presence of a murmur or echocardiographically demonstrated regurgitation in patients with mitral valve prolapse constitutes a moderate risk of endocarditis with invasive procedures. However, antibiotic prophylaxis is not recommended for routine caesarean section and is optional for

195

vaginal delivery, for which they recommend amoxycillin 2 g orally 1 hour before, or ampicillin 2 g i.m. or i.v. 30 min before. Pneumothorax in pregnancy Our patient had suffered three spontaneous pneumothoraces in the preceding 2 years. There are several case reports relating to parenchymal cyst formation and recurrent spontaneous pneumothoraces which suggest that this might be more common in patients with Ehlers-Danlos syndrome than in the general population. There is a risk that the raised intrathoracic pressures generated by labour could have precipitated a pneumothorax. This has been reported in patients without Ehlers-Danlos syndrome, usually in the second stage of labour.25 On the other hand general anaesthesia for planned caesarean section would generate positive airway pressures, which could result in a pneumothorax in a patient with Ehlers-Danlos syndrome. The decision for a planned caesarean section under a spinal block in our patient was largely because of concern about developing another pneumothorax, which could then result in hypoxia and compromise the fetus. If a pneumothorax had developed in labour the management is no different from that in the nonpregnant state. A small pneumothorax may resolve spontaneously but should be kept under close observation, especially during labour. The aetiology, when pneumothorax does occur in the second stage of labour, is thought to be ruptured alveoli secondary to straining against a closed glottis. Air from a ruptured alveolus may track subpleurally or along the vessels into the mediastinum and then along the vessels of the trachea into the neck, thus giving rise to mediastinal emphysema and cutaneous surgical emphysema as well as spontaneous pneumothorax, or a combination of all these. Mode of delivery The mode of delivery in patients with Ehlers-Danlos syndrome should be carefully considered. Types I, V and VI Ehlers-Danlos syndrome are associated with musculoskeletal deformities which, in combination with ligamentous relaxation in pregnancy, can lead to backache. Scoliosis and spondylolisthesis have been reported frequently in these types of Ehlers-Danlos syndrome.26 This may have a bearing on mode of delivery. Vaginal delivery has been reported to be associated with severe perineal tearing and perineal haematoma formation, particularly following the use of forceps, and there has been a recent report of

196

International Journal of Obstetric Anesthesia General anaesthesia may cause a sudden increase in arterial pressure during intubation, which we feared might result in major vessel rupture, as we did not know what type of Ehlers-Danlos syndrome our patient had. In addition, positive pressure ventilation might have provoked another pneumothorax. Injury to the cervical spine at intubation is also a theoretical risk in Ehlers-Danlos syndrome due to abnormal ligament laxity, as is bleeding in the upper airway from trauma at intubation. A single-shot spinal carries less risk of bleeding in the epidural space than do epidural or combined spinal epidural, and provides superior analgesia to an epidural. However, it could be argued that a combined spinal epidural technique would have been a better choice, given the potential for protracted surgery caused by difculty suturing friable vessels. The combined spinal epidural technique has previously been used successfully in this condition for exactly this reason. Caudal blocks have also been used, but the relative frequency of use of regional and general anaesthesia for caesarean section in these patients is unknown.30,31 This is because there is so little information about pregnancy in patients with Ehlers-Danlos syndrome other than type IV. The syndrome is rare and not surprisingly, few non-complicated cases are reported. In summary, Ehlers-Danlos syndrome comprises a group of inherited collagen disorders. Some of the features of type IV Ehlers-Danlos syndrome pose potentially life-threatening complications both during anaesthesia and surgery, and during pregnancy itself. The main complications of type IV Ehlers-Danlos syndrome are bleeding, major vessel rupture, and uterine rupture. Other types of Ehlers-Danlos syndrome do not appear to be associated with these serious problems. However, as our case demonstrates, other types may be associated with less well recognised complications such as mitral valve prolapse, cardiac arrhythmias, coagulopathy (even in the absence of abnormal clotting tests) and spontaneous pneumothorax. We would strongly recommend that all cases of Ehlers-Danlos syndrome should undergo typing before surgery of any kind and early in the antenatal period if the type is not already known. Senior obstetric, cardiological and anaesthetic staff should be involved in these potentially high-risk cases. We would endorse the recommendation that pregnancy in Ehlers-Danlos syndrome type IV should be managed in a specialised centre. We would advocate the use of invasive monitoring for planned caesarean section in type IV Ehlers-Danlos syndrome, given its potential for major blood loss, and recommend its consideration in other types of Ehlers-Danlos syndrome if

