What Is Known/What Is New
What Is Known
- Gastrointestinal bleeding occurs in 20% of children with biliary atresia at a median age 17 months.
- Characteristic endoscopic features based on variceal grade, red markings, and/or gastric varices are associated with a high risk of bleeding.
- Endoscopic primary prophylaxis of bleeding is effective and well tolerated.
- What Is New
- High-risk endoscopic signs are found during follow-up in one-third of children with biliary atresia who do not display these signs initially.
- Speed of emergence of high-risk signs depends on initial age, serum bilirubin, and pattern at the first endoscopy.
- A program of endoscopic surveillance can be set up to discuss primary prophylaxis of bleeding.
Portal hypertension develops early in children with biliary atresia and carries a risk of gastrointestinal (GI) bleeding: esophageal varices are found in 75% of children before age 2, and 20% of children bleed at a median age 17 months (1). Based on the grade of the esophageal varices, presence of esophageal red wale markings, and presence of gastric varices along the cardia, a characteristic upper GI endoscopic pattern has been shown to be associated with a high risk of GI bleeding in these children (1). In some children, high-risk gastroesophageal varices are present at the first endoscopic examination. Among other children who do not initially display high-risk gastroesophageal varices, follow-up shows some progression in the severity of endoscopic signs of portal hypertension in >50% of the patients (1). The question, therefore, arises as to whether it is possible to establish a program of surveillance to detect the emergence of high-risk gastroesophageal varices in children who do not display this pattern initially. The detection of high-risk varices could be used to discuss preemptive treatment and prevent GI bleeding. In the present study, we report on such an attempt. The results show that the probability of the emergence of high-risk gastroesophageal varices can be estimated based on the child's age, serum bilirubin concentration, and endoscopic pattern at the first endoscopic examination. From this analysis, a chronological program of endoscopic follow-up could be considered.
From 1989 to 2013, ≥1 upper GI endoscopic examinations were performed in 716 children with biliary atresia as part of their regular surveillance. The first upper GI endoscopy was performed at a median age 12 months in the children who displayed a clinical or ultrasonographic sign of portal hypertension. Each endoscopy was performed by the same endoscopist (M.D.) under general anesthesia or, in infants 1 year of age or younger, with intrarectal midazolam sedation (0.2–0.3 mg/kg), and the endoscopic findings were prospectively recorded from the beginning of the study and graded as follows: number and grade of esophageal varices—grade 1 when varices were flattened by insufflation, grade 2 when they were not flattened but separated by healthy mucosa, and grade 3 when varices were not flattened by insufflation and were confluent (2); red wale markings on the mucosal wall—these were estimated as previously described (3); gastric varices along the cardia described as GOV1 by Sarin et al (4) were recorded by retroflexion; and the presence of portal hypertensive gastropathy was also recorded. Each endoscopic result was ascribed to 1 of the following 2 groups: high risk and low risk, depending on its severity level. The high-risk group comprised the following features: grade 3 esophageal varices, grade 2 esophageal varices with red wale markings and/or gastric varices along the cardia, or grade 1 esophageal varices with gastric varices along the cardia. These signs have been reported to be predictive of a high risk of GI bleeding in this population of children with biliary atresia (1,5). The low-risk group consisted of endoscopic results in which these signs were not present. Informed consent of the parents was obtained in writing before each endoscopic procedure.
A total of 225 children who displayed a low-risk endoscopic pattern at the first examination underwent at least 2 upper GI endoscopic examinations, the second one ≤6 years after the first one, and form the basis of the present article. Data of follow-up were censored at 10 years after the first endoscopic examination. The median interval between the first and the last examination was 2.4 years (range, 1.5 months to 10 years). The medical records of the 225 children were studied to identify factors potentially related to the progression to high-risk gastroesophageal varices predictive of bleeding. The following data were recorded at the time of the first endoscopy and assigned according to 3 domains: clinical (age, type of and age at Kasai operation), biological (total serum bilirubin concentration in micromoles per liter, serum albumin concentration in grams per liter, prothrombin time in percentage, and blood platelet count per cubic millimeter), and endoscopic (number and grade of esophageal varices, presence of esophageal red wale markings, of gastric varices along the cardia, and/or of portal hypertensive gastropathy).
Comparisons of percentages were performed using a χ2 test. Means were compared using a Student t test. A survival study using the occurrence of high-risk varices in the 10 years after the first endoscopic examination as an event was performed to identify factors related to the emergence of these varices and to describe the probability of their emergence in 2 groups of children ages older than 18 months and 18 months or younger at the time of the first endoscopy. The role of clinical, biological, and endoscopic variables on the emergence of high-risk varices was first assessed using univariate Cox model. For each of the 3 main domains of variables (clinical, laboratory results, and endoscopic pattern) recorded at the time of the first endoscopy, a multivariate model was carried out to identify the variables remaining significant among those that were significant in the univariate model. Then, the final model comprised the significant variables of each domain. This was first applied to the entire population of 225 children, and then to the group of 126 children who were 18 months of age or younger at the time of the first endoscopy and to the group of 99 children who were older than 18 months at the time of the first endoscopy. Last, the Cox model allowed us to estimate after how many months following the first examination the probability of emergence of high-risk varices is 10% in different patterns according to prognostic variables (2). Statistical analyses were performed using Stata software, version 12 (StataCorp, College Station, TX).
