Documente Academic
Documente Profesional
Documente Cultură
ABSTRACT
Copyright 2010 by Lippincott Williams & Wilkins.Unauthorized reproduction of this article is prohibited.
JPGN Volume 50, Number 2, February 2010 Noninvasive Diagnosis of Varices in Children
system for gradation of esophageal variceal size. In this system, assessed using the Hosmer and Lemeshow goodness-of-fit test. The
varix size is graded on a 4-point Likert scale: grade 1 varices are small sample size in the present study limited the number of
small and flattened by insufflation of air, grade 2 are slightly larger independent variables to be included. We selected a priori 2 sets
and do not flatten, grade 3 are larger but do not touch in the middle of variables that were consistently reproduced in the adult literature
of the lumen, and grade 4 varices are large and touch each other in (1927); albumin or INR to be the first variable (reflecting the
the middle of the lumen (11). Routine practice at our institution is to severity of liver disease) and platelet count or platelet/spleen z score
measure spleen length during abdominal USS. These measurements to be the second variable (reflecting severity of portal hyperten-
were expressed as standard deviation scores (z scores) relative to sion). To avoid negative z scores, a factor of 5 was added to all of the
previously established reference standards of spleen length for spleen length z scores. The choice of either variable in each set
different ages (12). (albumin or INR) depended on optimizing the model by maximiz-
ing the c statistic and minimizing 2 log-likelihood after fitting all
of the possible models. The b-estimates of the predictors retained in
Statistical Analysis the best model were used to guide the weights of the variables in the
clinical prediction rule, after multiplication by 10 and rounding to
First, a descriptive univariate analysis was performed on the nearest 0.5.
variables that were determined subjectively by the investigators to The best cutoff point of the resulting clinical prediction
be potentially associated with esophageal varices. Categorical rule to differentiate children with and without varices was deter-
variables (eg, existence of collaterals on Doppler ultrasound exami- mined to be the point where the second diagonal crossed the
nation) are presented as proportions and compared using x2 or receiver operating characteristic (ROC) curve. However, other
Fisher exact tests. Continuous variables (eg, platelet count, albumin cutoffs to maximize sensitivity or specificity were also explored.
concentration, international normalized ratio [INR], spleen length) An area under the ROC curve (95% CI) of more than 0.7 was
are presented as mean standard deviation (SD) or median inter- considered a fair prediction rule, 0.8 good, and more than 0.9 excel-
quartile range and compared using unpaired Student t test or lent. Sensitivity, specificity, negative and positive predictive
Wilcoxon rank-sum test as appropriate for the normal distribution. values, and likelihood ratios were calculated for different cutoff
Univariate logistic regression was used to obtain the corresponding points. Ninety-five percent CI were calculated for all of the point
odds ratio (OR) and 95% confidence interval (CI) for each pre- estimates.
dictive variable. A correction factor of 0.5 was added to cells that All of the comparisons were made using 2-sided significance
contained 0 within the data tables. levels of P < 0.05. Statistical analyses were performed using SAS
Multivariate logistic regression was then modeled to asso- version 9.1 (SAS, Cary, NC) and SPSS version 15.0 (SPSS Inc,
ciate predictors with esophageal varices. Screening of variables for Chicago, IL).
the multivariable predictive rules using univariate statistical sig-
nificance levels involves multiple comparison problems and is
known to produce unreliable models (1317). Therefore, we fol- RESULTS
lowed the strong recommendation to set possible predictors a priori Of the 87 children with liver disease or portal vein throm-
based on extensive literature review and expert opinion (16,18). bosis who underwent EGD during the study period, 51 met the
Fine-tuning of the initial limited list of possible predictive variables eligibility criteria and were included in the present study (mean age
may be aided by fitting a few models while trying to maximize the c 11 years, range 2 months17 years, 25 males) (Table 1). The
statistic of the binary model. Calibration of the prediction rule was common reason for exclusion was a previous episode of bleeding
Diagnosis
Primary sclerosing cholangitis 2 (12%) 13 (38%)
Autoimmune hepatitis 3 (18%) 4 (12%)
Overlap syndrome 3 (18%) 4 (12%)
Portal vein thrombosis 4 (24%) 1 (3%)
Congenital hepatic fibrosis 2 (12%) 1 (3%)
Others 3 (18%) 11 (32%)
Male sex 6 (35%) 19 (56%) 0.17
Mean age 11 3.5 11.9 4.2 0.44
No. <2 y old 0 (0%) 1 (3%)
No. 210 y old 6 (35%) 5 (15%)
No. >10 y old 11 (65%) 28 (82%)
Child-Pugh (A/B/C) 14/3/0 29/4/0
Mean Child-Pugh 5.65 1.06 5.36 0.68 0.27
No. with splenomegaly 13/4 2/31 <0.001
EV esophageal varices.
