Reply to Harsanyi et al. Comment on “Hasan et al. Clinico-Pathological Features and Immunohistochemical Comparison of p16, p53, and Ki-67 Expression in Muscle-Invasive and Non-Muscle-Invasive Conventional Urothelial Bladder Carcinoma. Clin. Pract. 2023, 13, 806–819”

Abstract

Full text

We would like to thank Harsanyi et al. for their thoughtful and constructive commentary on our recent article [1,2]. Their study of a large patient cohort provides valuable evidence regarding the prognostic roles of p53 and Ki-67, and we appreciate their contribution to this important discussion.

We predefined a p53 cut-off of 20% and a Ki-67 cut-off of 13%. Using these thresholds, and within the constraints of our sample size, p53 did not show a significant association with grade or stage, whereas Ki-67 was higher in muscle-invasive than non-muscle-invasive tumors and aligned with histologic grade. Accordingly, our inferences are limited to the a priori thresholds and study design. We also commend the commenters for the scope and rigor of their large cohort and for the valuable findings they report, which meaningfully advance discussion on threshold selection.

The commenters propose higher thresholds (e.g., p53 ≥ 40%, Ki-67 ≥ 30%) to increase specificity for stage and grade stratification (e.g., NMIBC vs. MIBC; low vs. high grade). We agree that altering the cut-off inevitably shifts the specificity–sensitivity balance and that between-study differences in preanalytical handling, analytical methodology (antibody clone/platform, detection system), interpretation criteria (scoring schema and thresholds), and endpoint definitions can yield divergent results even when the same biomarkers are assessed. Because our analysis plan centered on predefined thresholds, any re-analysis at alternative values would be exploratory and should be interpreted as such.

Cut-offs for IHC markers are not universal; they vary with tumor type, cohort characteristics, assay conditions, and the clinical question being addressed. In urothelial carcinoma, Şentürk et al. [3] applied Ki-67 ≥ 13% and p53 ≥ 20% and demonstrated associations with stage and grade in bladder tumors, providing a precedent for thresholds in this range within UC cohorts. In other disease sites, for example, breast cancer, Soliman and Yussif [4] reported using a Ki-67 index of ~15% for risk stratification in defined contexts, underscoring that “optimal” values are context-dependent rather than fixed. Taken together, these examples support our decision to interpret our results strictly within our predefined thresholds while acknowledging that alternative cut-offs, particularly in larger cohorts and under different preanalytical/analytical conditions, may yield different operating characteristics.

We are grateful for Harsanyi et al.’s engagement with our work. We believe that such scientific dialog advances the field, and we fully support future collaborative efforts aimed at standardizing immunohistochemical evaluation in urothelial carcinoma.