Efficacy of Postoperative High-dose Steroids in Infants with Biliary Atresia: A Meta Analysis
Purpose: To evaluate whether high vs low-dose steroids or non-steroid improve the clearance rates of jaundice, cholangitis, and survival rates with the native liver in biliary atresia (BA) after Kasai portoenterostomy. Methods: Potential prospective and retrospective controlled trials were extracted from the computer database. Odds ratios (OR) with 95% confidence intervals (CI) were calculated for the clearance rates of jaundice, cholangitis and survival rates with native liver. Findings: The pooled OR significantly favour high-dose steroids (the cumulative dosage of prednisolone over 80 mg/kg) over low-dose steroids (the cumulative dosage of prednisolone less than 43 mg/kg) (1.804; 95%CI: 1.286, 2.531; P=0.001) or non-steroid (2.236; 95% CI: 1.122, 4.458; P=0.022) in the clearance rates of jaundice. Conclusion: High-dose steroids in postoperative BA improve the clearance rates of jaundice compared with low-dose steroids and non-steroid at a follow up of 6 months.
Keyword : Cholangitis; Jaundice; Kasai portoenterostomy; Steroids; Survival rates
The outcome of biliary atresia (BA) has improved since Kasai and Suzuki reported the hepatic portoenterostomy to facilitate bile flow, clear jaundice, and normalise bilirubin levels in infants with BA in 1959.1 However, the long-term problems in infants with BA after Kasai portoenterostomy, such as cholangitis, biliary cirrhosis and abnormal liver function became apparent.2-4 Furthermore, The need for liver transplanation is necessitated because of postoperative BA progress.5 As the availability of donor livers is limited, optimising the success Kasai portoenterostomy is very important to improve the survival rates in the infants with BA.6
Various reports suggest that steroids after Kasai portoenterostomy have improved the outcome of BA, such as reduction of bilirubin levels, the improvement of liver function and increased survival rates with native liver.7-11 However, the results from these studies have been inconclusive, some reports showed that steroids after Kasai portoenterostomy did not significantly improve liver function and cholangitis, reduce the need for liver transplanation.12-15 Further, the results from a meta analysis published in 2014 and another systemic review and meta-analysis published in 2011 did not find significant differences in normalising serum bilirubin levels, improving cholangitis rates and delaying the need for early liver transplantation in BA infants with postoperative steroids compared those with non-steroid.6,16 A meta-analysis published in 2016 showed that adjuvant steroid treatment following Kasai portoenterostomy may improve short-term (≤1 year) clearance rates of jaundice, but no significant effects on long-term (≥2 years) clearance rates of jaundice and native liver survival rates.17 These meta-analysis, however, is no consensus on what constitutes "high" dosage nor "low" dosage regimens and what kind type of steroids.
To evaluate the efficacy of high-dose steroids after Kasai portoenterostomy for BA, we conducted an updated meta-analysis to compare the outcomes of infants with BA after Kasai portoenterostomy who had been treated with high-dose or low-dose steroids or non-steroid according to the exact dosage of the steroids and used the "prednisolone" as the only standard of the steroids.
Articles in four databases-Pubmed/Mediline, Embase data, web of Science and Cochrane Library - were searched systematically by two independent reviewers. The research was limited to human studies, and included both observational studies and randomised controlled trials (RCTs). MeSH terms and text words including: biliary atresia, Kasai portoenterostomy, steroids, prednisolone, glucocorticoid, dexamethasone, methylprednisolone, corticosteroids and hydrocortisone, were searched. Relavant articles were retrieved based on consensus among authors. The reference lists of the retrieved articles were searched for other potentially relevant articles. The literature search was performed on May 1st, 2019.
The following inclusion criteria were used to identify published studies for this meta-analysis: 1) definition of high-dose and low-dose steroids: the cumulative dosage of prednisolone over 80 mg/kg is defined as high-dose steroids and the cumulative dosage less than 50 mg/kg is defined as low-dose steroids. 2) study design: comparative studies, patients who received high-dose or low-dose steroids or non-steroid therapy; 3) population and intervention: infants who were diagnosed as BA and underwent Kasai portoenterostomy and not received liver transplantation; 4) outcome measurement: variables of reports containing the primary outcome of interest were the clearance rates of postoperative bilirubin. The secondary outcomes were the cholangitis and survival rates with native liver after Kasai portoenterostomy. Decisions regarding which trials to be included were made independently by two reviewers.
Studies were excluded including patients who only received an unrelated outcome measurement after Kasai portoenterostomy; double reporting, delivering no baseline data and no aggregated results.
