Abstract African Americans face a higher rate of colorectal cancer (CRC) incidence and mortality compared to other ethnicities. Identifying key prognostic factors is essential for better outcomes in this group. This study focused on the prognostic value of stromal grading in the tumor microenvironment (TME) of African American CRC patients, using both manual grading and digital pathology. A retrospective analysis of CRC cases in African Americans was conducted. Two cohorts were established: a control cohort (patients without recurrence) and a recurrence cohort (patients with recurrence). Stromal grading was performed manually, categorizing tumors into Grade 1 (low risk), Grade 2 (moderate risk), and Grade 3 (high risk), and we also quantified the TME percentages by digital pathology: myxoid, collagen, immune and tumor. In the control cohort (N=40), stromal grading identified 30 Grade 1, 7 Grade 2, and 3 Grade 3 cases. Conversely, the recurrence cohort (N=17) demonstrated a starkly different distribution, with 1 Grade 1, 3 Grade 2, and 13 Grade 3 cases. The Chi-square statistic was notably significant (p<0.0001), indicating a strong association between higher stromal grades and the likelihood of recurrence. By unpaired t-test, high myxoid stroma (p<0.0001), low collagen (p<0.0001), and low tumor (p=0.0002) by digital pathology were associated with recurrence. TME phenotypes are a significant prognostic and recurrence indicator in African American patients with CRC. The marked difference in stromal grades and digital pathology features between the control and recurrence cohorts suggests that higher stromal grades, along with higher myxoid, lower collagen and lower tumor percentages are associated with an increased risk of recurrence. These findings underscore the need for tailored therapeutic strategies and more vigilant surveillance in African American patients exhibiting high-risk TME phenotypes. The efficacy of manual and digital pathology evaluation of the TME also warrants further exploration to enhance prognostic accuracy and treatment personalization. Citation Format: Dongling Wu, Rong Xia, Mouyed Alawad, Lili Lee, Tamas Gonda, Wenqing Cao, Sean Hacking. Tumor microenvironment phenotypes predict recurrence in African Americans diagnosed with colorectal cancer: A comparative study of manual grading with digital pathology [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 6180.
Read full abstract