Quantitative c-MET immunohistochemistry reveals prognostic subgroups in papillary renal cell carcinoma

Background: MET is a key driver gene in papillary renal cell carcinoma (PRCC), yet the prognostic significance of c-MET immunohistochemical (IHC) expression remains unclear. Previous studies have been limited by heterogeneous cohorts containing mixed renal tumor entities and by subjective variability inherent to manual IHC assessment. As c-MET targeted antibody-drug conjugates (e.g., Telisotuzumab vedotin) show therapeutic promise in several cancers, we aimed to quantitatively evaluate c-MET expression in a strictly defined PRCC cohort and to clarify its prognostic value using an image analysis platform. Methods: A total of 505 surgically resected PRCC cases (1989-2023, Asan Medical Center) were reclassified according to the 2022 WHO system after excluding non-PRCC entities using a 19-marker IHC panel. c-MET IHC (SP44, Ventana) was performed in the final 415 PRCC cases and scored both manually (H-score) and by HALO image analysis. Expression levels were dichotomized into high and low by the median value. Statistical analyses included Spearman correlation, and survival analyses for 5-year recurrence-free (RFS) and disease-specific survival (DSS). Results: HALO-based c-MET quantification showed a strong correlation with manual scoring (R = 0.88, P < 0.001), supporting the validity of automated evaluation. HALO scoring demonstrated significantly higher c-MET expression in type 1 compared to type 2 PRCC (median 64.03 vs. 33.26, P < 0.001). Type 2 PRCC exhibited worse RFS and DSS than type 1 (both P < 0.001). In subgroup analyses, type 2 PRCC with low c-MET expression had significantly poorer DSS than type 1 PRCC with either low (P = 0.022) or high (P = 0.029) expression, however type 2 PRCC with low c-MET expression did not show statistical significance compared to type 1 PRCC. For RFS, type 2 PRCC showed consistently worse outcomes than type 1 PRCC regardless of c-MET level. However, c-MET expression was not an independent prognostic factor in univariate or multivariate Cox models. Conclusions: This study's strength lies in its large, rigorously reclassified PRCC cohort and the combined use of manual and automated scoring methods. Importantly, automated HALO scoring provided an objective, reproducible measure of c-MET expression while maintaining excellent concordance with manual evaluation, supporting its utility in standardizing biomarker assessment. Although PRCC subtyping is no longer recommended in the current WHO classification, we found that low c-MET expression strongly correlated with type 2 morphology and poorer DSS, reflecting the fundamentally different underlying molecular pathways between type-specific phenotypes. Also, since type 2 PRCC with high c-MET expression showed survival patterns similar to type 1, quantitative assessment of c-MET expression may refine risk stratification in PRCC and furhter guide patient selection for emerging MET-targeted therapies.