Abstract
<title>Abstract</title> <p>Colorectal carcinoma (CRC) remains one of the leading causes of cancer-related morbidity and mortality worldwide despite significant advances in screening, molecular diagnostics, and targeted therapies. The density and spatial distribution of CD8⁺ tumour-infiltrating lymphocytes (TILs) are established prognostic markers and form the basis of the Immunoscore. However, routine manual quantification is laborious and subject to interobserver variability. Artificial intelligence (AI)-assisted digital pathology offers a standardised alternative, but validation studies from Indian centres remain limited. Aim: To compare intratumoral (IT) and peritumoral (PT) CD8⁺ TIL density in colorectal carcinoma using manual assessment by two independent pathologists and AI-assisted digital quantification with QuPath. Materials and Methods: This retrospective observational study included 30 treatment-naïve colorectal adenocarcinoma resection specimens. CD8 immunohistochemistry was performed using monoclonal antibody clone C8/144B. Two blinded pathologists independently quantified CD8⁺ TILs in representative IT and PT high-power fields (×400). The same annotated regions were analysed using QuPath version 0.5.x. Statistical analysis was performed using IBM SPSS Statistics version 27.0. Comparisons between manual and AI-assisted counts were made using two-tailed independent-samples t-tests, with p < 0.05 considered statistically significant. Results: The mean IT CD8⁺ TIL counts were 7.73 ± 12.00, 8.33 ± 15.60, and 8.53 ± 11.50 cells/high-power field (HPF) for Pathologist 1, Pathologist 2, and QuPath, respectively. Corresponding PT counts were 37.63 ± 28.40, 36.83 ± 26.10, and 42.27 ± 33.50 cells/HPF. No statistically significant differences were observed between the two pathologists or between manual and AI-assisted quantification (all p > 0.05). Peritumoral CD8⁺ TIL density was consistently higher than intratumoral density across all assessment methods. Conclusion: QuPath-based AI-assisted quantification demonstrated excellent agreement with manual assessment by experienced pathologists. The findings support its use as a reliable, reproducible, and cost-effective adjunct for routine evaluation of CD8⁺ TILs and may facilitate wider implementation of Immunoscore-based prognostic assessment in colorectal carcinoma.</p>