Pathological complete response at the para-aortic nodes as a possible surrogate endpoint in gastric cancer surgery with para-aortic node dissection after neoadjuvant chemotherapy

Abstract Purpose Gastric cancer with para-aortic node (PAN) metastasis has a chance to be cured with multidisciplinary treatment of D2 and PAN dissection (PAND) following neoadjuvant chemotherapy (NAC), but its prognosis remains unsatisfactory. To establish a better multidisciplinary treatment, a better surrogate endpoint is needed. The present study focused on a pathological complete response at the PANs alone as a new surrogate endpoint and evaluated its prognostic value. Methods The study examined patients who received radical gastrectomy with D2 and PAND after NAC for gastric cancer with PAN metastasis from 2004 to 2015. The study compared five methods of evaluating the response to NAC: RECIST, clinical disappearance of PANs (cPAN), histological response of the primary tumor defined by Japanese Classification of Gastric Carcinoma (JCGC histological criteria) and Becker's criteria, and pathological disappearance of PANs (pPAN). The efficacy of these methods was compared using the hazard ratio (HR) for death between responders and non-responders. Results Thirty-two patients were analyzed. The respective HR and 5-year overall survival rates of responders and non-responders were 1.316 and 49.1% vs. 60.0% by RECIST, 1.106 and 52.9% vs. 52.5% by cPAN, 0.246 and 71.3% vs. 28.6% by JCGC histological criteria, 0.239 and 76.2% vs. 36.8% by Becker's criteria, and 0.074 and 81.0% vs. 0.0% by pPAN. Conclusions A pathological complete response at the PANs had the lowest HR and clearly differentiated the survival, suggesting it might be a good surrogate endpoint for identifying future candidates for NAC in multidisciplinary treatment for gastric cancer with PAN metastasis.