Abstract 3967
Background
The choice of first-line chemotherapy in advanced HER2-negative gastric cancer (AGC) is based on the center's preferences and adverse effects profile. There is neither a standard accepted regimen nor predictive factors for drug response in clinical practice other than HER2 status. The objective is to evaluate whether Lauren's classification influences the efficacy of different chemotherapies and the survival of patients.
Methods
The data come from a national registry of AGC in which 31 Spanish centers participate. Eligibility criteria include the diagnosis of a stomach or gastroesophageal junction adenocarcinoma, HER2 negativity, and the use of two or three drug chemotherapy combinations. We used Cox proportional hazards (PH) regression with treatment-by-histology interaction tests to estimate the therapeutic effects.
Results
The registry contains 1215 HER2-negative tumors that could be analyzed for survival endpoints and 675 evaluable for overall response rate (ORR). Overall, the study failed to confirm a decrease in the ORR in the presence of diffuse component (Mantel-Haenszel, common odds ratio of 0.744 (CI 95%, 0.538-1.030), P = 0.088, nor heterogeneity of response according to histology. However, in the intestinal type, docetaxel-based or anthracycline-based regimens were more active than oxaliplatin-fluoropyrimidine regimens (ORR 62% and 60% vs. 45%, P
Conclusions
In this registry, tumor subtypes based on Lauren's classification predicted survival, and responded differently to chemotherapy. Future clinical trials should stratify estimates of effects based on histology.
Clinical trial identification
Legal entity responsible for the study
The investigators
Funding
None
Disclosure
All authors have declared no conflicts of interest.