A Randomized Control Trial to Evaluate the Effect of Local Instillation of Mitomycin-C at the Porta after Kasai Portoenterostomy in Patients of Biliary Atresia

Abstract

Background:

Kasai portoenterostomy (KPE) is the initial treatment for biliary atresia (BA). Even after initial jaundice clearance, a significant number of children presented with the reappearance of symptoms due to ongoing fibrosis involving porta and intrahepatic ducts. Mitomycin-C (MMC) is an antifibrotic agent, and the study hypothesized that local application of MMC at porta can decrease fibrosis, which can improve jaundice clearance and lead to better native liver survival (NLS).

Materials and Methods:

This prospective randomized control trial included children with BA, who were allocated to groups A or B. The patients in both groups underwent standard KPE; in addition, a 5 French infant feeding tube (IFT) was placed near the porta through the Roux limb in Group B children. During the postoperative period, MMC was locally instilled over the porta in Group B children through IFT. Postoperative jaundice clearance and NLS were assessed and compared.

Results:

A total of 27 children were enrolled in the study, 16 in Group A and 11 in Group B. Both groups were comparable preoperatively. Although the NLS was not statistically significant in Group B, the survival was quite higher, that was 91%, 81%, and 73% at 6 months, 1 year, and 2 years, respectively, compared to 63%, 50%, and 38% in Group A.

Conclusion:

Children in Group B clinically showed an early jaundice clearance and a better trend of serial bilirubin levels as well as longer NLS than Group A, but it was not statistically significant. The procedure was technically easy, and no complication was encountered related to surgical technique or MMC instillation.

INTRODUCTION

Extrahepatic biliary atresia (BA) is an inflammatory and sclerosing cholangiopathy with multifactorial etiology. Kasai portoenterostomy (KPE) is the initial treatment, and almost 40% of children require liver transplantation.[1]

In the majority, after KPE, there is a fall in bilirubin level with clinical improvement. During the follow-up course, there may be a reappearance of jaundice and deterioration of liver function. The ongoing inflammatory and fibrotic processes involving the porta hepatis and intrahepatic bile ducts consequently lead to the reemergence of symptoms.[2,3] In the current study, we hypothesized that the arrest/decrease of this inflammatory/fibrotic process at the porta after surgery could result in biliary drainage for a longer duration, which can improve native liver survival (NLS).

Mitomycin-C (MMC) has shown beneficial effects with its antifibroblastic property in various conditions. In this study, MMC was instilled locally during the early postoperative phase in the region of the portal plate after KPE. This was done through a tube placed at the porta incorporated during the portoenterostomy anastomosis. The study evaluated the technical feasibility of the procedure. It compared follow-up bilirubin levels and NLS between patients with and without MMC instillation.

MATERIALS AND METHODS

This prospective randomized control trial was conducted in the pediatric surgery department at a tertiary care center from December 2018 to December 2022. The Institutional Ethics Committee approved the study (NK/4422/258), and it was registered under the Clinical Trials Registry–India (CTRI/2018/12/016495). All patients of neonatal cholestasis with suspicion of BA underwent standard preoperative investigation protocol, including biochemical liver function tests, ultrasonography, and hepatobiliary iminodiacetic acid scan. Once investigated, an intraoperative cholangiogram was done to confirm the diagnosis. All diagnosed patients were treated by standard extended KPE. A liver wedge biopsy was taken during KPE and examined for various histopathological parameters as adopted previously.[4] All suspected children with BA were randomly allocated to groups A or B by block randomization. The study design was as shown in Figure 1. During the postoperative period, MMC was locally instilled over the porta in Group B children through the infant feeding tube (IFT) according to the protocol. The patients for whom KPE was not performed due to liver cirrhosis and those, whose parents refused to give consent, were excluded from the study. The cystic BA and syndromic patients were also excluded.

Figure 1.