Others include biliary atresia (1), Caroli syndrome (1), hepatitis B (2), hepatitis C (1), Budd-Chiari syndrome (1), nonalcoholic fatty liver disease (1)
glycogen storage disease (1), focal nodular hyperplasia (1), drug-related liver disease (1) parenteral nutrition cholestasis (1), hemochromatosis (1), a-1
antitrypsin deficiency (1), and cryptogenic liver disease (1).
www.jpgn.org 189
Copyright 2010 by Lippincott Williams & Wilkins.Unauthorized reproduction of this article is prohibited.
Gana et al JPGN Volume 50, Number 2, February 2010
(n 17). Ninety-two percent of all of the data fielded were complete best fit among all of the models, with the smallest 2 log-likelihood
for the 51 eligible children. Forty patients were taking medications, and highest c statistic (Table 3). After adjusting the b-coefficient as
the most frequent was ursodeoxycholic acid (n 25), but none were described, the resulting clinical prediction rule was:
receiving b-blockers. Basic characteristics did not differ between
the patients with or without esophageal varices, including age, sex, 0:75 Platelets
Child-Pugh Score, and Pediatric End-Stage Liver Disease (PELD) 2:5 Albumin
SAZ 5
or Model for End-Stage Liver Disease (MELD) scores. There were
no patients with evidence of encephalopathy.
EGD identified esophageal varices in 17 of the 51 children where platelet count is measured in units 109/L, SAZ is the
(33%), 9 of whom had large varices (grade 2 or more than 18% of spleen length z score, and albumin is measured in grams per liter.
the total, 53% of the varices group). Eight patients had portal Smaller values are associated with a higher likelihood of varices,
hypertensive gastropathy and 3 patients with large esophageal with the best cutoff value of 130 chosen to maximize sensitivity and
varices also had gastric varices. negative predictive value as required for a screening test (Fig. 1,
Variables found to differ significantly between children with Table 4). The model was well calibrated as evident from the Hosmer
and without varices included splenomegaly on physical examin- and Lemeshow goodness-of-fit test (P 0.81). The area under the
ation, spleen length z score measured by USS, presence of collat- ROC curve of this clinical rule to predict esophageal varices was
erals on USS, platelet/spleen length z score ratio, platelet count, 0.93 (0.850.99), implying excellent discriminant ability. Other
white blood cell count, aspartate aminotransferase/platelet ratio, cutoff values maximizing specificity are also presented (Table 4).
INR, aspartate aminotransferase/alanine aminotransferase ratio,
albumin, and creatinine (Table 2). DISCUSSION
We report the derivation of a predictive model that has a high
sensitivity and negative predictive value for the identification of
Derivation of the Noninvasive children with esophageal varices. The rule could be used to identify
Prediction Rule children most likely to have varices, thereby avoiding EGD in
children who do not have varices. We considered only simple,
Multivariate logistic regression was modeled to associate commonly available, reproducible variables because we believe
predictors with esophageal varices. A model with platelets, spleen that the other previously reported noninvasive predictors of eso-
length z score, and albumin as the independent variables had the phageal varices were less reproducible in clinical practice (28) and
190 www.jpgn.org
Copyright 2010 by Lippincott Williams & Wilkins.Unauthorized reproduction of this article is prohibited.
