Recurrence of primary sclerosing cholangitis after liver transplant in children: an international observational study.

Abstract

Background:

Recurrent primary sclerosing cholangitis (rPSC) following liver transplant (LT) has a negative impact on graft and patient survival; little is known about risk factors for rPSC or disease course in children.

Approach & Results:

We retrospectively evaluated risk factors for rPSC in 140 children from the Pediatric PSC Consortium, a multicenter international registry. Recipients underwent LT for primary sclerosing cholangitis and had >90 days of follow-up. The primary outcome, rPSC, was defined using Graziadei criteria. Median follow-up after LT was 3-years [IQR 1.1-6.1]. rPSC occurred in 36 children, representing 10% and 27% of the subjects at 2- and 5 years post-LT, respectively. Subjects with rPSC were younger at LT (12.9 vs. 16.2 years), had faster progression from PSC diagnosis to LT (2.5 vs. 4.1 years), and had higher alanine aminotransferase (112 vs. 66 IU/L) at LT; (all p < 0.01). Inflammatory bowel disease was more prevalent in the rPSC group (86% vs. 66%, p=0.025). After LT, rPSC subjects had more episodes of biopsy-proved acute rejection (mean 3 vs 1, p<0.001), and higher prevalence of steroid-refractory rejection (41% vs. 20%, p=0.04). In those with rPSC, 43% developed complications of portal hypertension, were re-listed for LT, or died within two years of the diagnosis. Mortality was higher in the rPSC group (11.1% vs. 2.9%, p=0.05).

Conclusion:

The incidence of rPSC in this cohort was higher than previously reported, and was associated with increased morbidity and mortality. Patients with rPSC appeared to have a more aggressive, immune-reactive phenotype. These findings underscore the need to understand the immune mechanisms of rPSC in order to lay the foundation for developing new therapies and improve outcomes in this challenging population.

Introduction:

Primary sclerosing cholangitis (PSC) is an immune-mediated liver disease thought to involve the innate and adaptive arms of the immune system. Its exact triggers and driving mechanisms remain poorly understood (1-4). PSC accounts for less than 3 % of liver transplants (LT) in children (5, 6). Among children with PSC, 30% will require a LT within ten years of diagnosis(7). Children with PSC undergo LT for life-threatening complications of cirrhosis and portal hypertension, while intractable pruritus, recurrent cholangitis(5, 8), and cholangiocarcinoma are less frequent indications (7). Unfortunately, the disease can recur in the allograft. In adults, recurrent PSC (rPSC) occurs in approximately 20% of LT recipients (6, 9, 10), and up to 12.4% with rPSC will require re-transplantation (11). However, despite decades of LT experience and the recognition that disease recurrence negatively impacts graft and patient survival, the true incidence of rPSC in children remains unknown. Specific diagnostic criteria are lacking, and disease recurrence features frequently overlap with other clinical presentations such as ischemia, infections, drug-induced liver injury, rejection, or biliary complications. Furthermore, the impact of recipient and allograft factors in disease recurrence is not well elucidated.

Prior studies reported that patients who received an extended-criteria donor graft (10), used steroid-free Thymoglobulin induction (6) or primary immunosuppression with tacrolimus (12), developed allograft rejection(13), had poorly controlled inflammatory bowel disease (IBD), or developed de novo IBD(5, 14), were all at a higher risk of disease recurrence. At the same time, a colectomy appeared to be protective (10, 12). The negative impact of rPSC on patient and graft survivals is well established for adults(11, 12, 15, 16). These risk factors are not well established in children, and most pediatric reports are small, single-center cohorts (6, 17, 18). A prior pediatric multicenter transplant registry study, published 10 years ago, reported that rPSC did not have a negative impact on patient or allograft outcomes (5).