Study design. BA: Biliary atresia, IOC: Intraoperative cholangiogram, MMC: Mitomycin-C, KPE: Kasai portoenterostomy

Surgical technique and postoperative protocol

All children underwent a limited laparotomy through a right subcostal incision for an intraoperative cholangiogram. In cases where the gall bladder was atretic, a cholangiogram was not done. After the diagnosis of BA was confirmed and the decision to proceed with KPE was taken, the incision was extended for formal laparotomy. A standard dissection was done at the porta, excising all the fibrotic elements, and a 25 cm Roux-en-Y hepaticojejunostomy was performed.

In Group B patients, to instill MMC in the postoperative period, a 5 French IFT was placed in the Roux limb [Figures 2 and 3]. The tip of the IFT was placed near the portal plate and fixed with chromic catgut suture to ensure the delivery of MMC over the target area. The other end of the IFT was made to exit through the Roux loop just like a feeding jejunostomy [Figures 2 and 3] and was secured to the abdominal wall after incision closure. This tube could now be used for drug delivery, as per the protocol mentioned below.

Figure 2.

Intraoperative image (a) showing the Roux limb (white arrow) through the retrocolic window (yellow star) and measurement of the infant feeding tube (yellow arrow) along the Roux limb in relation to the porta (red arrow). (b) showing the entry of the feeding tube (black arrow) in the Roux limb and feeding tube at the porta. (c) showing fixation of feeding tube with chromic catgut suture at the entry site with serosa and at the porta site (blue arrow) with mucosa

Figure 3.

Intraoperative image (a) showing completion of portoenterostomy (white arrow) and entry site of feeding tube in Roux limb (black arrow). (b) is showing external appearance after abdominal wall closure, fixation of feeding tube at abdominal wall and placement of corrugated rubber drain. Image (c) is a line diagram of the standard portoenterostomy while image (d) shows feeding tube (blue tube, blue arrow) in Roux limb. In both images (c&d) black arrow is representing Roux limb and red arrow represents porta area. [L= Liver, S= Stomach]

All children received 1 week of intravenous antibiotics (piperacillin-tazobactam, amikacin, and metronidazole) and were kept nil by mouth for 4 days. The follow-up was the same for both groups. In addition, for Group B children, MMC was instilled through the IFT according to the protocol. All children were followed weekly for the initial 3 weeks, then once a month for 3 months, followed by 3 monthly till 2 years of age. Any child who had presented in between with features of cholangitis was treated according to the severity of the disease.

All children had received oral steroids (prednisolone) for 1 month and prophylactic oral antibiotics for 3 months. Ursodeoxycholic acid (UDCA) and phenobarbitone were prescribed for a year.

Mitomycin-C treatment protocol

The patients in study Group B received three doses of MMC by local instillation through the IFT. The first dose was instilled on the postoperative day (POD) 10, the second dose on POD-17, and the last on POD-24. Before the procedure, the child was kept nil per oral for an hour. The IFT was flushed with 2 mL of normal saline just before the instillation. The child was held in a head-down position (Trendelenburg position) and remained in the same posture for 15 min after the instillation. The dose was 1 mg of MMC diluted in 2 mL (containing 0.5 mg/mL) of normal saline. The instillation was done slowly over 2 min, and no flushing of IFT was done after the instillation.

Data analysis

All preoperative variables which are shown in Table 1 hold normality and, therefore, summarized with a mean (standard deviation [SD]), and comparison between the groups was done by the independent t-test. The age and alanine transaminase (ALT) were summarized with the median, and hence, a comparison was done by the rank-sum test. All preoperative variables of liver biopsy parameters were qualitative, and thus, the Fisher's exact test was used. Bilirubin was recorded at a different time interval for both groups; however, the sample size was unequal; hence, a two-way repeated measure ANOVA for an unbalanced design was used. The point plot [Figure 4] with a mean (SD) was used to summarize the above-mentioned data. The survival was depicted using the Kaplan–Meier curve [Figure 5] and was compared using the log-rank test. All the above-mentioned statistical tests were conducted at a 5% significance level. Data entry was done using MS Excel on a Window machine, and analysis was conducted on R version 4.2.1 (R core team, R foundation for statistical computing, Vienna, Austria, 2013, http://www.R-project.org) withmacOs ventura, version 13.1.

