See Article on Page 1333
Liver transplantation provides the best option for most patients with end-stage liver disease. In children, for whom the primary causes of liver failure are diseases that do not recur in the graft, a full life after transplantation is a possibility. In fact, the actuarial 10-year graft survival rate for pediatric liver transplants in the United States was 62% in 1999, and it is likely to be better for later cohorts.1 A concern, however, is that 2 recent studies noted a high frequency of late graft fibrosis in children.2, 3 Ekong et al.2 reported that fibrosis was present in 61 of 63 children (97%) who were biopsied more than 3 years after transplantation; the majority (62%) had normal liver function tests. Portal inflammation with variable interface hepatitis was present in 44 of the cases (70%), and the inflammatory grade correlated with the fibrosis stage. Scheenstra et al.3 found that 69 children biopsied 10 years after transplantation had fibrosis; 29% had severe fibrosis. However, in that series, the presence of fibrosis was not associated with inflammatory changes.
In this issue of Liver Transplantation, Miyagawa-Hayashino et al.4 address the potential pathogenesis of late graft fibrosis after pediatric liver transplantation, and they suggest that donor-specific anti–human leukocyte antigen antibodies (DSAs) and humoral alloreactivity may contribute to the process. The authors analyzed protocol biopsy samples from 79 stable liver transplant patients more than 5 years after transplantation. At the time of biopsy, a blood sample was drawn for DSA determination. They noted DSAs were against human leukocyte antigen (HLA) class II (30 patients) and rarely against HLA class I. For the biopsy samples, the fibrosis score was graded with the nonalcoholic fatty liver disease staging system; complement component 4d (C4d) staining was also performed. The authors report that the DSA-positive group had a higher degree of fibrosis (mean fibrosis score = 3.1) than the DSA-negative group (mean fibrosis score = 1.1). In addition, severe fibrosis was more common in the DSA-positive group (88%) versus the DSA-negative group (17%; P ≤ 0.001). Diffuse C4d-positive endothelial staining was significantly higher in the DSA-positive group and was associated with higher fibrosis scores.
The importance of a positive crossmatch has been investigated in early liver transplant recipients for many decades, and more recently, the importance of DSAs has also been investigated. It was originally believed that they did not play a role in liver rejection or allograft loss.5 More recently, O'Leary et al.6 evaluated serum samples in 39 adult liver transplant recipients with chronic rejection and analyzed blood samples for DSAs with a single antigen bead test; they reported that 92% of the patients who had chronic rejection had DSAs. However, in that series, 61% of the comparator patients without chronic rejection also had DSAs. Fontana et al.7 noted that 24% of 95 consecutive liver transplant recipients developed HLA antibodies, and 4% developed DSAs. Among the patients with DSAs, 75% had biliary complications, and 66% had biopsy-proven chronic rejection. A recent prospective study followed serum DSA levels in 90 consecutive liver transplant recipients from the baseline to 4 months after transplantation.8 Twenty recipients (22.2%) had preformed DSAs. Seven days after transplantation, the DSA levels decreased markedly in all but 3 patients, in whom there was evidence of complement activation in the liver allograft and C4d deposition without a significant clinical impact in the first year.
Miyagawa-Hayashino et al.4 are the first to suggest that ongoing injury from DSAs may be the cause of the fibrosis seen late after pediatric transplantation. It is important to recognize that they have found an association and not a direct cause and effect relationship. In their series, there was no association between the mean fluorescence intensity for DSAs and the severity of fibrosis; in addition, there were some DSA-negative patients with stage 3 fibrosis and some DSA-positive patients with almost normal histology. An important observation was that the DSA-positive patients had a higher frequency of indeterminate or mild late acute rejection. It may be that both the development of DSAs and graft fibrosis occurred as a result of ongoing allograft injury due to the presence of ongoing inflammation. Others have reported that the severity of fibrosis correlates with the interval from transplantation to biopsy.3 Perhaps the ongoing inflammation results in C4d deposition and progressive fibrosis
There are other potential causes of the development of fibrosis in these grafts. It may be due to subclinical biliary or venous outflow obstructions or to a fibrotic milieu due to age-dependent variations in exposure or susceptibility to substances in the portal circulation (eg, endotoxin) or other environmental factors. Finally, in contrast to adults, grafts need to grow with children. Perhaps this physiological growth in the unique setting of transplant recipients (with chronic immunosuppression) results in fibrosis.9 The progressive increase in both the prevalence and severity of fibrosis in stable pediatric liver recipients after 5 years is a concern; however, whether it will lead to clinical graft dysfunction or significant rates of graft loss in the future is unclear.2, 4 Before any intervention trials are considered, it will be important to show that this late histological finding is associated with worse recipient outcomes.
The authors have made one additional important observation. Four patients successfully weaned from immunosuppression were DSA-negative. The biology of liver transplantation is unique in that a small, select proportion of recipients can be withdrawn from immunosuppression and not develop rejection. At present, there are no reliable markers for predicting which patients can be successfully withdrawn, and it is not a routine clinical practice. Recently, Feng et al.10 also reported a negative association between an increased C4d staining score for liver biopsy and the withdrawal of immunosuppression leading to the successful development of operational tolerance.10 Perhaps future withdrawal trials should assess baseline and longitudinal patterns of DSA and C4d staining to identify associations with tolerance, rejection, or both.
Miyagawa-Hayashino et al.4 need to be congratulated on bringing forward an important observation. Further studies combining serial determinations of DSAs and serial protocol biopsies are necessary to define the specific role of DSAs in the development of late graft fibrosis. The confirmation of these findings will potentially lead to trials to determine whether the progression of late graft fibrosis can be limited by intervention strategies.