Session 3: The Utility (or not) of Late Liver Allograft Biopsies and Serological Monitoring

These are great questions and timely. My views:

1. Protocol biopsies remain the best opportunity to understand these changes in late-allograft. One personal strong suspicion supported by several anecdotal and other studies is that sinusoidal endothelium and hepatocytes become the primary targets of alloimmunity after the initial stages, where bile duct epithelium and portal vein and hepatic vein endothelial represent the targets (as Banff scoring rightly recognized years ago). When we seen injuries i the late stages, they represent two variants: hepatitic (whether we call it idiopathic post-transplant, plasma cell rich, or isolated central perivenulitis),they share one thing in common: hepatitic injury to hepatocytes with relatively preserved bile duct epithelium, but sometimes with evidence of portal/hepatic vein endothelial injury as well. The second pattern is one of progressive subsinusoidal fibrosis, an indirect evidence of sinusoidal activity, hence the suggestion that sinusoidal endothelial cells becomes a target at some point (many studies especially those linking DSA to these features also support this suggestion). The problem is that while those with hepatitic injury sometimes (clearly not always) reveal a problem with changes in liver enzyme profile, would-be sinusoidal endothelial injuries have no serologic markers, and this likely explains why these patients almost always have normal enzymes and only discovered on protocol biopsies. There is also a subset of hepatitic group that maintains normal enzymes...curiously, despite sometimes quite robust necroinflammation.

2. The future? clearly protocol biopsies are needed as well, as protocol serologies with DSA, preferably times simultaneously with biopsies. The longitudinal access to tissue and blood would clarify several issues: when do hepatocytes/sinusoidal endothelium become preferred targets, if they do? Are there DSA (and autoantibodies) that mark this transition that could be monitored? Are these always related to DSA? Class I? Class II? Both? Neither? and what about not-HLA (??auto...?Allo-) antibodies?

3. If protocol serologies could be agreed upon, then it makes sense for he Banff group to sponsor serum aliquoting each time a protocol is achieved. This brings us to the cell-free DNA question. If sinusoidal endothelial injury is a driver for long-term fibrosis, ALT, AST, ALP, GGT would not identify this process. There is no serologic marker of endothelial injury that I know of, so, maybe utilizing cell-free DNA methods would provide appropriate serologic markers for sinusoidal endothelial but also all-mediated cell injuries (bile duct, hepatocytes, etc.). Asking each program for this is probably not realistic, so I would propose this as a centralized Banff-sponsored study. To be funded by....???

4. Lastly, while hoping to agree on some sort of unified monitoring, we still need to report these changes, uniformly. My suggestion as always is to keep things simple. Hepatitis is hepatitis, be it auto- allo- viral- or any other. Recognizing an alloimmune hepatitis as an entity remains the easiest and most realistic starting point, and reporting this as variants as well did in this paper (Am J Clin Pathol. 2023;159(3):283-292) has many positives to make it work, since we would hardly be introducing new terminologies, only unifying definition. Even if we choose to score interface versus lobular versus perivenular inflammation as separate variables, the existing terminologies could still be easily built into a new system, while keeping it simple and usable (trying not to say user-friendly).

Happy to know what others think but also looking forward excitedly to seeing everyone in Paris.

Dele