The changing clinical presentation of recurrent primary biliary cirrhosis after liver transplantation : Transplantation

Journal Logo


The changing clinical presentation of recurrent primary biliary cirrhosis after liver transplantation

Sanchez, Edmund Q.; Levy, Marlon F.; Goldstein, Robert M.; Fasola, Carlos G.; Tillery, Glenn W.; Netto, George J.; Watkins, David L.; Weinstein, Jeffrey S.; Murray, Natalie G.; Byers, Derek; Christensen, Laura L.; Klintmalm, Goran B.

Author Information
Transplantation 76(11):p 1583-1588, December 15, 2003. | DOI: 10.1097/01.TP.0000090867.83666.F7
  • Free


The discussion of recurrent primary biliary cirrhosis (RPBC) and its clinical existence and relevance have been explored in the past. Patient and graft survival times have increased because of improvements in posttransplant patient management and immunosuppression medications, which allow for longer posttransplant follow-up periods. Primary biliary cirrhosis (PBC) is a disease that manifests itself insidiously, as does RPBC. With longer posttransplant follow-up periods, we are now seeing increasing numbers of recurrent disease and are monitoring its clinical course.

PBC is characterized by the immune-mediated progressive destruction of interlobular bile ducts. Antimicrosomal antibodies to pyruvate dehydrogenase E2, branched chain 2-oxo-acid pyruvate dehydrogenase E2, and 2-oxo-acid pyruvate dehydrogenase E2 have been implicated in 95% of PBC cases. The remaining 5% have PBC-specific antinuclear auto antibodies directed against the nuclear pore complex and other targets (1). The activated lymphoid cells migrate into the biliary epithelial layer through complex interactions involving intercellular adhesion molecule-1, vascular cell adhesion molecule-1, leukocyte factor antigen-1, and vanillacetic-4 (2). This complex autoimmune interaction inevitably results in granulomas, chronic destructive cholangitis, and subsequent cirrhosis. It is unclear whether or not the “recurrent” disease is a continuation of the primary disease or a separate autoimmune entity.

In any event, transplantation for PBC has been shown to have an excellent prognosis. The timing of transplantation must be optimized and performed before the point of rapid clinical deterioration, not only to save the patient’s life but also to improve the quality of life posttransplant (3). Variables that affect the posttransplant quality of life include chronic pain, bone fractures, rejection, chronic rejection, immunosuppression-related side effects, and RPBC.

Other publications have evaluated clinical factors that could predispose a patient to recurrent disease after liver transplantation (recurrent primary sclerosing cholangitis, recurrent autoimmune hepatitis (4), viral hepatitis, etc.) as well as RPBC. None of the factors were found to be significant. Factors such as antibody status, antibody differences between sexes (5), the degrees of human leukocyte antigen (HLA) match and mismatch, presence of bacterial, or viral infections (6) have all been evaluated. The role of immunosuppression has also been evaluated in previous publications (7) but with small study populations.

Our current attempt was to determine those clinical factors (excluding liver biopsy) that could possibly assist us in predicting which patients were more likely to experience histologically proven recurrence. In addition, because of the autoimmune nature of PBC, we assessed the role of immunosuppression in the development of RPBC.


A retrospective analysis of our prospectively maintained liver-transplant database was performed. One thousand eight hundred and thirty-five patients between 1985 and 1999 received liver transplants. Patients were followed up after liver transplant and liver biopsies, and chemistries within the first 12 months were obtained upon clinical suspicion. Antibody titers (anti-smooth muscle and antimitochondrial), chemistries, serologies, Doppler ultrasounds of the allograft, and liver biopsies were obtained at 1, 2, 5, 10, and 15 year posttransplant follow-up times. This is our usual posttransplant protocol for patients without complications. Other liver biopsies were obtained on the basis of clinical indications.

All cases of RPBC were biopsy-proven and read by our transplant pathologists. Diagnosis of RPBC was based on the presence of lymphocytic (nonsuppurative) cholangitis or granulomatous lesions in the absence of bacterial or viral etiologies.

Factors that were statistically analyzed were sex, antibody status (anti-smooth muscle and antimitochondrial), number of acute cellular rejection (ACR) episodes, use of monoclonal antibody, and immunosuppression type (tacrolimus or cyclosporine). Comparisons between groups were analyzed using the two-tailed Fisher’s exact test for two by two tables and the likelihood ratio chi-square test for larger tables. All comparisons were made at the 0.05 level of significance.