The database was compiled anonymously within the restrictive requirements of the Commission Nationale Informatique et Liberté, the organization dedicated to privacy, information technology, and civil rights in France. According to French legislation, ethics committee agreement and provision of informed consent are not required for retrospective collection of data corresponding to the present practice.
Table 1 shows the results of the clinical, biological, and endoscopic items recorded at the time of the first endoscopic examination in the 225 children with biliary atresia who underwent at least 2 upper GI endoscopic examinations and who displayed low-risk signs at this first endoscopy. When compared with the other 491 children with biliary atresia and portal hypertension not included in the study, there were no significant differences in terms of age at Kasai operation, age at the first endoscopic examination, or platelet count. Significant differences, however, were recorded for the following items: the 225 children included had lower serum bilirubin concentrations and higher serum albumin levels and prothrombin times than the children not included in the study (data not shown).
Emergence of high-risk gastroesophageal varices on an upper GI endoscopic examination was found in 76 of the 225 children in the 10 years after the first endoscopic examination at a median age 4.25 years (range, 9 months to 16 years). Using the univariate Cox model, there were no significant differences in the probability of emergence of high-risk gastroesophageal varices according to age at and type of Kasai operation, and platelet count at the time of the first endoscopic examination. In this univariate model, there was a significant relation between the probability of emergence of high-risk gastroesophageal varices and the following items recorded at the time of the first endoscopic examination: number and grade of esophageal varices, serum bilirubin concentration, serum albumin concentration, prothrombin time, and age of the child (Table 2). Table 2 shows the result of the final multivariate analysis, which revealed that 3 variables were significantly related to the occurrence of high-risk gastroesophageal varices during follow-up in this population of 225 children: combined number/grade of esophageal varices, serum bilirubin concentration, and age of the child at the time of the first endoscopic examination. As shown in Figure 1, the emergence of high-risk gastroesophageal varices was faster in children who were 12 months of age or younger, in children who displayed a total serum bilirubin concentration >100 μmol/L, or in children who displayed >1 grade 1 esophageal varices at the time of the first upper GI examination. Of the 126 children who were 18 months of age or younger at the time of the first endoscopic examination, emergence of high-risk gastroesophageal varices on an upper GI endoscopic examination was found in 43 children in the 10 years after the first endoscopic examination. The multivariate analysis (Table 3) revealed that, in this population, serum bilirubin concentration was not an independent factor predicting the emergence of high-risk varices and that age ≤12 months and the presence of >1 grade 1 varices or of grade 2 varices were significantly related to the occurrence of high-risk GI varices during follow-up. As an example, the survival model estimates that a 10% probability of emergence of high-risk GI varices occurs after 1 year in a 12-month-old or younger child when no esophageal varices or 1 grade 1 varix is present, after 7 months when 2 to 4 grade 1 varices are present, and after 4 months when grade 2 varices are present at the first endoscopy. The same probability in a 12- to 18-month-old child occurs after 30 months when no varices or 1 grade 1 varix is present, after 10 months when 2 to 4 grade 1 varices are present, and after 5 months when grade 2 varices are present at the first endoscopy. Of the 99 children who were older than 18 months of age at the time of the first endoscopic examination, emergence of high-risk gastroesophageal varices on an upper GI endoscopic examination was found in 33 children in the 10 years after the first endoscopic examination. The multivariate analysis (Table 4) revealed that, in this population, serum bilirubin concentration and the presence of >1 grade 1 varices or of grade 2 varices were significantly related to the occurrence of high-risk GI varices during follow-up. The survival model estimates that a 10% probability of emergence of high-risk GI varices occurs after 4 years in a child whose serum bilirubin is ≤100 μmol/L when no esophageal varices or 1 grade 1 varix is present, after 3 years when 2 to 4 grade 1 varices are present, and after 18 months when grade 2 varices are present at the first endoscopy. The same probability in a child whose serum bilirubin is >100 μmol/L occurs after 15 months when no varices or 1 grade 1 varix is present, after 11 months when 2 to 4 grade 1 varices are present, and after 9 months when grade 2 varices are present at the first endoscopy.
Of the 76 children who displayed high-risk gastroesophageal varices during follow-up, 24 bled from varices and 40 underwent primary prophylaxis of bleeding. Of the 149 children who did not display high-risk gastroesophageal varices, 11 bled, including 9 who bled from ectopic sites.