Utilizing a large, multicenter cohort—the Pediatric PSC Consortium— we assessed the epidemiology, risk factors, and long-term outcomes for rPSC in children who underwent liver transplant for PSC. We aimed to provide insights into the incidence, diagnosis, risk factors, and clinical implications of rPSC in children with liver transplants.

Methods:

Study population and data source

The International Pediatric PSC Consortium is an active research registry involving 54 sites throughout Europe, the Middle East, Asia, North and South America (7). Thirty-four centers collected retrospective data from patients who underwent LT for a primary indication of PSC before 18 years of age, from 1986 to 2019. After a detailed review of medical records, data were abstracted, de-identified by the collaborating investigators at the local study sites, and submitted through the secure Research Electronic Data Capture (REDcap) platform (19). A detailed description of the whole cohort has been previously reported (7). The indication for and timing of LT were based on local clinical practices and established guidelines. Analyzed variables included demographics, laboratory data, donor and graft type, perioperative data, type of immunosuppression, histopathology reports, cholangiography, and endoscopy on each patient. LT recipients with less than 90 days of follow-up were excluded from this analysis. IBD activity was evaluated using physician global assessment at the time of LT as previously described(20). Patients with incomplete data were excluded from those sections of the analysis.

Outcomes

The study primary outcome was the diagnosis of rPSC as defined by Graziadei (21). Criteria for rPSC include: (a) confirmed diagnosis of PSC before LT (b) cholestatic biochemistry post- LT with cholangiography showing multifocal non-anastomotic biliary strictures with beading of bile ducts or fibro-obliterative lesions, (c) the absence of chronic ductopenic rejection, hepatic artery thrombosis/stenosis, or donor-recipient blood type incompatibility at least 90 days after LT. Any questions or discrepancies regarding the data points required to define disease recurrence were resolved between the local co-investigator and study authors.

To evaluate the impact of rPSC and the overall patient and graft outcomes, secondary clinical endpoints were established: (a) the development of portal hypertensive complications (ascites, hepatic encephalopathy, or esophageal varices with or without bleeding), (b) re-LT or listing for LT, or (c) death from liver disease progression or during re-LT. Event-free survival was defined as the absence of all (a-c) of the above.

Statistical analysis

Descriptive statistics were calculated: continuous variables were reported as median and interquartile range (IQR), and categorical variables as count and proportions. Continuous variables were compared using the rank-sum test, and categorical variables were compared with chi-square analysis. We created a retrospective cohort of all children with LT who met inclusion criteria, starting at the LT date. Observations were censored at the date of the last known follow-up. We used the Kaplan-Meier method to calculate outcome probabilities and Cox regression to compare the association between demographic, phenotypic, biochemical, donor, graft, and post-transplant medication factors and rPSC. The proportional hazards assumption was assessed graphically. Stata version 16.0 (StataCorp, College Station, TX) was used for statistical analysis.

Ethical considerations

The institutional review board of each participating center approved all research work. This research was conducted following the Declaration of Helsinki guidelines of good practice.

Results:

Study Population:

In this cohort of 1325 children with PSC, 172 underwent LT. Thirty-two patients with less than 90 days of follow-up after LT or incomplete primary data were excluded, leaving 140 for analysis. The median age at LT was 15.3 years [IQR 12.3-17.7]; 55% were male. The median time from PSC diagnosis to LT was 3.7-years [IQR: 1.6-5.9]. Patients were followed after LT for a median of 3-years [IQR 1.1-6.1], representing 592 total person-years of follow-up. The 1, 5 and 10-year graft survival was 97% [95% confidence interval (CI) 92-99%], 91% [95%CI 83-95%], and 71% [95%CI 52-84%], respectively.

Thirty-six subjects met the criteria for rPSC, at a median of 3.3 years [IQR 1.7-6.0] after LT. The probability of rPSC at two years after LT was 10% (9/86 subjects), at five years was 27% (13/47 subjects), and at ten years was 47% (6/12 subjects), Figure 1a.