JPGN Volume 50, Number 2, February 2010 Noninvasive Diagnosis of Varices in Children
Model 1
PLT/spleen 0.074 0.93 (0.880.98) 0.007 0.93 23.346
Alb 0.247 0.78 (0.620.98) 0.039
Model 2
PLT/spleen 0.063 0.94 (0.890.99) 0.019 0.91 28.697
INR 4.056 57.7 (0.01>100) 0.362
Model 3
PLT 0.016 0.98 (0.970.99) 0.002 0.88 25.09
Alb 0.313 0.73 (0.570.94) 0.016
Model 4
PLT 0.012 0.98 (0.970.99) 0.006 0.90 33.08
INR 4.373 79.3 (0.02>100) 0.298
Alb albumin; CI confidence interval; INR international normalized ratio; 2LL 2 log-likelihood; PLT platelets.
were subject to interobserver variability (29). We measured spleen or select patients for EGD who have a greater likelihood of
length using ultrasonography which is an easily obtainable, non- having varices.
invasive, and reproducible test (30,31). The expression of SAZ Data to support the noninvasive identification of varices in
enables appropriate comparison among children of different ages. children are sparse. In a recent study of children with portal
The platelet count/spleen diameter ratio has been previously shown hypertension, children having cirrhosis with splenomegaly were
to accurately identify varices in adults with cirrhosis and has a 14.6-fold more likely to have esophageal varices compared with
logical pathophysiological basis in children with portal hyperten- children having cirrhosis without splenomegaly. Hypoalbuminemia
sion. The increase in spleen length in patients with chronic liver increased the likelihood of varices (OR 4.17 [95%] CI 1.4312.18),
disease almost always reflects the increased portal pressure (32,33), whereas the significance of thrombocytopenia in the univariate
and thrombocytopenia may be the result of splenic pooling of analysis did not hold in the multivariable modeling (38). Fifteen
platelets because of portal hypertension, immune-mediated mech- children with portal vein cavernoma and esophageal varices were
anisms, or lower thrombopoietin synthesis (3436). evaluated in another pediatric study that analyzed ultrasound
Endoscopic screening is recommended for adults with cir- markers for esophageal varices. Abdominal USS revealed an
rhosis to identify those with varices who may benefit from pro- increased lesser omentum/aorta diameter ratio in children with
phylactic treatment, because there is evidence that prophylactic portal hypertension, compared with controls (P < 0.001) (39).
therapy with b-blockers or endoscopic variceal ligation is effective Several studies in adults with cirrhosis have demonstrated the
in reducing the incidence of variceal bleeding (10). However, EGD potential for noninvasive tests to identify esophageal varices (20
is invasive and unpleasant and will find no varices in up to 50% of 24,26,27,4050). The ratio between spleen diameter and platelet
adults with cirrhosis (10,37). There is, therefore, a strong interest in count has repeatedly shown high diagnostic accuracy in these adult
identifying a noninvasive test that can either replace screening EGD studies, ranging from 0.86 to 1 (24,25,4042). The optimal cutoff
value for this ratio has yet to be determined and depends in part
upon the severity of disease in the patient population under con-
sideration. Interestingly, patients identified by the ratio as false-
positives were more likely to have esophageal varices at follow-up
compared with the patients with true-negative results (25).
The limitations of the present study include its retrospective
design and small sample size. Screening EGD for esophageal
varices in children was not routinely performed in our center
and the study population is therefore also composed of children
undergoing endoscopy for investigations of gastrointestinal symp-
toms. For this reason, younger children with biliary atresia are
poorly represented, and there is an excess of older children with
primary sclerosing cholangitis who underwent endoscopy as a part
of their assessment for inflammatory bowel disease. Blinding of
endoscopists to the presence and degree of splenomegaly and
thrombocytopenia cannot be guaranteed in retrospect. The inclusion
of children with portal vein thrombosis as well as intrahepatic
disease may make the clinical prediction rule more widely
applicable to different patient groups, although validation studies
will be required in each group. The number of patients in the present
study did not allow for reliable subgroup analysis to compare
children with prehepatic and hepatic portal hypertension. None-
FIGURE 1. The prediction rule in patients with and without theless, EGD was performed in children with various severities of
varices. liver disease in the same center and using a single classification
www.jpgn.org 191
Copyright 2010 by Lippincott Williams & Wilkins.Unauthorized reproduction of this article is prohibited.
