Exploring Intratumoral Budding in Colorectal Cancer Using Computational Pathology: A Biopsy-Based Evaluation

Abstract

Owing to insufficient evidence, tumor budding (TB) is not currently evaluated in colorectal cancer (CRC) biopsies. This study investigates TB in CRC by establishing the value of intratumoral budding (ITB) in resection specimens and assessing the feasibility and clinical value of TB in biopsies. TB was assessed using an algorithm in all cases. In a test cohort of 555 primarily surgically treated CRC patients, we assessed the prognostic impact of ITB compared with peritumoral budding (PTB). The distribution of ITB in the uppermost 5 mm of resection specimens was analyzed to validate TB counting in biopsies. We further validated the prognostic and predictive impact of TB in biopsies of 285 rectal cancer patients, focusing on overall survival and response to neoadjuvant therapy. High-grade TB, whether intratumoral or peritumoral and in biopsies or resections, was associated with advanced pathological stage, lymphatic invasion, infiltrative tumor border, and poor overall survival in the test cohort. Superficial ITBs (0-3 mm from the lumen) accurately predicted the final TB grade based on PTB in 87% of tumors, with 87% of tumors having at least 1 superficial ITB hotspot. ITB (hazard ratio, 3.5; 95% CI, 1.1-10.8) was an independent predictor of overall survival, unlike PTB. In the validation cohort, TB presence in biopsies significantly reduced the likelihood of achieving a pathological complete response (odds ratio, 0.3; 95% CI, 0.1-0.7; P = .007). ITB is as prognostic as PTB, and evaluating both can improve risk stratification in CRC. TB assessment in biopsies can identify poor prognosis and predict response to neoadjuvant therapy.