Figure 1:

Kaplan-Meier probability of:1.a) Recurrence of primary sclerosing cholangitis after liver transplantation for PSC in children; 1b) Comparison of survival with allograft between the cohorts with and without recurrent PSC from time of transplantation; 1c) Event-free survival after diagnosis of rPSC; 1d) Survival with transplanted liver after diagnosis of recurrence of PSC. Abbreviations: CI: confidence interval, PSC: primary sclerosing cholangitis, rPSC: recurrent primary sclerosing cholangitis.

The diagnosis of rPSC was rendered by the treating physician based on a combination of biochemical, radiological and histological findings. All the subjects with rPSC had elevated liver enzymes with predominantly cholestatic laboratory abnormalities, and cholangiographic changes demonstrating multiple non-anastomotic strictures and beading of intrahepatic bile ducts. Figure 2. Seventeen patients (47%) had a liver biopsy demonstrating fibro-obliterative lesions or bile ductular proliferation.

Figure 2.

Left panel: Thick slab magnetic resonance cholangiography, demonstrating diffuse intrahepatic biliary ductal dilation, left greater than right, bile duct beading in the left bile duct, representing recurrent primary sclerosing cholangitisRight panel: Endoscopic retrograde cholangiogram after liver transplant with Roux-en-Y biliary reconstruction using push enteroscopy. Above balloon cholangiogram showing diffuse irregularity and beaded appearance of the right and left intrahepatic branches and the absence of the peripheral branches

Risk factors for disease recurrence in allograft:

The demographics and laboratory values at diagnosis of PSC and LT are shown in Table 1. Patients with rPSC (n=36) were younger at the time of LT, and the time from initial PSC diagnosis to LT was significantly shorter for those with rPSC. IBD was more prevalent in the rPSC group (86% vs 66%), particularly in the ulcerative colitis sub-group. The were no differences of the IBD activity using physician global assessment at the time of LT Table 1. Two patients underwent colectomy before LT, and neither developed rPSC. Five patients underwent colectomy after LT, and two developed rPSC.

Table 1:

Comparison of demographics and laboratory values at diagnosis, time of LT, 6 months, and 12 months after LT between the rPSC and no rPSC cohorts.