Information/data extracted included first author and year of publication, study design, treatment groups, number of patients, age and sex distribution of patients, and postoperative biliary drainage, cholangitis, and survival rates.
The statistical analysis was carried out through the statistical software Stata 12.0 (Texas, USA). Forest plots were used to present the results. Odds ratios (OR) with 95% confidence intervals (CI) were calculated for the clearance rates of bilirubin and survival rates with native liver and were compared between patients who received high-dose steroids treatment (high-dose steroids group) and low-dose steroids (low-dose steroids group) or those who had not received steroids treatment (non-steroid group). Heterogeneity among the studies was assessed by calculating the Cochran Q and the I statistic. For the Q statistic, P<0.10 was considered to indicate statistically significant heterogeneity. The I2 statistic indicates the percentage of the observed between-study variability caused by heterogeneity. Heterogeneity determined using the I2 statistic was defined as follows: 0-24%=no heterogeneity; 25%-49%= moderate heterogeneity; 50%-74%=large heterogeneity; and 75%-100% =extreme heterogeneity. If either the Q statistic (P<0.1) or I2 statistic (>50%) indicated heterogeneity existed between studies, a random-effects model of analysis was used. Otherwise, a fixed-effects model of analysis was used. Sensitivity analysis was performed for the clearance rates of jaundice, cholangitis and survival rates with native liver based on the leave-one-out approach. Funnel plot was used to observe the included studies' publication bias, asymmetry degree was measured by Egger's test, and a P value<0.05 was considered as an evidence of publication bias. To investigate the steadiness of our results, sensitivity analysis investigating the influence of each individual study on the overall meta-analysis summary estimates was carried out to identify potential outliners.
Initial electronic search retrieved 430 articles after removing duplicates, information insufficient, non-steroid therapy, reviews and meta-analysis, no Kasai portoenterostomy in which 409 articles were excluded after first-pass review of titles and abstracts. Nine studies were eliminated after full text review due to the lack of high-dose steroids therapy.7-14 One study was excluded from the meta-analysis because there was a short follow-up time (within 2 months after Kasai portoenterostomy) in each group.15 Two studies were excluded because there were only abstracts.18,19 Finally, a total of nine articles were identified (Figure 1).12,20-27
The characteristics of the studies are summarised in Table 1. Among the nine studies included, there were three randomised controlled trials, and six retrospective studies. The total number of patients in the studies ranged from 12 to 253 (overall n=948). Six studies reported on other adjuvant regimens, such as antibiotics, ursodeoxycholic acid.
The clearance rates of jaundice after Kasai portoenterostomy was defined by direct bilirubin <20 μmol/L (Dong et al) and serum bilirubin ≤20 μmol/L (Petersen et al and Wang et al), serum total bilirubin <25 μmol/L (Bezerra et al) and serum total bilirubin <34 μmol/L (Vejchapipat et al and Kobayashi et al).20-25 The studies reported by Davenport et al and Meyers et al did not provide a definition of jaundice clearance.26,27 Wang et al and Vejchapipat et al defined postoperative cholangitis as a combination of fever >38.5°C, change of yellow stool to alcholic stool, and leukocytosis with polymorphonuclear leukocyte predominance.22,24 The study reported by Kobayashi et al was defined postoperative cholangitis as fever (>38°C) accompanied by elevated serum bilirubin and leukocytosis.25
Clearance Rates of Jaundice
Two group studies were included in the meta-analysis of clearance rates of jaundice in infants with BA after Kasai portoenterostomy at a follow up of 6 months. In the first group (high-dose steroids vs non-steroid), after pooling of data, there was a significant heterogeneity among the studies (Q=13.270, df=6, P=0.039; I2=54.8%); therefore, a random-effects model of analysis was used. The pooled OR (2.236, 95% CI=1.122 to 4.458, P=0.022, Figure 2) significantly favours high dose steroids regimens over non-steroid regimens. In the second group (high-dose vs low-dose steroids), after pooling of data, there was no significant heterogeneity among the studies (Q=4.150, df=5, P=0.528; I2=0.0%); therefore, a random-effects model of analysis was used. The pooled OR (1.804, 95% CI=1.286 to 2.531, P=0.001, Figure 2) significantly favours high-dose steroid regimens over low-dose steroids. Overall, the pooled OR (1.857, 95% CI=1.297 to 2.658, P=0.132, Figure 2) significantly favour high-dose steroid regimens over low-dose steroids or non-steroid (P=0.001, Figure 2).
