Perioperative chemotherapy with docetaxel plus oxaliplatin and S-1 (DOS) versus oxaliplatin plus S-1 (SOX) for locally advanced gastric or gastro-esophageal junction adenocarcinoma (MATCH): An open-label, randomized, phase 2 study.

4031 Background: The optimal perioperative chemotherapy of locally advanced gastric or gastro-esophageal junction (G/GEJ) cancer has not been established in Asia, though FLOT was considered as standard therapy in Europe. Therefore, we aimed to assess efficacy and safety of DOS versus SOX as perioperative treatment in patients with locally advanced G/GEJ cancer in this study. Methods: This was a single center, open-label, randomized, phase 2 study. Eligible pts were aged ≥18 years with histologically diagnosed HER2 negative cT3-4 Nany M0 G/GEJ (Siewert II/III type) adenocarcinoma. Pts were randomized (1:1) to receive either 4 preoperative and 4 postoperative cycles of DOS (docetaxel 60mg/m2 D1, oxaliplatin 100mg/m2 D1, S-1 40-60mg bid D1-14 depending on body surface area (BSA), Q3W) or SOX (oxaliplatin 130mg/m2 D1, S-1 40-60mg bid D1-14 depending on BSA, Q3W). Pts underwent D2 gastrectomy after preoperative treatment. The primary endpoint was major pathological response (MPR) analyzed in the modified intention to treat (mITT) population according to Becker TRG criteria. Secondary endpoints included the 3-year progression free survival (PFS), the 3-year overall survival (OS), pathological complete response, R0 resection rate and safety. Results: From Aug, 2015, to Dec, 2019, 154 pts were enrolled, 7 pts withdrew consent, 147 pts were included as mITT population (DOS 71; SOX 76). 80.3% and 67.1% of the pts in the DOS and SOX groups accepted surgery respectively. More pts achieved MPR (25.4% vs. 11.8%, p = 0.035) and R0 resection (78.9% vs. 61.8%, p = 0.024) in the DOS group than the SOX group. With a median follow-up of 42.4 months, the 3-year PFS and the 3-year OS were 52.3% vs. 35% (p = 0.065, HR 0.667, 95%CI 0.432-1.029) and 57.5% vs. 49.2% (p = 0.114, HR 0.685, 95%CI 0.429-1.095) in the DOS and SOX groups, respectively. Pts who acquired MPR had a significant longer survival than non-MPR pts. The 3-year PFS were 89.4% vs. 38.6% (p < 0.001, HR 0.076, 95%CI 0.018-0.314). The 3-year OS were 100% vs. 50.1% (p = 0.008, HR 0.024, 95%CI 0.002-0.371). 56.1% and 60.8% of the pts completed at least 6 cycles of the perioperative chemotherapy in the DOS and SOX groups, respectively (p = 0.625). The most common grade ≥3 TRAEs included neutropenia (8.5% vs.10.5%), leucopenia (1.4% vs. 5.3%), thrombocytopenia (1.4% vs 15.8%, p = 0.002), anemia (1.4% vs. 3.9%) and diarrhea (1.4% vs. 2.6%). The incidence of grade 1-2 thrombocytopenia was also significant lower in the DOS group (15.5% vs. 31.6%, p = 0.022). No treatment-related deaths occurred. Conclusions: Perioperative DOS improved MPR significantly and tended to have better PFS compared with SOX in locally advanced G/GEJ cancer, might be regarded as a prefered option of perioperative chemotherapy. Clinical trial information: NCT02725424.