Gana et al JPGN Volume 50, Number 2, February 2010
TABLE 4. Performance characteristics of the clinical prediction rule at different cutoff values 0:75 Platelets 2:5 Albumin
SAZ 5
Cutoff value Sensitivity (%) Specificity (%) PPV NPV LR LR AUROC P
system. We focused on the presence of esophageal varices of any 9. Shneider BL. Approaches to the management of pediatric portal hyper-
size, rather than on the presence of large esophageal varices, tension: results of an informal survey. In: Groszmann RJ, Bosch J, eds.
because this is the first step in the diagnostic workup at which Portal Hypertension in the 21st Century. Dordrecht, the Netherlands:
point decisions regarding follow-up and treatment are made. Kluwer Academic Publishers; 2004, pp 16772.
Although our clinical prediction rule was the most accu- 10. Garcia-Tsao G, Sanyal AJ, Grace ND, et al. Prevention and management
rate test of those we studied, other simple tests such as spleno- of gastroesophageal varices and variceal hemorrhage in cirrhosis.
Hepatology 2007;46:92238.
megaly and platelet count also showed a high degree of diagnostic
11. Paquet KJ. Prophylactic endoscopic sclerosing treatment of the eso-
accuracy. A previous study suggested that splenomegaly detected phageal wall in varices a prospective controlled randomized trial.
on physical examination has a high sensitivity and specificity Endoscopy 1982;14:45.
for the diagnosis of esophageal varices (38). These simple tests 12. Megremis SD, Vlachonikolis IG, Tsilimigaki AM. Spleen length in
would be more practical to use on a daily basis. Nevertheless, the childhood with US: normal values based on age, sex, and somatometric
clinical prediction rule standardizes subjective assessments of parameters. Radiology 2004;231:12934.
physicians and we therefore speculate that the reliability of our 13. Harrell FE Jr, Lee KL, Califf RM, et al. Regression modelling strategies
clinical prediction rule will prove to be higher when measured for improved prognostic prediction. Stat Med 1984;3:14352.
in validation studies. The clinical prediction rule uses variables 14. Harrell FE Jr, Lee KL, Matchar DB, et al. Regression models for
that are commonly measured in routine clinical care of children prognostic prediction: advantages, problems, and suggested solutions.
with liver disease or portal vein thrombosis, thus reducing the Cancer Treat Rep 1985;69:10717.
potential cost of the present approach to the identification of 15. Harrell F. Regression coefficients and scoring rules. J Clin Epidemiol
varices. 1996;49:819.
16. Harrell FE Jr, Lee KL, Mark DB. Multivariable prognostic models:
issues in developing models, evaluating assumptions and adequacy, and
CONCLUSIONS measuring and reducing errors. Stat Med 1996;15:36187.
17. Austin PC, Tu JV. Automated variable selection methods for logistic
In summary, we found that variables linked to portal hyper- regression produced unstable models for predicting acute myocardial
tension (platelet count, spleen length) as well as those linked to the infarction mortality. J Clin Epidemiol 2004;57:113846.
severity of liver dysfunction (albumin, INR) are associated with the 18. Laupacis A, Sekar N, Stiell IG. Clinical prediction rules. A review and
presence of esophageal varices. We have derived a clinical pre- suggested modifications of methodological standards. JAMA 1997;
diction rule that uses simple noninvasive tests and shows accuracy 277:48894.
for the identification of children who require EGD to characterize 19. Sanyal AJ, Fontana RJ, Di Bisceglie AM, et al. The prevalence and risk
their esophageal varices. We are now undertaking a prospective factors associated with esophageal varices in subjects with hepatitis C
study to validate these encouraging results in an additional cohort and advanced fibrosis. Gastrointest Endosc 2006;64:85564.