Four studies were included in the meta-analysis of cholangitis rates at a follow up of 6 months (high-dose steroids vs non-steroid). After pooling of data, there was no significant heterogeneity among the studies (Q=0.090, df=3, P=0.993; I2=0.0%). The pooled OR (0.502, 95% CI=0.236 to 1.064, P=0.072, Figure 3) did not significantly favour high-dose steroid regimens over non-steroid.
Survival Rates with Native Liver
Five studies were included in the meta-analysis of survival rates with native liver at a follow up of 12 months (high-dose steroids vs non-steroid). After pooling of data, there was no significant heterogeneity among the studies (Q=7.950, df=4, P=0.093; I2=49.7%); The pooled OR (1.543, 95% CI=0.733 to 3.247, P=0.253, Figure 4) did not significantly favour high-dose steroids over low-dose steroids and non-steroid.
Figure 5 showed that the results of the sensitivity analysis, in which the studies were omitted one-by-one. For the clearance rates of jaundice (RR=1.35, 95%CI: 1.19-1.53, Figure 5A) and the cholangitis rates (RR=0.66, 95%CI: 0.42-1.04, Figure 5B), Pooled estimates for all groups were insensitive to the removal of individual studies and the corresponding pooled ORs were not substantially altered that indicating that our results were stable and reliable. For the survival rates with native liver, the direction and magnitude of the pooled estimated did not vary markedly with the removal of any study in a statistically significant finding (RR=1.13, 95%CI: 0.98-1.30, Figure 5C).
Egger's publication bias plot was performed to assess the publication bias of the literature. The symmetrical Egger's plots provided no evidence for publication bias in the total studies for the clearance rates of jaundice (t=0.400, P=0.695) (Figure 6), for cholangitis rates (t=0.200, P=0.863) and for survival rates with native liver (t=1.820, P=0.167).
Postoperative steroids as an important adjunct regimens were used to treat infants with BA after Kasai portoenterostomy, high-dose steroid protocols showed a beneficial efficacy of reducing serum bilirubin, improving liver function and cholangitis, and increasing survival rates with native liver.20,25-27 Some studies, however, reported that high-dose steroids did not improve the outcomes of BA infants after portoenterostomy.21-22 There is no a systemic review and meta-analysis to support the efficacy of high-dose steroids. Additionally, there still existed the lack of consistency between these studies in the types of steroids given, the duration of treatment, or the dosage of high and low-dose steroids. Our present meta-analysis includes nine studies and 948 infants with BA. For the first time, we used the "prednisolone" as the only standard of the steroids and further defined the cumulative dosage of prednisolone over 80 mg/kg throughout the total treatment courses as the standard of high-dose steroids and the cumulative dosage of prednisolone less than 43 mg/kg throughout the total treatment courses as the standard of low-dose steroids in our meta-analysis. We examined the effects of high-dose steroids on the important indicators after portoenterostomy compared with low-dose steroids or non-steroid so that we can support the sufficient evidences for clinical treatment.
The results of our meta-analysis suggest that high-dose steroids (the cumulative dosage of prednisolone over 80 mg/kg) after portoenterostomy provide a beneficial effect on the clearance rates of jaundice at a follow up of 6 months compared with low-dose steroid (the cumulative dosage of prednisolone less than 43 mg/kg) or non-steroid group. Overall, high-dose steroids after portoenterostomy showed a beneficial efficacy of improving the clearance rates of jaundice at a follow up of 6 months compared with low-dose steroids and non-steroid. Subsequent sensitivity analysis reveals that the meta-analysis of clearance rates of jaundice has robustness/reliability. Our data, however, show the efficacy of high-dose steroids after Kasai portoenterostomy in the aspect of reducing serum bilirubin, which is not consistent with those from a previously published meta-analysis and systematic review that did not find a significant effect of adjunct postoperative steroid therapy on normalising serum bilirubin levels.6,16 Four studies reported the effects of high-dose steroids vs non-steroid regimens on improving the cholangitis rates at a follow up of 6 months. Our meta-analysis reveals that high-dose steroids have no significant beneficial effects on the cholangitis rates at a follow up of 6 months compared with non-steroid groups (Figure 3). There are another three papers reported the effects of high-dose steroids vs low-steroids on improving the cholangitis that we can't perform a meta-analysis due to the lack of sufficient data. Total seven studies including five papers (high-dose steroids vs non-steroid) and two papers (high-dose steroids vs low-dose steroids) reported the survival rates with native liver at a follow up of 12 months. For the same reason, we performed one meta-analysis to compare between high-dose steroids with non-steroid. Our meta-analysis reveals that high-dose steroids have no significant beneficial effects on survival rates with native liver at a follow up of 12 months compared with non-steroid (Figure 4). Comparison between high-dose and low-dose steroids, Wang et al reported that high-dose steroids did not improve the survival rates with native liver at a follow up of 12 months.22 Additionally, one study by Davenport et al reported that there was no difference in 1-year and 4-year native liver survival rates between high-dose steroids (86% and 56%) and low-dose steroids (72% and 50%).26 Another study reported by Dong et al showed that two-year survival rate with native liver in the high-dose steroid group was much higher than that in the low-dose steroid group (53% vs 38%, P<0.01).20 As such, we suggest that more studies are needed to fully evaluate the survival rates with native liver between high vs low-dose steroids.
