Patients undergoing liver transplantation may have developed circulating anti-HLA antibodies due to previous blood transfusions, a prior transplant, pregnancy or a combination of these. These preformed anti-HLA antibodies are either against mismatched HLA antigens of the liver donor (donor specific antibodies, DSA) or against other HLA antigens (non-donor specific, non-DSA). In general, donor specific anti-HLA antibodies have no impact on posttransplant outcomes (1–10). Exceptions may occur in second transplants and in patients with a high level of anti-HLA Class II (DP, DQ and DR) antibodies (11–32). Anti-HLA antibodies decrease in amount after liver transplantation (33–36) and Class I DSAs generally disappear (33). In contrast, Class II DSAs are more likely to persist following liver transplantation (35). The absence of consequences of DSA in liver transplantation as compared to kidney or heart transplantation is not fully understood. Frequently cited theories for this phenomenon include a very large capillary bed (100x greater in the liver than in kidneys) expressing HLA Class I antigens (37–39), and secretion of soluble HLA Class I and Class II antigens capable of binding donor specific antibodies (38).

The primary focus of previous studies of anti-HLA antibodies in liver transplant recipients has centered on the potential harm caused by DSAs. This was not the focus of our study although data are presented that confirm previous findings on graft survival. Rather, the primary aim of this study was to observe the fate of all anti-HLA antibodies following liver transplantation. We aimed to determine if there are different patterns of clearance of DSAs compared with non-DSAs and if these patterns suggest mechanisms of anti-HLA antibody clearance. To our knowledge, published accounts of a comparison between non-DSA and DSA in liver transplant recipients are absent in the literature. We assessed the fate, at 12 months after transplant, of anti-HLA antibodies present at transplant, assessing both donor specific and non-donor specific antibodies.

Thirty-one patients had 76 DSAs identified in the single antigen bead assays performed at transplant. Twenty-two patients had Class I DSAs (N = 39), and twenty patients had Class II DSAs (N = 37). In the same single antigen bead assays performed at transplant, we identified 103 Class I non-DSAs and 78 Class II non-DSAs for comparison. The non-DSAs chosen were those with the highest MFIs, if there were more than five non-DSAs. Eleven patients had both Class I and Class II DSAs.

DSAs and non-DSAs decreased significantly between transplant and 12 months. However, while DSAs mostly disappeared between transplant and 12 months, non-DSAs did not ( Figure 1 ). The difference between DSA and non-DSA was most striking for Class I antibodies. The percent MFI reduction between transplant and 12 months was significantly greater for the Class I DSAs compared with the non-DSAs ( Figure 2 ). Class II DSAs performed differently from Class I DSAs between transplant and 12 months. While there still was a significant drop in the MFI values between transplant and 12 months, these antibodies mostly persisted at a low level and the percent reduction of Class II DSA was not significantly different from the percent reduction of Class II non-DSA ( Figure 3 ). The percent reduction for Class I DSA was significantly greater than for Class II DSA ( Figure 4 ). Among the 11 patients with both Class I and II DSAs, the percent reduction in Class I was also significantly greater than for Class II.

Although it was not the primary focus of this study to analyze graft survival in relation to the presence of anti-HLA DSAs, we found that graft survival of patients with DSAs at transplant was similar to that of patients without DSAs at transplant ( Figure 5 ). For this analysis, we studied only patients who survived the first year. Twelve of the 136 patients tested for the presence of antibodies at transplant failed before one year. Six patients died with function and six had graft failure. Four patients did not have anti HLA antibodies. Eight patients had anti HLA antibodies but none of these had DSAs.

There are several overviews in the published literature documenting that DSAs tend to disappear after liver transplant (40–43). Using established DSA monitoring assays (44, 45), our study demonstrated a significant difference in the fate of DSAs compared to non-DSAs following liver transplantation. We used patients as their own controls when comparing the reduction in DSA to the reduction in non-DSA. The difference between DSA and non-DSA was most striking for anti-HLA Class I antibodies. There was also a significance difference between Class I and Class II DSAs. Class I DSA was mostly absent at 12 months while Class II DSA persisted, although at lower levels. The differential outcomes of Class I DSAs and Class II DSAs have been previously described (33) but not with the detailed comparison to non-DSAs and the large cohort present in this study.