Our immunosuppression regimen has evolved over the years. There was a gradual change from cyclosporine-based immunosuppression to tacrolimus-based immunosuppression. Our first clinical experience with tacrolimus began in 1991. In addition, the subsequent introduction of mycophenolate mofetil has also been introduced into our regimen in the past few years. These regimens provided us with historic groupings that enabled our statistical comparison. The immunosuppression groups are listed in Table 1.

Table 1:
Immunosuppression groups in patients transplanted for primary biliary cirrhosis


One hundred sixty-nine patients were transplanted for PBC (Fig. 1). Thirteen patients were excluded because of death within 1 year without a documented liver biopsy or less than 1 year follow-up without liver biopsy. There were 148 females and 8 males. RPBC was seen in 17 (10.9%) patients.

Figure 1:
Transplants for primary biliary cirrhosis (PBC) and recurrences over 15 years.

The median follow-up time after liver transplant was 72.1 (range 12–180.1) months. The median time to histologic diagnosis of recurrence was 49.6 (range 13.2–85.2) months. There was a difference in the time to recurrence between the immunosuppression groups. Groups 1 and 2 had a longer time to recurrence (57.7 and 60.1 months, respectively) when compared with group 3 (24.3 months). The group sizes were too small to allow for statistical testing. The follow-up period after diagnosis had a median time of 11.5 months (range 8 days to 141.6 months).


There were 26 (16.7%) deaths in the 169 patients transplanted for PBC. There were only two (11.7%) deaths in the RPBC group (Fig. 2). Progression to cirrhosis was seen in one patient. The deaths occurred at 4.3 and 7 years posttransplant. Both deaths were caused by sepsis, and one patient also developed cirrhosis. There was no statistically significant difference in patient survival (P =0.15) between the two groups.

Figure 2:
Death in patients transplanted for PBC.


The typical female predominance of PBC was observed. There were 148 females and 8 males transplanted for PBC. Recurrence occurred in 15 females and 2 males (P =0.39). Age comparisons between RPBC patients and PBC patients were not statistically significant (P =0.3115). The mean age at transplant for PBC patients was 52.1±10.4 years compared with RPBC patients at 44.8±9.4 years.

Human Leukocyte Antigen

Our database contained complete HLA (both donor and recipient) data for 75 PBC patients; the data from the earlier patients receiving transplants was found to be partially complete. HLA of both donors and recipients were known for 10 of the 17 patients with RPBC (Table 2).

Table 2:
Human leukocyte antigens (HLA) of recurrent primary biliary cirrhosis patients and their donors

Comparisons of HLA mismatches were performed between donor and recipient for HLA-A, -B, and -DR. There was no difference in 0, 1, or 2 mismatches at any individual site (HLA A P =0.49, B P =0.74, and DR P =0.42). In addition, comparison of combined HLA A+B+DR (0–6 mismatches) did not reveal any significance (P =0.57) (Table 3).

Table 3:
Human leukocyte antigen (HLA) mismatch data in recurrent primary biliary cirrhosis patients

However, HLA allele frequency analysis demonstrated that patients with recurrence did have allele similarities that were significant. Donor alleles A1, B57, B58, DR44, DR57, and DR58 were found more frequently (P <0.05). The only recipient allele of significance was B48 (P =0.01) (Table 4).

Table 4:
Frequent human leukocyte antigen (HLA) alleles in recurrent primary biliary cirrhosis


Only one of the patients experiencing recurrence received a second allograft. The retransplant was performed 6 months after the initial transplant for biliary complications (not RPBC related) and persistent cholangitis despite surgical conversion to Roux-en-Y choledochojejunostomy. In addition, this patient developed de novo hepatitis C posttransplant (conversion to positive polymerase chain reaction). Biopsy-proven RPBC then occurred in the second allograft 27.8 months after retransplantation.

Antibody Status

ASMA and AMA titers were checked at time of pretransplant evaluation and at posttransplant follow-up. Four of the RPBC patients had no detectable anti-smooth muscle antibody (ASMA) or anti-mitochondrial antibody (AMA) posttransplant despite positive antibodies pretransplant. Otherwise, all other patients (in the recurrent and nonrecurrent groups) had positive antibody status posttransplant.