Variables	rPSC (n=36)	No rPSC (n=104)	p value
Demographics and phenotype (n=140)
Male sex	61%	52%	0.340
Age at PSC diagnosis (years)	9.5 [7.8-12.3]	12.3 [7.8-14.2]	0.082
Age at LT (years)	12.9 [10.7-14.7]	16.2 [13.6-18.0]	<0.001
Time between diagnosis and LT (years)	2.5 [0.7-4.6]	4.1 [2.3-6.6]	0.003
Features of overlap with AIH (%)	42	29	0.091
Presence of IBD (%)	86%	66%	0.025
IBD phenotype (% of the total IBD)
Ulcerative Colitis	75%	53%	0.050
Chron’s Disease	11%	13%
IBD active (%)	33%	27%	0.64
IBD in remission (%)	67%	73%
Laboratory values at PSC diagnosis (n=140)
Hemoglobin	12.2 [10.3-14.2]	12.3 [11.1-13.3]	0.852
Platelet count (x10(3)/uL)	279 [93-463]	216 [113-352]	0.401
International Normalized Ratio	1.2 [1-1.3]	1.1 [1-1.3]	0.942
Albumin (g/dL)	3.7 [3.2-3.9]	3.7 [3.3-4.0]	0.901
Gamma-glutamyltransferase (U/L)	234 [153-440]	307 [170-448]	0.968
AST (U/L)	144 [72-308]	143 [69-239]	0.841
ALT (U/L)	168 [105-298]	118 [82-196]	0.049
Total bilirubin (mg/dL)	2.7 [0.7-8.8]	1.3 [0.6-3.8]	0.217
Laboratory values at liver transplantation ( n=140)
Hemoglobin	10.9 [9.4-12.6]	11.7 [9.6-13.4]	0.232
Platelet count (x10(3)/uL)	97 [60-250]	89 [50-150]	0.463
International Normalized Ratio	1.4 [1.2-1.7]	1.3 [1.1-1.5]	0.338
Albumin (g/dL)	3.3 [2.9-3.6]	3.4 [2.9-4.1]	0.331
Gamma-glutamyltransferase (U/L)	157 [66-358]	167 [90-280]	0.966
AST (U/L)	112 [52-290]	108 [75-164]	0.315
ALT (U/L)	112 [46-319]	66 [51-114]	<0.001
Total bilirubin (mg/dL)	8.2 [3.3-21.3]	4.4 [2.1-10.3]	0.124
Laboratory values 6 months after liver transplantation (n=140)
Hemoglobin	12.6 [11.3-14.1]	12.2 [10.5-14]	0.542
Platelet count (x10(3)/uL)	161 [130-218]	171 [112-227]	0.916
International Normalized Ratio	1.1 [1-1.3]	1.1 [1-1.2]	0.399
Albumin (g/dL)	4.2 [4-4.4]	4.2 [3.9-4.5]	0.799
Gamma-glutamyltransferase (U/L)	41 [19-82]	31 [19-91]	0.701
AST (U/L)	33 [21-69]	26 [20-48]	0.388
ALT (U/L)	28 [21-71]	33 [19-61]	0.958
Total bilirubin (mg/dL)	0.7 [0.4-1.9]	0.7 [0.5-1.4]	0.768
Laboratory values 12 months after liver transplantation (n=140)
Hemoglobin	12.7 [10.7-14.4]	13.1 [11.7-14.7]	0.332
Platelet count (x10(3)/uL)	196 [138-292]	162 [129-234]	0.215
International Normalized Ratio	1.1 [1-1.3]	1.1 [1-1.2]	0.863
Albumin (g/dL)	4 [3.9-4.3]	4.2 [3.9-4.3]	0.376
Gamma-glutamyltransferase (U/L)	61 [35-312]	36 [19-74]	0.020
AST (U/L)	58 [31-77]	28 [20-46]	<0.001
ALT (U/L)	55 [27-93]	30 [19-57]	0.018
Total bilirubin (mg/dL)	0.9 [0.5-1.9]	0.7 [0.5-1.4]	0.597
IBD activity at LT (n=54)	rPSC(n=15)	No rPSC (n=39)
PGA 0	8	14	0.502
PGA 1	5	18
PGA 2	1	6
PGA 3	1	1

Abbreviation: AIH: autoimmune hepatitis, IBD: Inflammatory bowel disease, g/dL: grams per deciliters, LT: liver transplantation, mg/dL: milligrams per deciliters, PGA: Physician global assessment, PSC: primary sclerosing cholangitis, rPSC: recurrent PSC, U/L: unit per liters, uL: microliter.

At PSC diagnosis, liver transaminases, bilirubin, hemoglobin, and platelet count were similar in both groups Table 1. At LT, those with rPSC had higher alanine aminotransferase (ALT), aspartate aminotransferase (AST), and bilirubin. At 6 months post-LT, transaminases and bilirubin did not differ between the two groups. But at 12 months, patients with rPSC had significantly higher gamma-glutamyltransferase (GGT), AST, and ALT than those without rPSC. Model for end-stage liver disease (MELD) scores were similar in both groups at listing and LT. Type of graft and biliary anastomosis were not associated with rPSC Table 2. Biliary leaks occurred in only 7 (6.8%), with no reported anastomotic strictures. Five patients developed portal vein thrombosis, and five developed hepatic artery thrombosis. One patient in the rPSC cohort had transient hepatic artery thrombosis; the arterial flow was successfully re-established during the first 24h. The patient has normal graft function for 10 years before developing rPSC . The hepatic artery flow was normal at the time of rPSC diagnosis. No patient had both complications.