20. Zaman A, Hapke R, Flora K, et al. Factors predicting the presence of
of children.
esophageal or gastric varices in patients with advanced liver disease. Am
J Gastroenterol 1999;94:32926.
REFERENCES 21. Ng FH, Wong SY, Loo CK, et al. Prediction of oesophagogastric varices
1. Lykavieris P, Gauthier F, Hadchouel P, et al. Risk of gastrointestinal in patients with liver cirrhosis. J Gastroenterol Hepatol 1999;14:785
bleeding during adolescence and early adulthood in children with portal 90.
vein obstruction. J Pediatr 2000;136:8058. 22. Pilette C, Oberti F, Aube C, et al. Non-invasive diagnosis of esophageal
2. Goncalves ME, Cardoso SR, Maksoud JG. Prophylactic sclerotherapy varices in chronic liver diseases. J Hepatol 1999;31:86773.
in children with esophageal varices: long-term results of a controlled 23. Madhotra R, Mulcahy HE, Willner I, et al. Prediction of esophageal
prospective randomized trial. J Pediatr Surg 2000;35:4015. varices in patients with cirrhosis. J Clin Gastroenterol 2002;34:815.
3. Miga D, Sokol RJ, Mackenzie T, et al. Survival after first esophageal 24. Giannini E, Botta F, Borro P, et al. Platelet count/spleen diameter ratio:
variceal hemorrhage in patients with biliary atresia. J Pediatr 2001; proposal and validation of a non-invasive parameter to predict the
139:2916. presence of oesophageal varices in patients with liver cirrhosis. Gut
4. van Heurn LW, Saing H, Tam PK. Portoenterostomy for biliary atresia: 2003;52:12005.
Long-term survival and prognosis after esophageal variceal bleeding. 25. Giannini EG, Botta F, Borro P, et al. Application of the platelet count/
J Pediatr Surg 2004;39:69. spleen diameter ratio to rule out the presence of oesophageal varices in
5. Kobayashi A, Itabashi F, Ohbe Y. Long-term prognosis in biliary atresia patients with cirrhosis: a validation study based on follow-up. Dig Liver
after hepatic portoenterostomy: analysis of 35 patients who survived Dis 2005;37:77985.
beyond 5 years of age. J Pediatr 1984;105:2436. 26. Chalasani N, Imperiale TF, Ismail A, et al. Predictors of large esophageal
6. Toyosaka A, Okamoto E, Okasora T, et al. Outcome of 21 patients with varices in patients with cirrhosis. Am J Gastroenterol 1999;94:328591.
biliary atresia living more than 10 years. J Pediatr Surg 1993;28:1498 27. Zein CO, Lindor KD, Angulo P. Prevalence and predictors of esophageal
501. varices in patients with primary sclerosing cholangitis. Hepatology
7. Mitra SK, Kumar V, Datta DV, et al. Extrahepatic portal hypertension: 2004;39:20410.
A review of 70 cases. J Pediatr Surg 1978;13:517. 28. Valletta EA, Loreti S, Cipolli M, et al. Portal hypertension and eso-
8. Webb LJ, Sherlock S. The aetiology, presentation and natural history of phageal varices in cystic fibrosis. Unreliability of echo-Doppler flow-
extra-hepatic portal venous obstruction. Q J Med 1979;48:62739. metry. Scand J Gastroenterol 1993;28:10426.
192 www.jpgn.org
Copyright 2010 by Lippincott Williams & Wilkins.Unauthorized reproduction of this article is prohibited.
JPGN Volume 50, Number 2, February 2010 Noninvasive Diagnosis of Varices in Children
29. Tamayo SG, Rickman LS, Mathews WC, et al. Examiner dependence 40. Giannini EG, Zaman A, Kreil A, et al. Platelet count/spleen diameter
on physical diagnostic tests for the detection of splenomegaly: a ratio for the noninvasive diagnosis of esophageal varices: results of
prospective study with multiple observers. J Gen Intern Med 1993;8: a multicenter, prospective, validation study. Am J Gastroenterol 2006;
6975. 101:25119.
