Background
Phase III RCT, N=546, node-positive stage II or III gastric cancer (Korea), D2 gastrectomy within 6 weeks. ARTIST-2 evaluated whether adding oxaliplatin to S-1 (SOX) and whether adding radiotherapy to SOX (SOXRT) improves disease-free survival vs S-1 alone in lymph-node-positive patients informed by ARTIST subgroup findings.
Interventions and follow up
Arm A (S-1): S-1 (40–60 mg bid, 4 wk on/2 wk off) × 8 cycles (1 year)
Arm B (SOX): S-1 (40–60 mg bid, 2 wk on/1 wk off) + oxaliplatin 130 mg/m² q3wk × 6 cycles (6 months)
Arm C (SOXRT): SOX × 2 cycles → RT 45 Gy with concurrent S-1 → SOX × 2 cycles
Primary endpoint: Disease-free survival (SOX vs S-1)
mFollow up: 34.4 mo
Arm B (SOX): S-1 (40–60 mg bid, 2 wk on/1 wk off) + oxaliplatin 130 mg/m² q3wk × 6 cycles (6 months)
Arm C (SOXRT): SOX × 2 cycles → RT 45 Gy with concurrent S-1 → SOX × 2 cycles
Primary endpoint: Disease-free survival (SOX vs S-1)
mFollow up: 34.4 mo
Results
3-yr DFS (SOX vs S-1): 67.2% vs 53.6%, HR 0.74, 95% CI 0.56–0.99, P=.043
3-yr DFS (SOXRT vs SOX): No statistically significant improvement (NS)
3-yr OS (SOX vs S-1): Trend favoring SOX; OS data immature at time of report
3-yr DFS (SOXRT vs SOX): No statistically significant improvement (NS)
3-yr OS (SOX vs S-1): Trend favoring SOX; OS data immature at time of report
Adverse events
Hematologic/GI: Grade ≥3 neutropenia and nausea more frequent with SOX vs S-1
Neurologic: Peripheral neuropathy more common with oxaliplatin-containing arms
Tolerability: Discontinuation rate higher in SOXRT vs SOX due to CRT-related toxicity
Neurologic: Peripheral neuropathy more common with oxaliplatin-containing arms
Tolerability: Discontinuation rate higher in SOXRT vs SOX due to CRT-related toxicity
Conclusions
SOX for 6 months significantly improved 3-yr DFS vs S-1 monotherapy for 1 year in node-positive gastric cancer after D2 gastrectomy (67.2% vs 53.6%). Adding adjuvant radiation (SOXRT) to SOX provided no additional DFS benefit, definitively closing the role of postoperative RT in node-positive D2-resected gastric cancer.
Key Limitations
Korean single-country trial; D2 resection and S-1/SOX use are Asian-specific. SOX vs CAPOX comparison not studied; CLASSIC established CAPOX as standard in other Asian countries. OS data immature at primary analysis. Perioperative FLOT has since become preferred in Western centers, making adjuvant-only paradigm less relevant. P=.043 for DFS is modest; confidence interval wide.
Clinical Context
ARTIST-2 establishes SOX over S-1 for node-positive D2-resected gastric cancer and confirms that adjuvant RT adds no benefit after adequate D2 dissection. Perioperative FLOT (FLOT4) or durvalumab+FLOT (MATTERHORN) now preferred in Western resectable gastric cancer. In Asia, SOX or CAPOX adjuvant options both accepted for node-positive disease.