How, why, and when does primary biliary cirrhosis recur after liver transplantation?

The questions posed in the title of this editorial remain difficult to answer, despite considerable clinical data and much detailed experimental study of the auto immune destruction of the bile ducts of primary biliary cirrhosis (PBC). PBC, primary biliary cirrhosis; BEC, biliary epithelium cell, PDC-E2, pyruvate dehydrogenase complex-E2; AMA, antimitochondrial antibody; UDCA, ursodeoxycholic acid. At least one no longer has to argue over whether the disease does truly recur after transplantation as was the case when the Cambridge/King's Transplant Programme first reported it in 1982.1 The 3 patients described in that New England Journal of Medicine article had been investigated at 3.5 to 4.5 yr after transplantation because of the reappearance of pruritus and jaundice. Mitochondrial antibody was present in high titer in the serum and typical histological changes were found on liver biopsy. Further cases were described in 1989,2 but Starzl et al.,3 with the largest experience of liver transplantation in the world at that time, was far from convinced, as indeed were other centers for some years thereafter. Reappearance of the disease in the “new” liver would provide a wonderful opportunity to learn about the immunological processes involved in the development of the disease. In this respect we may not be much further on, although the pattern and clinical significance of recurrent PBC is now well established. In the latest long-term follow-up study published in this issue of Liver Transplantation, 74 of 154 PBC patients transplanted in the Mayo Clinic had histological evidence of recurrent disease on liver biopsy.4 Comparison with the nonrecurrent disease group showed that survival duration—free of death or transplantation—was not significantly altered when adjusted for age and gender. This is not to deny that a few cases will have progressive disease ultimately requiring retransplantation. Overall cross-sectional rates for recurrence based on the findings in serial protocol biopsies, increased from 18% at 3-5 yr to 23-32% at 10 yr with cumulative incidence rates of 22, 37, and 43% at 5, 10, and 15 yr, respectively.4 It might be anticipated that if the underlying immunological processes of the host are unaffected by removal of the diseased organ, its graft replacement will inevitably become affected. But even if cumulative incidence rates with longer follow-up get nearer to 100%, the severity of recurrent disease overall is significantly less than that of the primary condition. It is clear too that if a liver biopsy is carried out only when clinical features appear, the frequency of recurrence will be considerably underestimated. Currently, we have little knowledge of factors determining disease progression in the posttransplantation situation. The patients in the Mayo Clinic series were significantly older, more often male, had received grafts from older donors, were more likely to be on tacrolimus than cyclosporin and to have had a shorter duration of maintenance corticosteroid. But in multiple logistic regression mode the only significant predictors were tacrolimus immunosuppression and, surprisingly, male gender. That disease recurrence is more likely with tacrolimus as compared with cyclosporin-based immunosuppression was first noted in 2 patients in 1993.5 The 2 cases described had florid histological changes by 12 months after transplantation with return of pruritus and the first appearance of Sjogren's syndrome in 1 case. This observation, confirmed in a major study by Neuberger et al.6 in 2004, rather than indicating an adverse effect of tacrolimus, is more likely to represent a beneficial effect from cyclosporin. Two multicenter controlled treatment trials in the 1990s have shown that cyclosporine has efficacy in the treatment of PBC in relatively low doses.7, 8 We know also that prednisolone can affect histological improvement when used early in the disease. Although tacrolimus has similar immunomodulatory effect on T cell activity as cyclosporin through its interaction with calcineurin, the binding proteins—each with peptidyl-propyl isomerase activity—are different. The fact that maintenance of immunosuppression sufficient to prevent rejection is not sufficient to prevent redevelopment of the disease when the damage to the biliary epithelium cells (BECs) of the small bile ducts—the hallmark of PBC—is due to an autoimmune reaction directed against them, is nevertheless in keeping with the relatively weak effects of immunosuppressive therapy in preventing progression of the native disease. Histological studies have shown a clear relationship between autoreactive CD8+ cytotoxic cells in portal tract infiltrates and the ongoing apoptosis of BECs. There have been several key observations on the immunobiology of BECs that focus attention on the relationship between the immune response and the development of PBC. First, we should emphasize there is increasing evidence that the multilineage immune response to pyruvate dehydrogenase complex-E2 (PDC-E2), the immunodominant autoantigen of PBC, is in fact a pathogenic and effector response. In the past, the immune response to mitochondria focused entirely on the classic serologic antimitochondrial antibody (AMA), when in fact it is the qualitative and quantitative T cell response to mitochondrial antigens that is critical for disease. For example, the number of autoreactive CD4 PDC-E2–specific T cells is more than 100-fold higher in liver and regional lymph node in PBC compared to the number of such cells in peripheral blood.9 Similarly, the number of autoreactive CD8 PDC-E2–specific T cells is 10- to 15-fold higher in liver and regional lymph in PBC compared to the number of such cells in peripheral blood and their frequency is higher in