Portal hypertension is a major complication in children with biliary atresia, and GI bleeding carries many risks, including death, acute ischemic liver necrosis, impairment of liver function, ascites with the ensuing risks of bacterial peritonitis, hyponatremia, and hepatorenal syndrome, as well as hospitalizations and major stress for the child and his or her family (6–9). Although several noninvasive indices have been reported in adults and children to indicate the presence of esophageal varices, none allows for indicating the presence of high-risk gastroesophageal varices related to high-risk bleeding. As a result, upper GI endoscopy remains necessary for this purpose (10). We have previously reported the endoscopic features that are combined with and predictive of GI bleeding in children with biliary atresia in a wide range of situations in terms of liver function (1). We reported that preemptive endoscopic treatment is effective and well tolerated in these children regardless of their liver condition even at a young age (5). Because there are no data supporting the use of medical therapy in children with portal hypertension, the choice for primary prophylaxis in children who display high-risk gastroesophageal varices at the first endoscopic examination is between endoscopic therapy and hastening of liver transplantation depending on the child's liver condition (11–14). The role of primary prophylaxis of GI bleeding is not, however, as widely accepted in the pediatric gastroenterological community for children with portal hypertension as it is in adults with cirrhosis (11,14,15). The results reported in the present study may contribute to the debate. They indicate that in children with biliary atresia in whom the initial upper GI endoscopy did not show high-risk gastroesophageal varices and whose liver condition allows 1 additional upper GI endoscopy during the following years, high-risk gastroesophageal varices can be found in 10 years in one-third of children (76/225) at ages ranging from 9 months to 16 years. The population represented by these 225 children has, as a whole, less severe liver disease and is less likely to require early liver transplantation than other children with biliary atresia. The management of portal hypertension is, therefore, of particular significance: these children have biliary cirrhosis with a cholestasis of varying degree and many are in a state of relatively compensated cirrhosis. In addition to the risks already mentioned, an episode of GI bleeding can severely alter this balance and precipitate the need for liver transplantation that would otherwise be waived or delayed for many years (16).
The results confirm and strengthen the notion that progression of varices can be extremely rapid in children 12 months of age or younger who display signs of portal hypertension and that a 10% risk of emergence of a high-risk endoscopic pattern can be estimated after a time interval as short as 4 months in some instances. The survival analysis used in the present study has some limitations because it relies on endoscopic examinations that were not set at fixed intervals for each child. Therefore, we cannot exclude that the high-risk pattern was already present before it was found. Second, the 10% risk of displaying high-risk varices to suggest a time interval between endoscopies is arbitrary; it was chosen because it had been suggested as a guideline in a similar study on adult patients with cirrhosis (2). Third, the surveillance of gastroesophageal varices does not provide information about the risk of bleeding from ectopic sites (eg, varices of the ascending jejunal loop or rectorrhagia).
1. Duché M, Ducot B, Tournay E, et al. Prognostic value of endoscopy in children with biliary atresia
at risk for early development of varices and bleeding. Gastroenterology
2. Calès P, Desmorat H, Vinel JP, et al. Incidence of large oesophageal varices in patients with cirrhosis: application to prophylaxis of first bleeding. Gut
3. Beppu K, Inokuchi K, Koyanagi N, et al. Prediction of variceal hemorrhage by esophageal endoscopy. Gastrointest Endosc
4. Sarin SK, Lahoti D, Saxena SP, et al. Prevalence, classification and natural history of gastric varices: a long-term follow-up study in 568 portal hypertension
5. Duché M, Ducot B, Ackermann O, et al. Experience with endoscopic management of high-risk gastroesophageal varices with and without bleeding in children with biliary atresia
6. Miga D, Sokol RJ, MacKenzie T, et al. Survival after first esophageal hemorrhage in patients with biliary atresia
. J Pediatr
7. Gartner JC Jr, Jaffe R, Malatack JJ, et al. Hepatic infarction and acute liver failure in children with extrahepatic biliary atresia
and cirrhosis. J Pediatr Surg
8. Pugliese R, Fonseca EA, Porta G, et al. Ascites and serum sodium are markers of increased waiting list mortality in children with chronic liver failure. Hepatology
9. Yousef N, Habes D, Ackermann O, et al. Hepatorenal syndrome: diagnosis and effect of terlipressin therapy in 4 pediatric patients. J Pediatr Gastroenterol Nutr
10. Lemoinne S, Thabut D. Screening for esophageal varices. CLD
11. Molleston JP, Shneider BL. Preventing bleeding in infants
and children: is less more? Gastroenterology
12. Wanty C, Helleputte T, Smets F, et al. Assessment of risk of bleeding from esophageal varices during management of biliary atresia
in children. J Pediatr Gastroenterol Nutr
13. Lampela H, Kosola S, Koivusalo A, et al. Endoscopic surveillance and primary prophylaxis sclerotherapy of esophageal varices in biliary atresia
. J Pediatr Gastroenterol Nutr
14. Ling SC, Walters T, McKiernan PJ, et al. Primary prophylaxis of variceal hemorrhage in children with portal hypertension
: a framework for future research. J Pediatr Gastroenterol Nutr
15. Garcia-Tsao G, Sanyal AJ, Grace ND, et al. Prevention and management of gastroesophageal varices and variceal hemorrhage in cirrhosis. Hepatology
16. Lykavieris P, Chardot C, Sokhn M, et al. Outcome in adulthood of biliary atresia
: a study of 63 patients who survived for over 20 years with their native liver. Hepatology