Different Models to Predict the Risk of Recurrent Hepatocellular Carcinoma in the Setting of Liver Transplantation

Abstract

Liver transplantation is the preferred therapeutic option for non-resectable hepatocellular carcinoma in early-stage disease. Taking into account the limited number of donor organs, liver transplantation is restricted to candidates with long-term outcomes comparable to benign indications on the waiting list. Introducing the morphometric Milan criteria as the gold standard for transplant eligibility reduced the recurrence rate. Even with strict patient selection, there is a risk of recurrence of between 8 and 20% in the transplanted liver, and this is of even greater importance when using more expanded criteria and downstaging protocols. Currently, it remains challenging to predict the risk of recurrence and the related prognosis for individual patients. In this review, the recurrence-risk-assessment scores proposed in the literature are discussed. Currently there is no consensus on the optimal model or the implications of risk stratification in clinical practice. The most recent scorings include additional biological markers for tumour behavior, such as alfa-foetoprotein, and the response to locoregional therapies, in addition to the number and diameter of tumoral nodules. The refinement of the prediction of recurrence is important to better inform patients, guide decisions about prioritization and listing and implement individualized surveillance strategies. In the future, this might also provide indications for tailored immunosuppressive therapy or inclusion in trials for adjuvant treatment.

From prediction to prevention: Machine learning revolutionizes hepatocellular carcinoma recurrence monitoring

“…Researchers and clinicians alike recognize the limitations of relying solely on morphologic and molecular analyses[ 15 ]. The quest for increasing predictive accuracy has led the scientific community to explore uncharted territory and harness the transformative power of technology, especially in the fields of imaging, genetics and computational sciences[ 16 ].…”