Is Adjuvant Chemotherapy Beneficial to All Patients With pT3N0M0 Stage Gastric Cancer?

Background: The efficacy and benefits of adjuvant chemotherapy (AC) for patients with gastric cancer pT3N0M0 remain controversial.

Methods: We prospectively collected and retrospectively analyzed 235 patients with pT3N0M0 gastric cancer who underwent radical resection between February 2010 and January 2016. Patients were divided into two groups: the surgery-alone (SA) group (n = 82) and the AC group (n = 153). We analyzed the effects of AC on the overall survival (OS) and recurrence-free survival (RFS), and the relationship between the number of chemotherapy cycles (CC) and recurrence rate (RR).

Results: The 5-year OS and RFS of the participants were 80.9% and 87.7%, respectively, and those in the AC group were significantly higher than those in the SA group (86.9% vs. 69.5%, p = 0.003). The RFS of the AC and SA groups were 88.9% and 85.4%, respectively; the difference was not statistically significant (p = 0.35). The independent risk factors affecting the OS were perineural invasion-positive (PNI+) (HR = 2.64, 95%CI: 1.45-4.82, p = 0.003) and age ≥ 65 years (HR = 2.58, 95%CI: 1.39-4.8, p = 0.003). The independent risk factor affecting the RFS was also PNI+ (HR3.11; 95%CI: 1.48-6.54, p = 0.003). Stratified analysis revealed that postoperative AC can significantly improve the OS of PNI+ patients (AC group versus SA group: 84.1% vs. 45.5%, p = 0.001) and RFS (86.4% vs. 63.6%, p = 0.017). However, perineural invasion negative (PNI-) patients did not show the same results (p = 0.13 and p = 0.48, respectively). According to the number of CC, divided into CC < 3 groups and CC ≥ 3 groups, the cumulative RR in the CC ≥ 3 group of patients with PNI+ was significantly lower than that of the CC < 3 group (7.4% vs. 28.2%, p = 0.037).

Conclusion: For pT3N0M0 gastric cancer patients with PNI+, at least three cycles of postoperative AC can significantly reduce the overall RR. This finding should be verified by using large external sample data.