Table 1.

Preoperative demographics and clinical characteristics of children in both groups

Parameter	Group-A (KPE) n=16	Group-B (KPE + M) n=11	P
Age (days), Median (IQR)	78.5 (62, 85.5)	72 (62, 76.5)	0.430
Weight (kg), Mean (SD)	4.1 (0.8)	4.2 (0.8)	0.947
M: F, n (%)	7 (43.8): 9 (56.2)	3 (27.3): 8 (72.7)	0.448
Bilirubin (mg/dl), Mean (SD)	10.6 (3.6)	9.7 (2.3)	0.479
AST (IU/L), Mean (SD)	261.6 (145.2)	310.9 (158.8)	0.411
ALT (IU/L), Median (IQR)	147 (94, 246.8)	173 (107.5, 237.5)	0.863
GGT (IU/L), Mean (SD)	547.7 (358.7)	442.8 (273.9)	0.565
ALP (IU/L), Mean (SD)	690.9 (186)	562.5 (304.2)	0.184

Age=Age at Surgery, n=number of patients, KPE=Kasai portoenterostomy, KPE + M = Kasai portoenterostomy plus Mitomycin-C, IQR=interquartile range, SD=standard deviation, kg=Kilogram, M: F=Male: Female, ALT=Alanine transaminase, AST=Aspartate transaminase, ALP=Alkaline phosphatase, GGT=Gamma-glutamyl transferase, IU/L=International unit/litre

Figure 4.

Dot plot showing the comparison of bilirubin levels (mg/dl) between both the groups sequentially at 3 months, 6 months, 12 months, and 24 months. The connected dots showing the mean of each group at a particular time point. Bil: Bilirubin level, KPE: Kasai portoenterostomy, KPE + M: Kasai portoenterostomy plus Mitomycin-C

Figure 5.

Kaplan–Meier survival curve of both the groups till 24 months of surgery. The X axis is showing the percentage of survived children, while the Y axis showing the timeline in months. (KPE = Kasai portoenterostomy group, KPE + M: Kasai portoenterostomy plus Mitomycin-C group)

RESULTS

A total of 33 patients were screened during the study, and 27 patients who fulfilled the inclusion criteria were enrolled in the study. Six patients were excluded from the study due to the following reasons: two parents refused to give consent, two had an advanced stage of cirrhosis, and one each had BA splenic malformation syndrome and cystic BA. In Group A, there were 16 patients, while in Group B, 11 patients were included. The demographics and preoperative investigations of both groups were comparable [Table 1]. The postoperative bilirubin level trend was different [Figure 4], and also, the NLS between both the groups [Table 2 and Figure 5] was dissimilar. All aforementioned preoperative parameters between groups were reported as nonsignificant. All liver enzymes, including ALT, aspartate transaminase, alkaline phosphatase, and gamma-glutamyl transferase, were raised in both groups. In liver histology, most children showed severe grades of cholestasis and fibrosis, while moderate hepatocellular damage and bile duct proliferation. There was a mild grade of portal edema and inflammation present in all children. Although the maximum bile ductule diameter at the portal plate showed a wide range from 70 μ to 220 μ with an average of 110 μ, the difference was not statistically significant. Furthermore, there was no statistically significant difference between the two groups’ liver histology parameters.

Table 2.