Table 2:

Comparison of graft type and perioperative variables between the recurrent PSC (rPSC) and no rPSC cohorts. Complete data from 103 subjects, 25 with rPSC

Variables	rPSC (n=25)	No rPSC (n=79)	p value
Donor type
Living-related	18%	20%	0.601
Living-unrelated	0%	6%
Deceased (braindead)	64%	62%
Deceased (after cardiac death)	18%	12%
Donor age (Median [IQR])	27 [16-42]	24 [15-39]	0.951
Graft type
Whole	73%	68%	0.594
Technical variant graft	27%	32%
Biliary anastomosis
Duct-to-duct	13%	25%	0.210
Duct-to-roux	87%	75%
Ischemia time in minutes
Warm (Median [IQR])	23 [0-45]	43 [31-45]	0.265
Cold (Median [IQR])	204 [65-342]	170 [33-257]	0.860
MELD
At listing (Median [IQR])	15 [11-23]	15 [10-21]	0.845
At LT (Median [IQR])	19[15-29]	22 [17-30]	0.941
Increase, from listing to LT (Median [IQR])	+4[2-7]	+2 [0-14]	0.377

Abbreviation: IQR: interquartile range, LT: liver transplantation, rPSC: recurrent primary sclerosing cholangitis.

Complete data on perioperative variables, post-LT immunosuppression, rejection and viral replication were available in 103 subjects, including 25 with rPSC. Neither induction nor maintenance immunosuppression was associated with rPSC Table 3. The rPSC patients had a higher number of biopsy-proven allograft acute rejection episodes (median 3 vs.1, p<0.001), and those rejection episodes were more likely to be steroid-resistant (41% vs. 20%, p=0.04). The majority of rejection episodes occurred before the diagnosis of rPSC, Table 4. Those with rPSC had a higher prevalence of Epstein Barr virus (EBV) viremia during follow-up (46% vs 21%), although this difference did not reach statistical significance.

Table 3:

Comparison of Immunosuppression management between the rPSC and no rPSC cohorts. Complete data from 103 subjects, 25 with rPSC.

Variables	rPSC (n=25)	No rPSC (n=79)	p
Immunosuppression induction
Steroid induction	91%	90%	0.809
Basiliximab	8%	19%	0.234
Thymoglobulin	8%	10%	0.809
Initial steroid taper duration (months)	4 [2-6]	4 [2-8]	0.831
Immunosuppression Maintenance
Tacrolimus monotherapy	44%	42%	0.510
Tacrolimus + mycophenolate	36%	43%
Tacrolimus + thiopurines	4%	5%
Tacrolimus + mTOR inhibitor	16%	6%
Cyclosporine monotherapy	0%	3%
Ursodeoxycholic Acid post-LT	65%	54%	0.326

Abbreviation: Post-LT: post liver transplantation, mTOR: mammalian target of rapamycin, rPSC: recurrent primary sclerosing cholangitis

Table 4:

Differences in allograft rejection and viral replication between the rPSC and no rPSC cohorts. Complete data from 103 subjects, 25 with rPSC.

Variables	rPSC (n=25)	No rPSC (n=79)	p value
Allograft Rejection
Number of episodes	3 [1-5]	1 [0-2]	<0.001
Episodes of rejection per year after LT	0.8	0.3	<0.001
Time to first episode	3.8mo [1mo-1yr]	6.3mo [1.5mo-1.9yr]	0.345
Steroid resistant	41%	20%	0.046
Viral replication and serologies
EVB viral replication	41%	26%	0.103
EBV recipient positive	37%	47%	0.434
EBV donor positive	83%	63%	0.173
EBV recipient negative/donor positive	73%	43%	0.425
CMV replication	11%	9%	0.731
CMV patient positive	42%	37%	0.661
CMV donor positive	65%	53%	0.411
CMV recipient negative/donor positive	27%	29%	0.852

