EP856 - The T2/T3 Boundary: A Source of Error in Rectal Cancer Clinical Staging

Purpose/Background: Clinical staging of rectal cancer is fundamental in treatment planning, dictating the critical decision between initial surgical resection and neoadjuvant therapy. However, the inherent limitations of current diagnostic modalities lead to discrepancies when clinical staging (cTNM) is compared to the final, definitive pathologic staging (pTNM) obtained after surgical resection. These discrepancies can result in both the unnecessary administration of aggressive pre-operative treatments due to stage overestimation or inadequate initial therapy due to underestimation. This study aims to characterize the frequency, directionality, and specific patterns of discordance between clinical and pathological staging.

Methods/Interventions: A retrospective chart review of the 2020 NCDB was performed. All rectal cancer patients over 18 who underwent surgical resection of their primary site as first line treatment were included in the study. The patient's overall stage as well as T and N stages were compared between clinical and pathological classification. Concordance, defined as the same overall numeric stage on both clinical and pathologic staging, and discordance were determined for each patient.

Results/Outcomes: 9411 total patients met the inclusion criteria. In the evaluation of the overall stage, 22% of patients had discordance between clinical and pathologic staging, with 78% of those being upstaged on their final pathology vs 22% being downstaged. Clinical T-stage was statistically more likely to be discordant than N stage (p < 0.01, CI 6.67-9.92%). Although there was no difference in the rate of upstaging, clinical T stage was more likely to be downstaged compared to N stage (p < 0.01, CI 8.21-10.2%). When looking specifically at the differences among T substages, patients with a clinical T2 were most likely to have discordance than all other substages (p < 0.01). We also noted differences in directionality in T substages: patients with T2 staging were more likely to be upstaged, and T3 patients were more likely to be downstaged (p < 0.01, CI 57.5-64.6%).

Conclusion/Discussion: T stage was more frequently discordant than N stage, primarily due to higher rates of downstaging on pathology. This suggests that aggressive pre-operative therapies based on overestimated T stages may be unnecessary in a subset of patients. Furthermore, this discordance is not uniform, with a significant number of cT2 tumors being upstaged and another significant number of cT3 patients being downstaged. This suggests that the interface between muscularis propria invasion and perirectal tissue invasion (T2/T3 boundary) is a significant point of imaging ambiguity. Future efforts to improve clinical staging should focus intensely on improving the differentiation between these substages, as this represents the most significant area of diagnostic uncertainty and potential treatment error.