Comparisons of postoperative outcomes between the groups-

Parameter	Group KPE n=16	Group KPE + M n=11	P
NLS at 6-Month, n (%)	10 (62.5)	10 (90.9)	0.10
NLS at 1-Year, n (%)	8 (50)	9 (81.8)	0.09
NLS at 2-Year, n (%)	6 (37.5)	8 (72.7)	0.06

NLS=Native liver survival, N=number of patients, KPE=Kasai portoenterostomy, KPE + M = Kasai portoenterostomy plus Mitomycin-C

The difference between preoperative bilirubin levels and postoperative bilirubin levels at different time points was significantly (P = 0.0001) low in both groups as shown in the dot plot [Figure 4]. The postoperative bilirubin levels of Group A were 4.07, 3.1, 3.65, and 1.68 mg/dl at 3 months, 6 months, 1 year, and 2 years, respectively. Similarly, the value for Group B was 2.71, 2.84, 1.76, and 1.62 mg/dl correspondingly as depicted. Although Group B's bilirubin level was substantially lower at each time point, it was not statistically significantly lower than Group A's (P = 0.297).

The NLS characteristics of both the groups as depicted in Table 2 and by the Kaplan–Meier survival curve [Figure 5], the P value of the log-rank test was not statistically significant. However, it is important to mention that Group B (KPE plus mitomycin-C [KPE + M]) median survival was quite higher for the overall period compared to Group A (KPE). The above-mentioned results of statistical tests were subjected to sample size limitation, and with a balanced design and higher number of participants may turn NLS to be statistically significant.

DISCUSSION

In 1955, Professor Kasai and Suzuki did the first portoenterostomy.[5] Even today, KPE operation is the initial treatment for children with BA and delays the need for transplant.[6,7]

The NLS mainly depends on the status of liver fibrosis/cirrhosis at the time of surgery, the appropriate surgical technique (KPE), and the incidence of postoperative cholangitis. The few contraindications for KPE are liver failure, portal hypertension, and uncorrectable coagulopathy. These children deteriorate rapidly in the postoperative period, resulting in high morbidity and mortality.

In developing countries, many patients cannot afford liver transplantation; from our study group, no patient could undergo a liver transplant due to financial issues or the nonavailability of transplant services. In addition, there are only a few centers that perform pediatric liver transplants, and most of the institutes (like our institute) prefer to do transplants only after 1 year of age when the child's weight is around 10 kg. It further strengthens the role of KPE and the longer duration of bilio-enteric drainage.

Mitomycin-C

MMC is isolated from Streptomyces caespitosus, which has been shown to have antitumor and antifibrotic activity. In humans, MMC is rapidly cleared from the serum after intravenous administration. It is indicated in cancer of the stomach, pancreas, colon, liver, and lungs. MMC toxicity is consistent in all species studied to date. In laboratory animals, the LD50 (lethal dose, 50%) varies from 1.0 to 2.5 mg/kg for intravenous administration, which corresponds with toxicity in humans. Oral toxicity was similar to intravenous toxicity at doses 8–12 times the intravenous doses. In our study, the dose was 1 mg; hence, compared to the intravenous safe dose (20 mg/m²), it was approximately four times less (considering body surface area = 0.27 m² for an infant having 4.5 kg weight and 58 cm height). It has been reported that, in a human, oral administration of 10 mg MMC gave a maximum serum concentration of 0.1 μg/mL, which was approximately 1/20 the maximum concentration, which can be obtained by intravenous administration.[8] In the index study, the dose used was substantially less, and hence, the blood level of MMC was not measured for any drug toxicity.

MMC has shown in vivo and in vitro antifibroblastic activities. It has gained wide acceptance in ophthalmology as an agent to reduce scar formation and restenosis in conditions such as glaucoma, dacryocystorhinostomy, and pterygium surgery.[9,10] The topical application of MMC has been proven to reduce scar formation in the larynx and trachea.[11,12] It has proven its antifibrotic role in esophageal stricture treated with endoscopic dilatation plus 1 mg/mL MMC.[13] Occleston et al. have shown that a single MMC exposure can arrest fibroblasts’ growth and proliferation. However, they noted fibroblasts continue expressing growth factors and forming extracellular matrix molecules.[14]