The mechanisms of reduction in anti-HLA antibodies following liver transplantation are not completely understood and were not pursued as a part of this study. A reduction in non-DSA following liver transplantation is likely due to clearance of antibodies in concert with a natural decrease in the production of antibodies, enhanced by the use of immunosuppression. The further reduction in DSA is likely to be related to the liver allograft itself. While no mechanism has been proven, authors have speculated that the liver’s large capillary bed (21 square meters) contributes to the disappearance or reduction in DSA (37–39). HLA molecules expressed on endothelial cells in this large capillary bed could potentially bind donor specific anti-HLA antibodies. If so, this would have to occur without activating complement or propagating other inflammatory processes that could damage the liver. Another previously mentioned mechanism is the secretion of cell free HLA molecules by the liver that could bind donor specific antibodies in the circulation. The differential handling of DSA versus non-DSA also raises a possibility of tolerance induction. However, anti-Class I DSAs disappear in nearly all patients while tolerance (the ability to remove immunosuppression) occurs in only 10–15% of liver transplant recipients (46).

The reasons for the differential handling of anti-Class II antibodies compared with anti-Class I are not clear. If one implicates the binding and removal of antibodies by liver tissue as the primary mechanism, the absence of expression of Class II on most liver cells could explain the persistence of anti-Class II DSAs.

The strength of this observational study derives primarily from the large number of patients who had high titer donor specific antibodies at the time of transplant and who also had high titer non-donor specific antibodies. This study was only possible because it had been previously determined that DSAs have almost no adverse consequences in liver transplantation (confirmed by our analysis in the present study) and therefore DSAs are not assessed and avoided prior liver transplantation. Indeed, pre transplant crossmatches are only performed in liver transplantation when patients are considered for multi-visceral or other multi-organ transplants that include the liver or a retransplant is considered after graft failure from severe antibody mediated rejection. This is not true for other organs, in which donors with HLA antigens against which a recipient has antibodies are avoided because of their demonstrated adverse effects on outcomes (42).

A weakness of this study is the fact that it is retrospective and observational only. No attempt was made to determine the mechanisms of antibody reduction observed. Binding of antibodies to the endothelia of the large capillary beds of livers would require serial liver biopsies. The presence of secreted HLA molecules in liver transplant recipients and their potential for binding anti-HLA antibodies could be studied in the future. Donor-recipient interactions relative to gender, blood type, and age were also not considered for their impact on DSA and non-DSA expression. The cause of liver failure, while described as a demographic feature, was not analyzed as a predictor of de novo DSA or non-DSA. For instance, those patients with PSC or PBC as an indication for liver transplantation – both of which are considered on the auto-immune spectrum - may have a propensity for forming de novo antibodies.

This was a detailed assessment of DSA and non-DSA expression against both Class I and Class II HLA that explored the change in antibodies at one year from liver transplantation. A more complete picture of this temporal change is crucial as a first step toward understanding how the effects of antibody exposure to the liver graft evolves and should provide direction for future investigations. These efforts are particularly important as the transplant community develops a more nuanced understanding of AMR in liver transplantation.

The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding author/s.

The studies involving humans were approved by Institution Review Board, Oregon Health & Science University. The studies were conducted in accordance with the local legislation and institutional requirements. The ethics committee/institutional review board waived the requirement of written informed consent for participation from the participants or the participants’ legal guardians/next of kin because this was a retrospective study of data collected for clinical reasons and no PHI is included.

DN: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Writing – original draft, Writing – review & editing. CE: Formal analysis, Writing – review & editing. WN: Formal analysis, Writing – review & editing. RE: Data curation, Methodology, Writing – review & editing. CS: Data curation, Formal analysis, Methodology, Writing – review & editing.