Conditional recurrence-free survival after curative resection of esophageal cancer: Integrated analysis of individual patient data from 11 randomized controlled trials.

345 Background: Most recurrences after curative surgery for esophageal cancer occur within two years. However, conventional recurrence-free survival (RFS), calculated from the time of surgery, does not accurately reflect the prognosis of patients who remain recurrence-free during this initial period. The aim of this study was to evaluate conditional RFS and recurrence timing to explore the potential for individualizing follow-up intervals. Methods: Individual patient data (IPD) analysis of phase III randomized controlled trials (RCTs) comparing perioperative treatments for resectable advanced esophageal and gastroesophageal junction cancer was conducted. Conditional RFS (i.e., the probability of recurrence-free surviving y additional years given that a patient has already recurrence-free survival for x years; RFS y |RFS x ) was performed. Results: IPD were available from 10 phase III RCTs (JCOG1109, JCOG9907, JCOG9204, FFCD9901, FFCD9102, SAKK75/08, CROSS, KOK, CMISG1701, NeoRes2), in addition to one phase II RCT (NeoRes), including 2,268 patients who underwent R0 resection (cStage IV, cT1N0 and cT4b excluded). Of these, 1,597 patients had squamous cell carcinoma (SCC), and 664 patients had adenocarcinoma. The 5-year RFS rate (RFS 5 |RFS 0 ) for patients with SCC was 47.9%; however, RFS 5 |RFS 1 , RFS 2 , RFS 3 , RFS 4 gradually increased to 63.0%, 72.5%, 78.2%, and 81.2%, respectively. At baseline (RFS 5 |RFS 0 ), patients with pN-positive disease or pM1 disease had substantially worse 5-year RFS than those with pN0 or pM0 disease. However, among patients who remained recurrence-free for 4 years after surgery (RFS 5 |RFS 4 ), the pattern was reversed, with the advanced groups showing better subsequent 5-year RFS. Among patients who experienced recurrence, those with pN-positive disease showed earlier recurrence patterns, with 58.8% recurring within one year and 81.7% within two years of randomization, compared to 42.8% and 69.9% in the pN0 group. These overall trends were also observed in patients with adenocarcinoma. Conclusions: Conditional RFS improved over time in patients with esophageal cancer, especially in those with advanced pTNM stage. Although patients with more advanced-stage disease are typically monitored more intensively after surgery, these findings suggest that similar follow-up intensity may be appropriate across all stages once a patient has remained recurrence-free for a certain postoperative period. Additionally, these findings may provide valuable insights for guiding survivorship care in patients who remain recurrence-free for several years, even if they were initially diagnosed with advanced disease.