Abbreviations: CMV: cytomegalovirus, EBV: Epstein Barr virus, rPSC: recurrent primary sclerosing cholangitis

Impact of disease recurrence on patient and graft survival:

Overall, survival with allograft after LT was worse in rPSC vs. patients without recurrence, as shown in figure 1b. Seven patients with rPSC developed portal hypertensive complications at a median of 1.7 years after the diagnosis of disease recurrence. Overall, event-free survival after diagnosis of rPSC was 71% (95%CI 52-84%) at 1 year and 57% (95%CI 36-74%) at 2 years, as shown in Figure 1c. One patient died of post-transplant lymphoproliferative disorder. Six patients were re-listed for a second LT. One patient died of complications of end-stage liver disease while awaiting a second LT. Five patients underwent a second LT. Two patients died, one of perioperative complications 1mo after the second LT, and one of sepsis and multiorgan failure 4.5 years after the second LT. Three patients were alive at 3, 4, and 5 years after a second LT at the end of follow-up. Overall, after the diagnosis of rPSC, allograft survival was 90% (95%CI 72-97%) at one year, 81% (95%CI 59-92%) at two years, and 63% (95%CI 37-80%) at five years, as shown in Figure 1d. In those without rPSC after LT, there were three deaths. One patient died from metastatic cholangiocarcinoma and 2 of sepsis with multiorgan failure. In summary, Seven patients (4.3% of 140) died during post-LT follow-up; mortality was higher in the rPSC cohort (11.1% vs. 2.9%, p=0.05). There were no other graft losses in patients who were alive at follow-up.

Discussion:

We analyzed a multicenter cohort of children who underwent LT for PSC. In this cohort, 27% developed rPSC by five years after LT (n=47 with 5 years of follow-up). rPSC was associated with a more aggressive, immunoreactive phenotype of PSC pre-LT characterized by younger age at LT, faster progression to end-stage liver disease, higher prevalence of IBD, and more frequent and difficult to treat episodes of allograft rejection. rPSC was associated with substantial morbidity and mortality. We identified no modifiable peri-transplant risk factors associated with rPSC.

Data regarding rPSC incidence in children are sparse and heterogeneous (5, 6). In this multicenter, multi-national series of children transplanted for PSC, 27% developed rPSC by 5 years post-LT. This recurrence rate is higher than the 17.5% reported by Gordon et al.(22) and12% reported by Campsen et al. (16), but lower than the 37% reported by Vera et al. (23); all 3 of these are adult cohorts with similar follow-up time. Miloh et al.(5) identified rPSC in 6 of 61 (10%) of children followed for 18.7+13.8 months post-LT, which is similar to our observed 2 years disease recurrence rate. Our cohort has a longer duration of follow-up, with 33.5% having a 5-year follow-up compared to only 20% in the group reported by Miloh (5), which could explain the higher detection rate of rPSC in our study. However, it is important to note that we do not yet have 5 or 10 years follow-up on the majority of children in the Pediatric PSC consortium; longer-term follow-up on a larger cohort will be critical to understanding rPSC fully. The high variability in reporter rPSC incidence could also be due to differences in diagnostic criteria, length of follow-up, and varying disease phenotype (24).

Miloh et al. (5) reported that the prevalence of complications after LT for PSC did not differ significantly from the complication rate in children transplanted for other indications and that disease recurrence did not have a negative impact on patient and graft survivals. However, in our cohort, at 2.5 years after rPSC diagnosis, almost half of our patients had a complication related to disease recurrence. The negative impact of rPSC on patient and graft survival is well established for adults(11, 12, 15, 16), but this is the first sizeable pediatric cohort reporting this finding. Careful monitoring, prompt diagnosis, and new therapeutic strategies are needed to improve graft and patient survivals.

