Early vs. Late Liver Allograft T-Cell Mediated Rejection: A Banff Allograft Pathology Update
Claudia Mescoli et al
Read the full manuscript on SSRN
Abstract
The often smouldering accumulation of alloimmune damage after the initial post-transplant months, categorised as late-TCMR, is difficult to sensitively recognise and monitor because: a) liver injury tests can be insensitive and nonspecific, b) lack of protocol liver biopsy practice delays detection until graft dysfunction manifests and c) late rejection does not usually resemble early-TCMR histologically; differences in tempo and pathogenesis lead to distinct clinical and histologic manifestations. Late- compared with early-TCMR is characterized by more prevalent interface and perivenular inflammatory activity, a lymphocyte-predominant infiltrate and fibrosis, with less conspicuous portal, duct-centered and subendothelial venular inflammation. The traditional Banff RAI scheme that works consistently well to diagnose early-TCMR is less well calibrated to account for these late shifts of inflammatory target and density, risking underdiagnosis of clinically significant late-TCMR. This document lays out the histopathologic characteristics of early- vs. late-TCMR and the rationale for updating the Banff scoring system for liver allograft rejection to better incorporate currently unrepresented histology feature sets correlating with a molecular rejection signature into the Banff rejection scheme.
Join the conversation by providing your comment(s) on the:
Clinical feasibility and utility of the classification
Areas of improvement for clarity and reduction of ambiguity
Inaccuracies and inconsistency
Evidence and knowledge gaps
The comment period will be open until August 31, 2025. All comments will be reviewed and considered by the Banff meeting report author groups prior to finalizing the manuscripts and submission for formal peer review.
How to comment:
Click on the comment box and you will be prompted to create an account
Once your account is created, you will return to the article, please enter your comment and click publish
Make sure to follow the post for updates on new comments and replies
I think consideration should be given to having two different systems for grading late TCMR: one for research/pharma trials and one for clinical practice. This approach was adopted for reporting mesangial hypercellularity in the Oxford System for grading IgA nephropathy.
For research and clinical trials, detailed measurements of arc length and circumferences by morphometric tools or pixel based assessment can certainly be explored further. My personal prediction based on experience with the kidney literature is that only modest benefits will emerge. In one study on assessing fibrosis in the kidney by Dr Robert Colvin (PMID 21115619), the correlation of the VISUAL assessment on duplicate slides was nearly as good as with the MORPHOMETRIC techniques (R2 = 0.62 to 0.90 for visual versus R2 = 0.69 to 0.96 for morphometry). Correlation coefficients between % fibrosis assessed visually vs Sirius red assessment were 0.42 and 0.45 respectively. I believe these disappointing differences reflect the patchy nature of pathology lesions. More accurate mathematical measurements do not help as much as one might expect.
Clinically oriented practicing pathologists need an eyeballing technique similar to the current RAI index that spans from 0-9. They could use the system for grading central venulitis that the Banff group has already published. Similarly, the Knodell scoring system exists for interface activity in viral hepatitis and most pathologists are familiar with this and are using effectively.
I completely agree that RAI is effective for diagnosing early TCMR, but its efficacy decreases in late TCMR. However, the histological features characteristic of late TCMR are very similar to those of early TCMR that persists despite steroid pulse therapy or ATG therapy. Is there any possibility that late-TCMR represent a continuum of early-TCMR? When early and late represent distinct pathophysiology, is it possible to histologically differentiate late TCMR from early TCMR with therapy modification? We, faculties pediatric hospital, feel it is difficult to distinguish between early TCMR from late TCMR through frequent biopsies for early-TCMR.
At our hospital, we rarely see ductopenia in CR, likely because tacrolimus is our primary immunosuppressant. Even in neonatal cases where cyclophosphamide is sometimes used, we have rarely observed subsequent ductopenia. Neonatal cases often present as acute liver failure of unknown cause (fulminant hepatitis). In such cases, if TCMR is suspected and a biopsy is performed early post-transplant, even in early TCMR, central perivenulitis is frequently the main finding, necessitating differentiation from recurrence of the primary disease. Moving forward, when evaluating rejection using the Banff scheme, it may be necessary to consider factors such as age and the donor's/recipient's primary disease.
The recognition that the current Banff Rejection Activity Index (RAI) underestimates late-TCMR—particularly interface and perivenular activity—is well supported by emerging molecular and histopathologic data. The emphasis on protocol biopsies, molecular correlations, and integration of non-invasive tools reflects an important step toward more precise and personalized allograft monitoring.
That said, a few considerations may strengthen the proposed framework:
Reproducibility and Simplicity: Incorporating interface and perivenular inflammation is essential, but highly granular schemes may reduce reproducibility across centers. A simplified composite or tiered scoring system (e.g., global impression plus key modifiers) could maintain sensitivity while ensuring day-to-day usability.
Integration of Non-HLA Allo-immunity: While the review acknowledges the role of non-HLA antibodies (AT1R, MICA, others), their practical integration into diagnostic criteria remains limited. Given their association with microvascular inflammation, plasma cell–rich rejection, and fibrosis progression, explicit inclusion in future Banff formulations would be valuable.
Therapeutic Implications and Risk Stratification: Early- and late-TCMR are biologically distinct. Even mild late-TCMR may accelerate fibrosis in pediatric and young adult recipients with decades of expected graft survival, whereas its clinical weight in older recipients may be different. Prospective studies stratified by age, comorbidity, and alloantibody profile will be critical to define intervention thresholds.
Overall, this initiative represents a major advance in codifying the spectrum of late alloimmune injury. Moving forward, prioritizing reproducible scoring, explicit incorporation of non-HLA alloimmunity, and prospective clinicopathologic correlation will help ensure broad applicability and clinical impact.
Would it be useful to define clear operational thresholds for key histologic features and provide guidance for mixed phenotypes, while also considering how best to integrate non-invasive monitoring tools? Any plans for multicentre validation?