Due to a limitation on number of literature, we should consider all factors interfering with our results. Firstly, Our meta-analysis includes nine studies and 948 infants with BA. More clinical trials with a large sample size are needed because insufficient number of samples possibly affect the objectivity and authenticity of the results. Secondly, different study design affects the reliability of the results. Although our publication bias assessment of all studies did not provide the evidences for publication bias in the publications, total nine studies including six studies in our meta-analysis were retrospective and therefore may have varying degrees of bias inherent with the design. Only three randomised controlled trials further elucidate the efficacy of high vs low-dose steroids or non-steroid after Kasai portoenterostomy. Therefore, more randomised, double-blind, multicentre trials are needed. Thirdly, although we used the "prednisolone" as the an unified standard of the steroids in this study, there still exist differences in half-life and bioavailability among different types of the steroids. Therefore this study may have varying degrees of bias inherent. Fourthly, there was obvious heterogeneity between the studies, for example, the effects of high-dose steroids vs non-steroid on native liver survival rates. Further, in several studies, high-dose steroid regimens were not standardised resulting in within-study heterogeneity, which affected the outcomes of our meta-analysis. Therefore, well-designed prospective studies, especially randomised controlled trials, are needed to explain the important efficacy of high-dose steroids after Kasai portoenterostomy. Finally, not every research result in all the studies was the outcome of interest. As such, the results of our meta-analysis were partially limited by the availability of data.
Our results in this meta-analysis suggest that high-dose steroids (the cumulative dosage of prednisolone over 80 mg/kg) after Kasai portoenterostomy for BA may significantly improve the clearance rates of jaundice at a follow up of 6 months compared with low-dose steroid and non-steroid regimens and can not improve the cholangitis at a follow up of 6 months and the survival rates with native liver at a follow up of 12 months compared with non-steroid regimens. Given that adjunct steroids will undoubtedly continue to be routinely prescribed, we recommend to use high-dose steroid regimens (the cumulative dosage of prednisolone over 80 mg/kg) after Kasai portoenterostomy and further conduct more well-designed prospective clinical controlled trials with a larger sample size to provide more definite evidences to support the application of high-dose steroid protocols.
The study was supported by grants from Nanjing Scientific Development Project (#201605045).
Conflict of Interest
The authors declare that there is no actual or potential conflict of interest that could inappropriately influence this work.
1. Kasai M, Suzuki SA. New operation for "non-correctable" biliary atresia: hepatic portoenterostomy. Shujutsu 1959;13:733-9. [In Japanese].
2. Tyraskis A, Parsons C, Davenport M. Glucocorticosteroids for infants with biliary atresia following Kasai portoenterostomy. Cochrane Database Syst Rev 2018;5:CD008735.
3. Sasaki H, Tanaka H, Nio M. Surgical and long-term problems in biliary atresia. Nihon Geka Gakkai Zasshi 2014;115:317-22.
4. Chan KW, Lee KH, Wong HY, et al. From laparoscopic to open Kasai portoenterostomy: the outcome after reintroduction of open Kasai portoenterostomy in infant with biliary atresia. Pediatr Surg Int 2014;30:605-8.
5. Lee S, Park H, Moon SB, et al. Long-term results of biliary atresia in the era of liver transplantation. Pediatr Surg Int 2013;29:1297-301.
6. Zhang D, Yang HY, Jia J, Zhao G, Yue M, Wang JX. Postoperative steroids after Kasai portoenterostomy for biliary atresia: a meta-analysis. Int J Surg 2014;12:1203-9.
7. Suzuki T, Hashimoto T, Kondo S, Sato Y, Hussein MH. Evaluating patients' outcome post-Kasai operation: a 19-year experience with modification of the hepatic portoenterostomy and applying a novel steroid therapy regimen. Pediatr Surg Int 2010;26:825-30.