Our data suggest that rPSC may represent a more aggressive form of the disease in the allograft—which may be in conjunction with, triggering, or following other immune activation (allograft rejection). Patients with rPSC showed a more severe, immune-reactive phenotype of PSC pre-LT. Our observations concur with previous reports demonstrating that patients with rPSC were younger at LT(6), had a faster progression from diagnosis to LT(12, 14), and higher ALT at LT(25). In addition to more immune-reactive disease, patients with rPSC also had a significantly higher frequency of IBD, as has been reported in other pediatric (5, 6) and adult studies(10, 12, 14, 15, 23). One explanation for the well-known association between disease recurrence and IBD activity(23) is that intestinal microbiota modulates liver disease(26). It has been suggested that antigens associated with gut inflammation cause immune activation and liver inflammation, (4) influencing overall clinical outcomes. Recently, environmental factors, the microbiome, and imbalanced bile acid composition have been implicated as potential contributing factors(26, 27).

In agreement with other reports (13, 28), rPSC subjects in this cohort had more frequent and more steroid-refractory allograft rejection occurring before the diagnosis of rPSC. Increased risk of rejection and recurrence of the primary disorder makes post-LT management—of immunosuppression and other interventions—very challenging. It is possible, though, that some patients were misdiagnosed with rejection rather than receiving the appropriate diagnosis of rPSC. This finding highlights the need for new diagnostic algorithms—and high clinical suspicion— to facilitate accurate differentiation of alloreactivity from rPSC. The rPSC cohort also had higher liver enzyme levels (ALT, AST, GGT) at 1-year post-LT, which aligns with Miloh et al.(5) that report significantly higher levels of aminotransferases at 5 years post-LT in patients with PSC when compared with patients undergoing LT for non-PSC indications. Despite the lack of specificity of liver enzyme elevations, there is an overall tendency to default to the diagnosis of rejection in transplant recipients; our findings suggest that consideration of early disease recurrence is essential. The allografts of LT recipients for PSC should undergo a detailed histological and radiological evaluation to exclude other possible etiologies of graft dysfunction before increasing immunosuppression. This strategy could prevent over-immunosuppression, limiting the side effects associated with it.

Interestingly, the higher frequency of EBV viremia in the rPSC group may indirectly indicate a higher exposure to immunosuppression (29). Also, it seems likely that they received more immunosuppression, given that they had more rejection episodes. Alternatively, augmented immunosuppression during graft dysfunction due to misdiagnosing rPSC as rejection could have also contributed. Translational investigations have demonstrated that T and B-cells activated and expanded pre-LT remain in the body; they can trigger allo- or autoimmune responses post-LT due to cross-reactive antigen recognition(30, 31). A more substantial lymphocyte expansion pre-LT could induce stronger alloimmune and autoimmune responses post-LT. Post-LT graft dysfunction in these children might thus represent a combination outcome. Increased risk of rejection and recurrence of the primary disorder makes post-LT management—of immunosuppression and other interventions—very challenging. It is also possible, though, that some patients were misdiagnosed with rejection rather than receiving the appropriate diagnosis of rPSC. This observation highlights the need for new diagnostic algorithms—and high clinical suspicion— to facilitate accurate differentiation of rejection from rPSC and more precise therapy.

We did not identify any modifiable peri-or post-transplant risk factors for rPSC—highlighting the need for developing new therapies to prevent and treat rPSC to improve long-term outcomes. Others have reported that colectomy in the peri-transplant period seems protective for rPSC(10, 12). Regrettably, given our limited sample size—particularly those with colectomy—we could not explore its potential protective effects. As reported by others, graft and donor type were not associated with rPSC in our cohort (16) . Interestingly, the type of biliary reconstruction was also not associated with rPSC, and a Roux-en-Y hepaticojejunostomy did not prevent rPSC (32, 33), which differs from previous reports(34). Our observation that ursodeoxycholic acid does not affect disease recurrence aligns with our prior report of it lacks effect in preventing PSC progression in the native liver (35).