ACR was clinically suspected with rising serum total bilirubin, transaminases, alkaline phosphatase, and gamma glutamyl transferase numbers. ACR was diagnosed through liver biopsy. Rejection was confirmed histologically by the presence of polymorphous portal infiltrates, bile ductule damage, and endotheliitis. The mean number of rejection episodes in patients with RPBC was 1.3. Nonrecurrent patients experienced 1.0 rejections (P =0.34). In addition, there was no significant difference in the incidence of ACR between cyclosporine and tacrolimus treatment.

Antibody Use

The main indication of OKT3 use was for steroid-resistant rejection. Induction therapy is unusual and was only given for the rare patient who was anuric for more than 48 hours posttransplant. OKT3 was used in 41% (7/17) of RPBC patients before recurrence. Thirty-three of 139 (23.7%) individuals with non-RPBC received OKT3. All patients received only one course of OKT3 (maximum 70 mg total). There was no significant difference in development of RPBC between the groups (P =0.36). Also, although there appears to be anincreased incidence of steroid resistant rejection in the tacrolimus group versus cyclosporine (5 vs. 2 patients), this did not achieve statistical significance.


Group 1 (n=71) demonstrated six (8.4%) recurrences, group 2 (n=49) demonstrated five (10.2%) recurrences, and group 3 (n=36) had six (16.7%) recurrences. A comparison of the groups did not reveal any statistical difference. Comparison of the cyclosporine groups against the tacrolimus group also was not statistically significant (P =0.11).

The median time to recurrences for groups 1 and 2 was 57.7 and 60.1 months, respectively. Group 3 (tacrolimus) had a median recurrence time of 24.3 months. This did not reach statistical significance (P =0.1) (Table 5).

Table 5:
Groupings of recurrences by immunosuppression, OKT3, and median recurrences


Publications focusing on recurrent disease after liver transplant are difficult to compare. Disparity in study methods, study group sizes, and intrinsic institutional differences in patient management and follow-up are just a few of the factors involved. In any event, RPBC is becoming an accepted entity after liver transplantation, and its impact on long-term graft and patient survival is coming to the forefront. The data we present constitutes our single institution experience.

Recurrence has been reported to be low. Neuberger (8), reports that by 10 years, 30% to 50% of patients will develop recurrence. This is assuming a gradual progression. Our results did not support this finding. We found 10.9% of our patients developed RPBC, which was not related to length of time posttransplant. Our patients that received transplants more than 10 years ago did not have a gradual increase in recurrence. It is possible that this phenomenon may be related to the immunosuppression regimen that they are on.

Unlike previous smaller reviews, our study presents one of the largest single-center populations of patients transplanted for PBC with extensive follow-up periods. Despite our numbers, because of the indolent nature of RPBC, we are still left with a small study population that is subject to sampling error. In any event, our attempts to identify significant factors were consistent with previous studies. There were no clinical factors that were found to be statistically significant. Episodes of ACR were not statistically different between recurrent and nonrecurrent groups, but the use of OKT3 (which is reserved for steroid-resistant rejection) was higher in the RPBC patients. This may suggest an immunologic predilection for recurrence, although we do not have the data to support this conjecture at this time.

The HLA data in our database, although only partially complete, was complete enough to allow for statistical testing. On the basis of the data that we obtained, donor alleles A1, B57, B58, DR44, DR57, and DR58 were found at an increased frequency. The only recipient allele of significance was B48. In addition, the degrees of HLA match and mismatch were not significant in recurrence. These findings are interesting; however, the population size of RPBC patients in our experience did not allow us to make a conclusive statement regarding these findings. It does allow us to continue to imply an immunologic basis for the recurrence on the basis of the allele frequency findings. Additional follow-up and inclusion of more RPBC patients will be needed to validate the these statements.

Our data do suggest, however, that the incidence of RPBC was noticeably higher in our tacrolimus population, which is consistent with previous reports. One factor that may have caused the lower incidence of pathologic diagnosis of RPBC in the cyclosporine population was that recurrence was not an accepted entity in the earlier era of liver transplantation. This obviously would decrease the number of recurrences and corresponding statistics during the cyclosporine (early) era of this review. Again, an additional contributing factor may also be the small study population of RPBC patients. In any event, the median time to developing RPBC was shorter with tacrolimus when compared with cyclosporine. This may be responsible for the increasing numbers of RPBC seen in the last 6 years (Fig. 1) rather than recurrence developing in individuals receiving transplants in the long-term. The mechanism for this recurrence is not known, and it may involve subtle immunosuppressive effects, or lack thereof, on the disease itself. This does raise the question as to whether cyclosporine should be the immunosuppressive agent of choice in PBC. The difference between cyclosporine and tacrolimus did not achieve statistical significance, and future controlled clinical trials should be established to investigate this possibility. In addition, reinspection of the microscopic slides with additional staining should be undertaken to determine the correctness of the pathologic diagnoses made in the earlier era to confirm whether a difference exists.