early and more severe disease.10, 11 In contrast, there is no relationship between the level and titer of serologic AMA and the severity of PBC. Further, the sera titer of AMA does not significantly decrease after liver transplantation. Interestingly, high titer sera AMA can be detected in patients with PBC for more than a decade prior to clinical disease and the titer is often extraordinary high, with levels by immunoblot analysis approaching 10−6. It is therefore not surprising that the sera AMA titer does not change after liver transplantation or following immunosuppression.4 We envisage that the sera AMA, perhaps by cross-priming, remains a player in recurrence of disease. These observations, including the recurrence of PBC following liver transplantation, suggest that the BEC becomes a target based upon its unique biologic characteristics and is not an innocent victim. More than a decade ago, based on immunofluorescence data, PDC-E2 was found on the apical surface of biliary epithelial cells only in PBC.12 Subsequently, however, it was demonstrated that the fluorescence was not due solely to PDC-E2 on the apical surface, but rather secondary to an immune complex of immunoglobulin A and PDC-E2 bound to the BEC polyimmunoglobulin receptor.13 Interestingly, when the biliary epithelial cell goes through apoptosis, PDC-E2 comes out intact; this is in sharp contrast to the degradation of PDC-E2 when control cells go through apoptosis.14, 15 We submit, therefore, that disease recurrence in PBC is based upon the presence of the polyimmunoglobulin receptor of the donor BEC and the ability of the serologic AMAs to cross-prime and produce CD8 responses.10 It would be of interest to perform major histocompatibility complex (MHC) typing in patients with recurrent disease. We would argue that patients at greatest risk for recurrence have similar major histocompatibility complex class I alleles, allowing cross-reactivity of their CD8 responses. The complex interactions between the innate immune response with the production of antibodies to PDC-E2 and the initiation of adaptive immunity to the self-antigen involving not only cytotoxic CD8+ cells but also CD4+ cells and the regulatory T cell subset that inhibits such processes, is beyond the scope of this editorial—for excellent reviews, see Refs.16-18. The greatest paradox of all in PBC is how its ubiquitous self-protein PDC present on the inner mitochondrial membrane comes to be expressed on the surface of the BECs. Molecular mimicry, in which bacterial antigens or self-antigen xenobiotically altered by environmental chemical agents show homology with self molecules and trigger the reactivity of T cells, is one explanation for the loss of tolerance to a normal self-antigen. The observation that the PDC-E2 is also highly expressed on the surface of BECs of the graft is a pointer to the same mechanisms being operative in the development of disease in the graft.19, 20 Similar considerations would appear to apply to the use of ursodeoxycholic acid (UDCA) in recurrent PBC as in the nontransplantation situation with favorable affects on biochemical markers of cholestasis and some symptomatic relief of pruritus and fatigue. Whether use of the agent improves long-term survival by slowing down histological progression and the need for transplantation or retransplantation is the subject of conflicting reports. A recently published study from the Barcelona group21 of treatment for the native disease showed that the biochemical response to UDCA at 1 yr (seen in 61% of the series) was associated with similar survival of the patient group to a matched control population. A minor or no biochemical response appeared to be a marker of a more rapid disease progression. In the Mayo Clinic transplant series, 38 of the 52 patients with recurrent PBC received UDCA over a 36-month period.4 Over 50% of these showed normalization of serum alkaline phosphatase and alanine aminotransferase levels compared with 22% of untreated patients. Although no significant difference in the rate of histological progression could be detected between the treated and untreated subgroups, the proportion of individuals with histological progression was significantly lower in those that showed a biochemical response—27% vs. 73% regardless of treatment—consistent with the findings in nontransplantation situation. In the other recent long-term follow-up study of 100 PBC patients after transplantation from the Berlin group,22 in whom a recurrence rate of 32% over a 13-yr period was observed, the patients had been treated with UDCA from the time of the transplantation. Nearly one-half had received primary immunosuppression with tacrolimus, so it is possible that UDCA was delaying or preventing progression of the disease in those cases in which a more rapid course might have been anticipated. Only 2 of the 10 deaths in that series were as a result of recurrent disease. Starting UDCA at the first sign of histological recurrence on biochemical changes has also been shown to have beneficial effects.23 But what determines a response to UDCA remains uncertain. UDCA is known to decrease toxic hydrophobic bile acids, to stabilize cell membrane, and to have immunomodulatory properties. How do these relate to the complex processes of autoimmune reactivity referred to above? Is UDCA response a marker of a distinct subgroup of disease as referred to in a recent review?—“PBC no longer seems to be behave as a single disease but rather appears as a syndrome with varying modes of clinical presentation, responses to medical therapy and prognosis.”24 Whatever the answers, our current repertoire for delaying histological progression and minimizing the impact of recurrent disease, are limited to the use of UDCA started early after transplantation together with a cyclosporine-based immunosuppression regime.