delivery through central rupture of the perineum in a patient with type II Ehlers-Danlos syndrome.2728 Proponents of planned caesarean section would argue that bleeding can be better dealt with under calm, controlled conditions in the operating theatre than can a sudden haemorrhage from a torn perineal vessel whilst in labour. Uterine rupture has been well documented as a complication of labour in type IV EhlersDanlos syndrome.2 It would seem logical that caesarean section would reduce the incidence of this complication. However, caesarean section may increase the risk of uterine rupture in subsequent pregnancies. Although a denite complication of type IV Ehlers-Danlos syndrome, uterine rupture does not appear to be reported in other types. Another consideration in deciding whether to allow vaginal delivery in Ehlers-Danlos syndrome is the perceived risk of major vessel rupture as a result of straining in labour. In a review of 81 women with biochemically conrmed Ehlers-Danlos syndrome type IV who had 183 pregnancies with 167 deliveries of live-born infants at term, 12 women died during the peripartum period or within 2 weeks after delivery (ve of uterine rupture during labour, two of vessel rupture at delivery, and ve in the postpartum period after vessel rupture).11 Finally, vaginal prolapse has been described following vaginal delivery in Ehlers-Danlos syndrome. In one series, vaginal prolapse occurred in ve out of 29 patients, and recurred in two of these following repair.27 Caesarean section has been shown to reduce the incidence of subsequent vaginal prolapse compared with vaginal delivery.29 Choice of anaesthesia There are a number of issues regarding choice of anaesthetic technique that need to be carefully considered in each patient. Again the type of Ehlers-Danlos syndrome is all important. In some types (see above and Table 1), skeletal abnormalities may make regional anaesthesia difcult or impossible. A bleeding diathesis would be a contraindication to a regional block, but there was no evidence of this in our patient. Risk of major haemorrhage is often considered to be a relative contraindication to a regional technique, as compensation for hypovolaemia is impaired by sympathetic block. Patients with non-vascular types of Ehlers-Danlos syndrome (that is all types except type IV), however, are not at risk of sudden major vessel rupture and therefore at no more risk of major haemorrhage than are normal parturients. However, they may still be at risk of continuous oozing due to the coagulation problems described, and difculty suturing friable blood vessels.

Ehlers-Danlos syndrome, complicated by mitral valve prolapse there are additional risk factors for bleeding. Appropriate awareness of the potential dangers involved, adequate preparation for these potential complications and appropriate choice of technique are essential to prevent serious morbidity and mortality in this condition. In other types of Ehlers-Danlos syndrome there is insufcient information in the literature to enable the relative risks of anaesthetic and surgical morbidity and mortality to be assessed. However, this in itself would suggest that in non type IV Ehlers-Danlos syndrome, pregnancy, surgery and anaesthesia are relatively safe.

197

REFERENCES 1. Sorokin Y, Johnson M P, Rogowski N, Richardson D A, Evans M I. Obstetric and gynecologic dysfunction in the Ehlers-Danlos Syndrome. J Reprod Med 1994; 39: 281284. 2. Rudd N L, Holbrook K A, Nimrod C, Byers P H. Pregnancy complications in type IV Ehlers-Danlos syndrome. Lancet 1983; 1: 5053. 3. Beighton P. The Ehlers-Danlos syndrome. London: Heinemann, 1970. 4. Ploeckinger B, Ulm M R, Chalubinski K. Ehlers-Danlos syndrome type II in pregnancy. Am J Perinatol 1997; 14: 99101. 5. Kiilholma P, Gronroos M, Nanto V, Paul R. Pregnancy and delivery in Ehlers-Danlos syndrome. Acta Obstet Gynecol Scand 1984; 63: 437439. 6. Taylor D J, Wilcox I, Russell J K. Ehlers-Danlos syndrome during pregnancy: A case report and review of the literature. Obstet Gynecol Surv 1981; 36: 277281. 7. Ainsworth S R, Aulicino P L. A survey of patients with Ehlers-Danlos syndrome. Clin Orthop 1993; 286: 250256. 8. Atalla A, Page I. Ehlers-Danlos syndrome type III in pregnancy. Obstet Gynecol 1988; 71: 508509. 9. Rochelson B, Caruso R, Davenport D, Kaelver A. The use of prophylactic desmopressin (DDAVP) in labor to prevent hemorrhage in a patient with Ehlers-Danlos syndrome. N Y J Med 1991; 91: 268269. 10. Lauwers G, Nevelsteen A, Daenen G, Lacroix H, Suy R, Frijns J-P. Ehlers-Danlos syndrome type IV: A heterogeneous disease. Ann Vasc Surg 1997; 11: 178182. 11. Pepin M, Schwarze U, Superti-Furga A, Byers P. Clinical and genetic features of Ehlers-Danlos syndrome type IV, the vascular type. N Engl J Med 2000; 342: 673680.