Nonetheless, several factors may influence MMC efficacy. These factors include the dose delivered to the tissues (which is concentration dependent), volume, duration of exposure, preparation method, administration, and tissue-related factors.[15] Different application techniques have been described. The most frequent was local application through a cotton pledget soaked in MMC solution under direct endoscopic visualization.[16] Spraying onto the stricture is another possible technique.[17] A further alternative, previously reported only in adult studies, involves the injection of MMC directly into each quadrant of the stenosis after dilation.[18,19]

It was mostly reported to be freshly prepared immediately before the application. A systematic review showed that the concentrations of MMC used were between 0.1 and 2 mg/mL (median and mean values of 0.4 and 0.5 mg/mL, respectively). The number of MMC applications varied between 1 and 12, with a mean of 2 and 2.6 in pediatric and adult patients, respectively, although the majority of children (79%) required only 1–2 applications. When MMC was applied more than once, intervals ranged from 1 week to 13 months, with a median of 4 weeks.[20] To date, no study compares the effectiveness of different concentrations and dosages of MMC; the concentration of 0.4 mg/mL is the most used and appears to be effective. In our study, the MMC was freshly prepared; the dose was 0.5 mg/mL with three doses weekly.

The mechanism of MMC to prevent fibrosis in esophageal stricture at the molecular level has been explained by Zhang et al. MMC inhibited fibrosis by enhancing cell apoptosis and inhibiting autophagy through upregulating long noncoding RNA-ATB and downregulating miR-200b. The expression of apoptosis-related proteins (caspase-3 and caspase-8) was upregulated, and fibrosis-related proteins (α-SMA and collagen-1) were downregulated. In addition, MMC could inhibit autophagy according to the results from the transmission electron microscope through the expression of LC3 II and ATG5. After treatment with MMC, no more autophagosomes were observed by transmission electron microscope.[21]

The literature also suggested the role of a redo-KPE in selected cases, especially when patients initially had jaundice clearance. During redo-KPE, the newly developed fibrous extrahepatic tissue is removed, the hilar plate is superficially dissected just before baring the liver parenchyma, and hepaticojejunostomy is done again. Considering the index study intervention from this perspective, one can understand the various mechanisms by which one can get persistent bilio-enteric drainage. To begin with, flushing the feeding tube with saline before MMC instillation could remove the inspissated bile from the ductule and open the blocked ductules. The vital mechanism would be, MMC may act on ductules and prevent their subsequent fibrosis. In addition, MMC can prevent the development of new fibrosis (fibrotic tissue) over the portal area.

Although the difference in bilirubin levels and NLS between the groups was not statistically significant, there was an apparent clinical improvement. To achieve a statistically significant better result in a study, the intervention should affect the result drastically. Here, in BA, the aim is to achieve a “marginal gain” with each intervention after KPE. The addition of steroids, UDCA, and phenobarbitone also provides only a marginal gain. Unless the etiopathogenesis of different types of BA is clear and an intervention to act on the basic cause is developed, the gain would be marginal with any kind of intervention. This study provides a unique channel to access the porta after KPE; hence, different drugs can be tried (like topical steroids) to prevent fibrosis. This study is an innovative work with a novel intervention that has not been described in English literature.

There were some weaknesses in the study, the foremost being a smaller number of participants. As this study was conducted during the COVID pandemic, it was difficult for the patients to reach the hospital. BA was considered a semi-emergency in our hospital during the COVID pandemic, and surgical intervention was never delayed once the patient arrived. In addition, the outcome had not considered cytomegalovirus (CMV) status and the incidence of postoperative cholangitis episodes. In the future, a more extensive study with longer follow-ups considering CMV and cholangitis can explain the results more precisely.

CONCLUSION

Children in the KPE + M group clinically showed an early jaundice clearance and a better trend of serial bilirubin level as well as longer NLS than the standard KPE group. The aforementioned results, while clinically promising, are restricted in terms of being statistically insignificant. Henceforth, this limitation (statistical nonsignificance) requires confirmation with a balanced design with a sufficient sample size. The procedure was technically easy, and no complication was encountered related to the surgical technique or MMC instillation.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.