A consideration is that the level of immunosuppression may have been less in the recent era, allowing for more rejection and recurrences to occur. In the recent era of tacrolimus use, our institutional protocol has remained the same. The doses of tacrolimus are adjusted to keep trough levels at 10 to 20 ng/mL in the first 3 months and from 5 to 10 ng/mL at 3 months and beyond. This is, of course, patient specific, and if these trough levels cannot be achieved because of nephrotoxicity, a third agent such as mycophenolate mofetil is added to provide additional immunosuppression. In addition, our protocol dictates that our PBC patients are to remain on steroids for at least 1.5 years posttransplantation. Finally, there is some reluctance by our group to discontinue steroids in these patients. Therefore, we do not feel that, in the current era, these patients are getting less immunosuppression than those in the early (cyclosporine) era.

Concurrently, almost half of our cyclosporine patients had received azathioprine and the other half mycophenolate mofetil. The comparison of groups 1 and 2 did not reveal a significant difference in development of recurrence and the amount of time to recurrence. Therefore, it appears that these antimetabolite immunosuppressive drugs do not influence recurrence.

The role of immunosuppressive drugs and their impact on RPBC are still to be validated. It appears by this initial retrospective review that the possibility of immunosuppression altering the clinical course of RPBC has been identified. To confirm this, future controlled randomized studies must be undertaken. If future studies do demonstrate superiority of cyclosporine over tacrolimus in RPBC, then an important follow-up clinical inquiry should be begun. The question of switching patients on tacrolimus to cyclosporine would then need to be addressed.

Unfortunately, recurrence at this time can only be recognized through liver biopsy (9). Other studies do not demonstrate adequate noninvasive data identifying risk factors for development (10). As with the primary disease, the chemistries often do not match the histologic appearance of the biopsy. However, early recurrence may be missed because of the inherent sampling variation occurring with liver biopsies. Therefore, liver biopsies may be controversial in diagnosing early RPBC (11). Early lesions are focal within the liver and segmental within the duct system. They might not be sampled by liver biopsy needles, and, in a proportion of patients with a clinical diagnosis of PBC, histology reveals a nonspecific portal hepatitis with variable ductopenia (12). In addition, overlapping histologic features may also be identified with the same liver biopsy. Sometimes it can be difficult to distinguish RPBC against viral hepatitis or chronic rejection. Special stains such as antibody to cytokeratin-7 (13,14) and antibody to C355.1 (15) may be needed and can assist in making the distinction. These stains were not used in our study at this time; however, a retrospective approach will be undertaken to review slides that require further evaluation for more definitive diagnoses. This may increase our incidence of recurrence. Subsequent reports will be published.

Currently, recurrence is best discovered on liver biopsy at long-term follow-up. The liver biopsy data evaluated by Sebagh et al. (9), who found plasma cell presence in the portal infiltrate at 1-year posttransplant, seemed to be an early marker of recurrence of PBC. We too will be looking retrospectively at liver biopsies, specifically for any evidence of preliminary characteristic changes that may predict RPBC. This will be published in a follow-up study from our institution.

Once this distinguishing feature is noticed on biopsy, the clinician should refocus on the future management of the affected patient. At our institution, we do not have an established protocol for RPBC. Recently, we have been reinstituting ursodeoxycholic acid at time of diagnosis of RPBC. Ursodeoxycholic acid is the only medication that is shown to be beneficial in PBC patients (16,17). The effects of this medication have been shown to delay or even prevent the progression of disease. Presumably, the effect should be the same on RPBC. Future trials should focus on the role of ursodeoxycholic acid in RPBC and on the continued use of ursodeoxycholic acid immediately after liver transplantation for PBC.