12. Cikrit D F, Miles J H, Silver D. Spontaneous arterial perforation: The Ehlers-Danlos spectre. J Vasc Surg 1987; 5: 248255. 13. Pope F M, Narcisi P, Nicholls AC, et al. Clinical presentations of Ehlers-Danlos syndrome type IV. Arch Dis Child 1988; 63: 10161025. 14. Bruno P A, Napolitano V, Votino F, Di Mauro P, Nappi C. Pregnancy and delivery in Ehlers-Danlos syndrome type V. Clin Exp Obstet Gynecol 1997; 24: 152153. 15. Morales-Rosello J, Hernandez-Yago J, Pope M. Type III Ehlers-Danlos syndrome and pregnancy. Arch Gynecol Obstet 1997; 261: 3943. 16. Anstey A, Mayne K, Winter M, Van de Pette J, Pope F M. Platelet and coagulation studies in Ehlers-Danlos syndrome. Br J Dermatol 1991; 125: 155163. 17. Jaffe A S, Geltman E M, Rodey G E, Uitto J. Mitral valve prolapse: a consistent manifestation of type IV Ehlers-Danlos syndrome. The pathognomonic role of abnormal production of type III collagen. Circulation 1981; 64: 121125. 18. Watanabe S, Ishimitsu T, Inoue K, et al. Type IV EhlersDanlos syndrome associated with mitral valve prolapse: a case report. J Cardiol 1988; 18(suppl): 97105. 19. Dolan A L, Mishra M B, Chambers J B, Grahame R. Clinical and echocardiographic survey of the Ehlers-Danlos syndrome. Br J Rheumatol 1997; 36: 459462. 20. Freed L A, Levy D, Levine R A, et al. Prevalence and clinical outcome of mitral valve prolapse. N Engl J Med 1999; 341: 17. 21. Degani S, Abinader E, Scharf M. Mitral valve prolapse and pregnancy: A review. Obstet Gynecol Surv 1989; 44: 642649. 22. Page R L. Treatment of arrhythmias during pregnancy. Am Heart J 1995; 130: 871876. 23. Lip G Y H, Beevers M, Churchill D, et al. Effect of atenolol on birth weight. Am J Cardiol 1997; 79: 14361438. 24. Alcantara L, Marx G. Cesarean section under epidural analgesia in a parturient with mitral valve prolapse. Anesth Analg 1987; 66: 902903. 25. Bending J J. Spontaneous pneumothorax in pregnancy and labour. Postgrad Med J 1982; 58: 711713. 26. Beighton P, Horan F T. Orthopaedic aspects of Ehlers-Danlos syndrome. J Bone Joint Surg (Br) 1969; 51: 444. 27. Beighton P. Obstetric aspects of the Ehlers-Danlos syndrome. Br J Obstet Gynaecol 1969; 76: 97. 28. Georgy M S, Anwar K, Oates S E, Redford D H A. Perineal delivery in Ehlers-Danlos syndrome. Br J Obstet Gynaecol 1997; 104: 505506. 29. Handa V L, Harris T A, Ostergard D R. Protecting the pelvic oor: Obstetric management to prevent incontinence and pelvic oor prolapse. Obstet Gynecol 1996; 88: 470478. 30. Brighouse D, Guard B. Anaesthesia for Caesarean section in a patient with Ehlers-Danlos syndrome type IV. Br J Anaesth 1992; 69: 517519. 31. Abouleish E. Obstetric anaesthesia and Ehlers-Danlos syndrome. Br J Anaesth 1980; 52: 12831286.