Immunological and Other Late Complications

Clinical Relevance of Serum Non–organ-specific Autoantibodies and de novo Autoimmune Hepatitis

Positive titers of anti-nuclear, anti–smooth muscle, or anti–liver-kidney-microsomal antibodies are reported in up to 70% of recipients by 5 to 10 years after LT, regardless of the primary indication. They have been reported in LT recipients with normal liver biochemistry, with chronic hepatitis, or isolated allograft fibrosis progressing toward cirrhosis.

The occurrence of chronic hepatitis with liver dysfunction (namely, elevation of aminotransferases and/or gamma-glutamyl-transferase and/or bilirubin levels), newly recognized serum non–organ-specific autoantibodies, raised serum immunoglobulin levels (particularly immunoglobulin [Ig]G), and dense plasma cell infiltrates in liver biopsy was described more than two decades ago in children undergoing LT for conditions other than autoimmune disorders, and it was termed de novo autoimmune hepatitis (AIH). Since the initial pediatric report in 1998, several series of de novo AIH were reported in children, with prevalence up to 22% of LT recipients in some series, as well as in adults. The patients usually do not respond well to increases of calcineurin inhibitor (CNI) doses, but rather to predniso(lo)ne, often combined with azathioprine (or mycophenolate mofetil), using the same schedule as for classical AIH. Lifelong maintenance therapy with steroids is recommended and reported to provide excellent graft and patient survival.

Alternative names, such as alloimmune hepatitis, de novo immune hepatitis, plasma cell hepatitis, and graft dysfunction mimicking AIH, have been used mostly in adult series, reflecting the incompletely understood pathogenic mechanisms. Several overlapping pathogenic mechanisms have been proposed. The overall consensus is that autoantibody positivity reflects rather than causes the graft injury, and routine annual monitoring is recommended.

In its latest update in 2016, the Banff Working Group on Liver Allograft Pathology proposed replacing the term de novo AIH with plasma cell–rich rejection and included de novo anti–human leukocyte antigen (HLA) DSAs and antibodies to anti–glutathione S-transferase T1 (GSTT1) in null recipients of GSTT1-positive donors as contributory but not mandatory features for the diagnosis of plasma cell–rich rejection ( Table 27.1 ). In this update, non–organ-specific autoantibodies are no longer included as diagnostic criteria, whereas C4d immunostains for detection of complement components are recommended on all biopsies diagnosed as plasma cell–rich rejection. A recent review argues against the use of this nomenclature in pediatric patients who fulfill the criteria for de novo AIH, including increased serum IgG levels and circulating non–organ-specific autoantibodies, as they respond well to the immunosuppressive regimen used for classical AIH ( Table 27.2 ). The authors suggest that pediatric de novo AIH should be categorized separately from adult de novo AIH/plasma cell hepatitis.

Clinical Significance of Serum Anti–Human Leukocyte Antigen Donor-Specific Antibodies and Antibody-Mediated Rejection

Although the liver allografts have long been considered less susceptible to antibody-mediated rejection (AMR) caused by anti-HLA DSAs compared with cardiac or kidney allografts, an increasing number of publications suggest a pathogenic role for anti-HLA DSAs also in the LT. As reviewed by Del Bello et al., any kind of aggression, such as T-cell–mediated rejection or viral and bacterial infection, may induce production of HLA class I and II antigens by hepatocytes, bile duct epithelium, or endothelial cells within the liver transplants. Donor Kupffer cells and interstitial dendritic cells are gradually replaced during the post-transplant period by recipient accessory cells that express self-major histocompatibility complex molecules, which are then capable of presenting antigens to T-lymphocytes. All these changes could then facilitate the development of acute and chronic rejection and potentially also idiopathic graft hepatitis, graft fibrosis, and unexplained biliary complications.

Studies of serial biopsies in LT recipients have suggested that the increased prevalence of graft fibrosis with time may be related to low-grade AMR. Anti-HLA class 2 DSA (mostly anti-DQ) are found in 50% – 60% of children after LT, most of which are formed de novo . Histopathological lesions most strongly associated with persistent DSA presence include low-grade portal, periportal, and perivenular lymphoplasmacytic inflammation with low-grade interface and perivenular necro-inflammatory activity and noninflammatory fibrosis. Putative chronic AMR lesions are associated with microvascular endothelial C4d deposition. In the study of Miyagawa-Hayashino et al., which included 67 LT recipients who had undergone a protocol liver biopsy 5 to 20 years post-transplantation (median, 11 years), a higher degree of fibrosis associated with diffuse C4d-positive endothelial staining was demonstrated in patients with positive DSAs compared with DSA-negative patients.

Low levels of CNI and noncompliance have been identified as predictive factors for de novo DSA formation. In studies examining children who were weaned off all immunosuppression after parental living donor LT, the prevalence of DSAs was higher in those who failed successful weaning (i.e., nontolerant patients), but persistent de novo DSAs were also found in the tolerant patients. Consequently, the absence of DSAs did not appear as a reliable indicator of expected tolerance when planning weaning of immune suppression.

The currently most sensitive Luminex single-antigen bead assay detects anti-HLA DSAs without the ability to specify potential cytotoxic effects. This makes it difficult to assess which antibodies have an influence on the graft outcome and which are innocuous. Anti-DQ DSAs have been associated with worse allograft outcomes, but severe graft injury may also occur with negative or low DSAs. There are data suggesting that DSA of IgG3 subclass and/or those that bind the first subcomponent of complement complex C1 (C1q) are particularly damaging. One recent study has shown that the ability of DSAs to bind complement components (C3d) correlated with the risk of graft failure in LT recipients. These findings may help to identify patients at the higher risk for allograft loss, but they need to be validated in other cohorts of LT recipients.

In standard clinical practice, routine DSA monitoring may help IS management (e.g., to avoid minimization of immune suppression), but further work is needed to better define the following : (1) DSA-associated tissue injury patterns; (2) antibody characteristics (C1q, median fluorescence intensity [MFI], titer, IgG subclass, etc.); and (3) appropriate immune suppressant levels. Stringent chronic AMR criteria ( Table 27.3 ) will help to avoid overdiagnosis until the entire spectrum of pathomorphological manifestations is appreciated. Treatment of AMR, however, is not well defined; plasmapheresis, rituximab, polyclonal antibodies, or bortezomib have been used with variable results.