Background: Intrahepatic cholangiocarcinoma (ihCC) is a rare aggressive liver malignancy with rising incidence. For resectable cases, surgery is the only curative approach, but recurrence rates remain high. These challenges highlight the need for personalized, evidence-based clinical decision-making to improve patient outcomes. To address this, we developed the Multiphasic Evidential Decision-making Matrix (MedMax) to support systematic, individualized therapeutic strategies for ihCC. Methods: In this retrospective single-center study, between 2010 and 2020, we assessed the ability of the MedMax matrix to make treatment decisions in 489 consecutive patients with ihCC or suspected ihCC. Patients were divided into two cohorts depending on whether their tumor was operable (surgical cohort, n = 335) or non-operable (non-surgical cohort, n = 154). We assessed the accuracy of diagnostic confirmation and treatment allocation by MedMax and evaluated how the model’s recommendations corresponded to those made by the tumor board. Results: In the surgical cohort, MedMax achieved 100% accuracy in diagnostic confirmation and 97.9% accuracy in treatment allocation. There was 74.3% concordance between the resection type proposed by MedMax and actual extent of resection. This discrepancy was caused by deviations from the preoperative plan based on intraoperative findings, which could not have been predicted preoperatively. In the non-surgical cohort, MedMax again achieved 100% accuracy in diagnostic confirmation and 98.7% accuracy in treatment allocation. All discrepancies between the decisions made by MedMax and those made by the tumor board were attributed complex, high-risk patient profiles. MedMax reliably identified risk factors (such as advanced comorbidities and multifocal disease) in both cohorts. Conclusions: The MedMax matrix can make accurate, reliable and transparent decisions about the diagnosis and treatment of patients with ihCC thanks to its modular, evidence-based approach. It can also stratify and document risks in both surgical and non-surgical settings.
Ramouz et al. (Fri,) studied this question.