8. Davenport M, Parsons C, Tizzard S, Hadzic N. Steroids in biliary atresia: single surgeon, single centre, prospective study. J Hepatol 2013;59:1054-8.
9. Lao OB, Larison C, Garrison M, Healey PJ, Goldin AB. Steroid use after the Kasai procedure for biliary atresia. Am J Surg 2010;199:680-4.
10. Stringer MD, Davison SM, Rajwal SR, McClean P. Kasai portoenterostomy: 12-year experience with a novel adjuvant therapy regimen. J Pediatr Surg 2007;42:1324-8.
11. Muraji T, Higashimoto Y. The improved outlook for biliary atresia with corticosteroid therapy. J Pediatr Surg 1997;32:1103-7.
12. Chung HY, Wong KKY, Lan LCL, Tam PKH. Evaluation of a standardized protocol in the use of steroids after Kasai operation. Pediatr Surg Int 2008;24:1001-4.
13. Escobar MA, Jay CL, Brooks RM, et al. Effect of corticosteroid therapy on outcomes in biliary atresia after Kasai portoenterostomy. J Pediatr Surg 2006;41:99-103.
14. Muraji T, Nio M, Ohhama Y, et al; Japanese Biliary Atresia Society. Postoperative corticosteroid therapy for bile drainage in biliary atresia-a nationwide survey. J Pediatr Surg 2004;39:1803-5.
15. Japanese Biliary Atresia Society, Nio M, Muraji T. Multicenter randomized trial of postoperative corticosteroid therapy for biliary atresia. Pediatr Surg Int 2013;29:1091-5.
16. Sarkhy A, Schreiber RA, Milner RA, Barker CC. Does adjuvant steroid therapy post-Kasai portoenterostomy improve outcome of biliary atresia? Systematic review and meta-analysis. Can J Gastroenterol 2011;25:440-4.
17. Zhang MZ, Xun PC, He K, Cai W. Adjuvant steroid treatment following Kasai portoenterostomy and clinical outcomes of biliary atresia patients: an updated meta-analysis. World J Pediatr 2017;13:20-6.
18. Alterti D, Colusso M, Cheli M, Stroppa P, Bravi M, Casotti V,Sonzogni A, Antiga LD. Efficacy of high versus low dose adjuvant corticosteroid treatment in children with biliary atresia a single-centre, controlled study. J Pediatr Gastroenterol Nutr 2011;52(suppl 1):E54.
19. Colusso M, Zaranko E, Terzi C, et al. Three year follow-up after Kasai portoenterostomy for biliary atresia: outcomes and role of high dose steroid therapy. Dig Liver Dis 2009;41:S200-1.
20. Dong R, Song Z, Chen G, Zheng S, Xiao XM. Improved outcome of biliary atresia with postoperative high-dose steroid. Gastroenterol Res Pract 2013;902431.
21. Petersen C, Harder D, Melter M, et al. Postoperative high-dose steroids do not improve mid-term survival with native liver in biliary atresia. Am J Gastroenterol 2008;103:712-9.
22. Wang W, Zheng S, Xiao XM. Effect and safety of high-dose steroids in postoperative treatment of biliary atresia. Chin J Pediatr Surg 2006;27:460-3. [In Chinese].
23. Bezerra JA, Spino C, Magee JC, et al; Childhood Liver Disease Research and Education Network (ChiLDREN). Childhood Liver Disease Research and Education Network (ChiLDREN). Use of corticosteroids after hepatoportoenterostomy for bile drainage in infants with biliary atresia: the START randomized clinical trial. JAMA 2014;311:1750-9.
24. Vejchapipat P, Passakonnirin R, Sookpotarom P, Chittmittrapap S, Poovorawan Y. High-dose steroids do not improve early outcome in biliary atresia. J Pediatr Surg 2007;42:2102-5.
25. Kobayashi H, Yamataka A, Koga H, et al. Optimum prednisolone usage in patients with biliary atresia postportoenterostomy. J Pediatr Surg 2005;40:327-30.
26. Davenport M, Parsons C, Tizzard S, Hadzic N. Steroids in biliary atresia: single surgeon, single centre, prospective study. J Hepatol 2013;59:1054-8.
27. Meyers RL, Book LS, O'Gorman MA, et al. High-dose steroids, ursodeoxycholic acid, and chronic intravenous antibiotics improve bile flow after Kasai procedure in infants with biliary atresia. J Pediatr Surg 2003;38:406-1.