Patients with PBC have been shown to have an excellent prognosis after liver transplantation. As with other chronic liver diseases, transplantation must be performed before the patient’s clinical course deteriorates rapidly (18,19). After transplantation, the natural progression of recurrent disease comes into perspective because of longer graft and patient survival. As in the primary disease, recurrent disease progression is slow and is an unlikely cause for medium-term graft loss. The clinical course has changed in that the appearance of recurrence is occurring in a shorter period of time when compared with the earlier era of liver transplantation. Retardation of the disease process should become the focus with these patients. Again, we have demonstrated that clinical factors are not reliable in predicting which patients are at risk for RPBC. Liver biopsy is currently the single mode of detection in RPBC. Therefore, the focus should be placed on continual attempts at delaying or preventing recurrence with the continued use of ursodeoxycholic acid and possibly a cyclosporine-based regimen posttransplantation. Further evaluation is needed to prove the latter hypothesis. All of these patients should undergo protocol biopsies at scheduled intervals for long-term surveillance.


1. Strassburg CP, Manns MP. Autoimmune tests in primary biliary cirrhosis. Baillieres Best Pract Res Clin Gastroenterol 2000; 14: 585.
2. Nakanuma Y, Tsuneyama K, Sasaki M, et al. Destruction of bile ducts in primary biliary cirrhosis. Baillieres Best Pract Res Clin Gastroenterol 2000; 14: 549.
3. Gross CR, Malinchoc M, Kim WR, et al. Quality of life before and after liver transplantation for cholestatic liver disease. Hepatology 1999; 29: 356.
4. Neuberger J. Recurrence of primary biliary cirrhosis, primary sclerosing cholangitis, and autoimmune hepatitis. Liver Transpl Surg 1995; 1 (5 Suppl 1): 109.
5. Nalbandian G, Van de Water J, Gish R, et al. Is there a serological difference between men and women with primary biliary cirrhosis? Am J Gastroenterol 1999; 94: 2482.
6. Tanaka A, Prindiville TP, Gish R, et al. Are infectious agents involved in primary biliary cirrhosis? A PCR approach. J Hepatol 1999; 31: 664.
7. Dmitrewski J, Hubscher SG, Mayer AD, et al. Recurrence of primary biliary cirrhosis in the liver allograft: the effect of immunosuppression. J Hepatol 1996; 24: 253.
8. Neuberger J. Recurrent primary biliary cirrhosis. Baillieres Best Pract Res Clin Gastroenterol 2000; 14: 669.
9. Sebagh M, Farges O, Dubel L, et al. Histological features predictive of recurrence of primary biliary cirrhosis after liver transplantation. Transplantation 1998; 65: 1328.
10. Haagsma EB. Clinical relevance of recurrence of primary biliary cirrhosis after liver transplantation. Eur J Gastroenterol Hepatol 1999; 11: 639.
11. Slapak GI, Saxena R, Portmann B, et al. Graft and systemic disease in long-term survivors of liver transplantation. Hepatology 1997; 25: 195.
12. Portmann BC. Recurrence of primary biliary cirrhosis after transplantation. The pathologist’s view. Eur J Gastroenterol Hepatol 1999; 11: 633.
13. Yabushita K, Yamamoto K, Ibuki N, et al. Aberrant expression of cytokeratin 7 as a histological marker of progression in primary biliary cirrhosis. Liver 2001; 21: 50.
14. Rubio CA. Qualitative and quantitative differences between bile ducts in chronic hepatitis and in primary biliary cirrhosis. J Clin Pathol 2000; 53: 765.
15. Van de Water J, Gerson LB, Ferrell LD, et al. Immunohistochemical evidence of disease recurrence after liver transplantation for primary biliary cirrhosis. Hepatology 1996; 24: 1079.
16. Corpechot C, Carrat F, Bonnand AM, et al. The effect of ursodeoxycholic acid therapy on liver fibrosis progression in primary biliary cirrhosis. Hepatology 2000; 32: 1196.
17. Kumar D, Tandon RK. Use of ursodeoxycholic acid in liver diseases. J Gastroenterol Hepatol 2001; 16: 3.
18. Kim WR, Wiesner RH, Therneau TM, et al. Optimal timing of liver transplantation for primary biliary cirrhosis. Hepatology 1998; 28: 33.
19. Wiesner, RH. Liver transplantation for primary biliary cirrhosis and primary sclerosing cholangitis: predicting outcomes with natural history models. Mayo Clin Proc 1998; 73: 575.
© 2003 Lippincott Williams & Wilkins, Inc.