Contrary to previous findings (13, 16), we did not observe any differences in immunosuppression management between the 2 groups, and immunosuppression medications—although variable—did not appear to impact rPSC incidence. However, this finding could be due to our modest sample size and incomplete data (e.g., no doses or tacrolimus troughs available) suitable to the study's retrospective nature. Non-adherence can trigger rejection and possibly increase the recurrence of autoimmune liver disease. Adherence to prescribed medications may represent one of the few modifiable risk factors to prevent graft loss in these patients. Autoimmune liver disorders have been reported among the highest causes of graft loss and mortality in young adults during the transition of care(36), and most children with end-stage liver disease from PSC are teenagers who transition care to adult programs soon after LT. Unfortunately, out of the scope of this retrospective study is the impact of patient non-adherence to immunosuppression on rPSC and graft dysfunction, but it is a formidable topic for future study. Over the three decades from which our registry includes data, no new therapies have been developed to improve these patients' outcomes.

Strengths and Limitations:

This study includes the largest reported number of pediatric patients transplanted for PSC, a rare disease for analysis, and was possible due to collaboration between many centers participating in the Pediatric PSC Consortium database. Limitations include the retrospective nature of data collection, creating the potential for missing cases of rPSC. We also do not yet have long-term follow-up (longer than 5 years) on the majority of cases. Furthermore, our report lacks detail in areas of interest such as liver histology, IBD activity, HLA phenotyping, and immunosuppression. The gaps reported in this cohort reflect the lack of standard diagnostic or management practices in our multicenter clinical cohort, particularly in the surveillance and workup for rPSC and IBD activity. The actual rate of rPSC may have also been higher had all patients undergone protocol liver biopsy and/or cholangiography at set time points. We should also acknowledge that a limitation in the field is the lack of a standard clinical assessment that provides accurate differentiation between allograft rejection and rPSC. Additional longitudinal studies that include histology review—to identify biopsies or other early markers of rPSC and differentiate it from acute rejection—would also help earlier, more accurate diagnoses of recurrence, and potentially allow treatment optimization or even prepare the field for clinical trials of drugs development. Further studies of these remaining gaps could facilitate the development of specific immunosuppression management protocols to mitigate immunoreactivity, avoiding rejection and reactivation of donor-reactive lymphocytes.

Conclusions:

In summary, while LT is a successful therapy for patients with refractory complications related to PSC, disease recurrence affects around one-quarter of children after 5 years of follow-up. A more aggressive, immune-reactive phenotype of the condition characterized by faster progression to end-stage liver disease, higher frequency of IBD, and more severe hepatobiliary inflammation at LT is associated with rPSC. Graft rejections were diagnosed more frequently, occurred earlier after transplantation, and were refractory to conventional management in subjects who developed rPSC. Liver transaminases were higher at 1-year post-LT in patients with rPSC and should prompt histological and radiological evaluation for rPSC. Disease recurrence is associated with substantial morbidity and mortality early after diagnosis.

This is the first pediatric study describing the negative impact of rPSC on patient and allograft survival, emphasizing the importance of conducting prospective well-designed multicenter collaborations to elucidate the risk factors for rPSC and to modify disease behavior subsequently. More meticulous radiological and histological monitoring of the allograft might aid accurate assessment of alloreactivity and disease recurrence timing. Societal guidance should reflect this need in order to change current clinical practices. There is an urgent need to unravel the pathogenesis of PSC and rPSC, to support the development of novel diagnostic and therapeutic approaches to control the immune reactivity that fuels a resurgent immune attack on the allograft.

Abbreviations:

PSC

primary sclerosing cholangitis

LT

Liver transplant

rPSC

recurrent primary sclerosing cholangitis

ALT

alanine aminotransferase

AST

aspartate aminotransferase

GGT

gamma-glutamyltransferase

IBD

inflammatory bowel disease

IQR

interquartile range

CI

confidence interval

MELD

Model for end-stage liver disease

EBV

Epstein Barr Virus