Development of a novel apoptosis-based tumor regression grade to assess the efficacy of preoperative chemoradiotherapy for rectal cancer: a retrospective single-center study.

Background: Preoperative chemoradiotherapy is used preferably for locally advanced rectal cancer, followed by a watch-and-wait strategy for cases showing clinical complete response. However, there is a discordance between pathological and clinical complete response rates. We aimed to propose a tumor regression grade (TRG) that truly reflects the therapeutic effects of preoperative chemoradiotherapy in locally advanced rectal cancer.

Methods: Overall, 293 consecutive patients with T3/T4a/T4b rectal cancer who underwent chemoradiotherapy followed by radical surgery between Sep 2003 and Dec 2018 were retrospectively reviewed. We assessed apoptosis using M30 cytoDEATH immunostaining and correlated that with conventional TRG (convTRG) evaluated using hematoxylin-eosin staining, and created a new TRG by evaluating apoptosis and convTRG. The modified TRG1-4 (modifTRG) classification was as follows: modifTRG1 comprised poor TRG, modifTRG2 moderate TRG, modifTRG3 good TRG, modifTRG4 complete apoptosis and convTRG3 (pathological complete response). We assessed the overall survival, relapse-free survival, and local recurrence rate.

Results: Pathological complete response rate was 10.6% when evaluated using conventional staining. Using M30 staining, apoptosis was observed in the residual disease in convTRG 1a 0%, convTRG 1b 0.3%, convTRG 2 9.2%. Combining the two, modifTRG4 was observed in 20.1%. The survival rates were similar between modifTRG4 and convTRG3, suggesting that modifTRG4 was equivalent to pathological complete response. However, in multivariate analysis, modifTRG but not convTRG was an independent risk factor for local and distant recurrences.

Conclusion: The proposed modifTRG truly reflected the therapeutic effects of chemoradiotherapy and may be superior to the convTRG to stratify rectal cancer patients treated with chemoradiotherapy.