30. Winkfield B, Aube C, Burtin P, et al. Inter-observer and intra-observer 41. Sethar GH, Ahmed R, Rathi SK, et al. Platelet count/splenic size ratio:
variability in hepatology. Eur J Gastroenterol Hepatol 2003;15:959 a parameter to predict the presence of esophageal varices in cirrhotics.
66. J Coll Physicians Surg Pak 2006;16:1836.
31. ODonohue J, Ng C, Catnach S, et al. Diagnostic value of Doppler 42. Zimbwa TA, Blanshard C, Subramaniam A. Platelet count/spleen
assessment of the hepatic and portal vessels and ultrasound of the diameter ratio as a predictor of oesophageal varices in alcoholic
spleen in liver disease. Eur J Gastroenterol Hepatol 2004;16:147 cirrhosis. Gut 2004;53:1055.
55. 43. Schepis F, Camma C, Niceforo D, et al. Which patients with cirrhosis
32. Liangpunsakul S, Ulmer BJ, Chalasani N. Predictors and implications of should undergo endoscopic screening for esophageal varices detection?
severe hypersplenism in patients with cirrhosis. Am J Med Sci 2003;326: Hepatology 2001;33:3338.
1116. 44. Thomopoulos KC, Labropoulou-Karatza C, Mimidis KP, et al. Non-
33. Pockros PJ, Duchini A, McMillan R, et al. Immune thrombocytopenic invasive predictors of the presence of large oesophageal varices in
purpura in patients with chronic hepatitis C virus infection. Am J patients with cirrhosis. Dig Liver Dis 2003;35:4738.
Gastroenterol 2002;97:20405. 45. Thabut D, Trabut JB, Massard J, et al. Non-invasive diagnosis of large
34. Peck-Radosavljevic M. Thrombocytopenia in liver disease. Can J oesophageal varices with FibroTest in patients with cirrhosis: a pre-
Gastroenterol 2000;14 (Suppl D):60D6D. liminary retrospective study. Liver Int 2006;26:2718.
35. Peck-Radosavljevic M. Review article: coagulation disorders in 46. Kazemi F, Kettaneh A, Nkontchou G, et al. Liver stiffness measurement
chronic liver disease. Aliment Pharmacol Ther 2007;26 (Suppl 1): selects patients with cirrhosis at risk of bearing large oesophageal
218. varices. J Hepatol 2006;45:2305.
36. Giannini E, Botta F, Borro P, et al. Relationship between thrombopoietin 47. Cottone M, DAmico G, Maringhini A, et al. Predictive value of
serum levels and liver function in patients with chronic liver disease ultrasonography in the screening of non-ascitic cirrhotic patients with
related to hepatitis C virus infection. Am J Gastroenterol 2003;98: large varices. J Ultrasound Med 1986;5:18992.
251620. 48. Zaman A, Becker T, Lapidus J, et al. Risk factors for the presence of
37. de Franchis R, Primignani M. Natural history of portal hypertension in varices in cirrhotic patients without a history of variceal hemorrhage.
patients with cirrhosis. Clin Liver Dis 2001;5:64563. Arch Intern Med 2001;161:256470.
38. Fagundes ED, Ferreira AR, Roquete ML, et al. Clinical and laboratory 49. Vanbiervliet G, Barjoan-Marine E, Anty R, et al. Serum fibrosis markers
predictors of esophageal varices in children and adolescents with portal can detect large oesophageal varices with a high accuracy. Eur J
hypertension syndrome. J Pediatr Gastroenterol Nutr 2008;46:178 Gastroenterol Hepatol 2005;17:3338.
83. 50. Sethar GH, Ahmed R, Rathi SK, et al. Platelet count/splenic size ratio:
39. Moreno E, Torres P, Trejo C, et al. Predictive value of ultrasonography in a parameter to predict the presence of esophageal varices in cirrhotics.
portal hypertension. G E N 1991;45:1014. J Coll Physicians Surg Pak 2006;16:1836.
www.jpgn.org 193
Copyright 2010 by Lippincott Williams & Wilkins.Unauthorized reproduction of this article is prohibited.