Impact of pathological complete response on survival in gastric cancer after neoadjuvant chemotherapy: a propensity score matching analysis.

PURPOSE

The survival benefits of neoadjuvant chemotherapy (NAC) for locally advanced gastric cancer (LAGC) patients are inconsistent. This study aims to investigate how different tumor regression grades (TRG) influence the survival gains associated with NAC treatment.

METHODS

This study compared the treatment outcomes of patients who underwent CSC (neoadjuvant chemotherapy - surgery - adjuvant chemotherapy) with those receiving traditional SC (surgery - adjuvant chemotherapy) treatment. Propensity score matching (PSM) was employed to minimize potential biases arising from differences in baseline characteristics and intervention factors between the treatment groups. After PSM, the CSC cohort was stratified according to TRGs, and their survival outcomes were compared to assess the impact of TRGs on survival gains associated with NAC.

RESULTS

Before PSM, a total of 506 patients were enrolled: 291 in the CSC cohort and 215 in the SC cohort. The CSC cohort had a lower 3-year survival rate (3Y-SR) than the SC cohort (64.6% vs. 76%). In the CSC cohort, patients who achieved pathological complete response (pCR, 12.1%, 26/215) demonstrated significantly improved 3Y-SR (95.5%). After PSM, 110 patients were matched in each cohort. The 3Y-SR was similar between the CSC cohort (68.3%) and the SC cohort (63.6%). In the CSC cohort, 12.7% (14/110) of patients achieved pCR. Subgroup analysis revealed that the pCR subgroup (3Y-SR 100%) was the only subgroup within the CSC cohort that maintained significantly improved survival compared to the SC cohort. Better tumor differentiation was the only pre-treatment factor significantly associated with achieving pCR (p < 0.001).

CONCLUSION

In this retrospective study, LAGC patients who achieved pCR after NAC demonstrated significantly better survival outcomes compared to other response groups. The study found tumor differentiation was a potential predictor of pCR.