Determining the optimal radiation dose for locally advanced esophageal cancer: A pooled analysis of reconstructed individual patient data from randomized clinical trials

Abstract

Background

The optimal radiation dose of definitive concurrent chemoradiotherapy (dCCRT) for esophageal cancer (EC) has always been a concern in radiation oncology and has remained controversial for several decades, we performed a meta-analysis based on individual patient data (IPD) to explore the optimal dose.

Methods

Randomized clinical trials (RCTs) comparing high-dose radiotherapy (HD-RT,≥59.4 Gy) with standard-dose radiotherapy (SD-RT, 50 Gy/50.4 Gy) were identified. Graphical reconstructive algorithms were employed to extract time-to-event outcomes from Kaplan-Meier curves presented in the original RCTs. Using reconstructed individual patient data, summary overall survival (OS), progression-free survival (PFS) and locoregional progression-free survival (LRPFS) for HD-RT versus SD-RT were recalculated. Hazard Ratios (HRs) of OS, PFS and LRPFS reported were also pooled by the fixed or random effects model.

Results

Six RCTs, including 1722 patients, were included. IPD for OS, PFS, and LRPFS were from 1287, 462, and 722 patients, respectively. Overall, HD-RT had no significant benefits in 3-year OS (RR = 1.00, P = 0.990), 3-year progression-free survival (PFS) (RR = 0.96, P = 0.320) and 3-year locoregional progression-free survival (LRPFS) (RR = 0.88, P = 0.204), compared with SD-RT. Consistent with above results, the pooled HRs of OS, PFS and LRPFS for HD-RT versus SD-RT were 0.99 (P = 0.854), 0.94 (P = 0.628) and 0.91 (P = 0.410), respectively. However, HD-RT had higher grade ≥ 3 treatment-related adverse effects (TRAEs) (OR = 1.26, P = 0.025). Subgroup analyses were also performed based on the RT techniques, histology, size of the RT target, dose-escalation mode, and stage editions. We found that dose escalation, even in subgroups, did not benefit long-term survival but resulted in a higher incidence of grade ≥ 3 TRAEs.

Conclusion

The results provide robust evidence that corroborates current guidelines and supports the clinical practice of employing SD-RT. Additionally, it provides implications for the feasibility of further research into novel drug combinations (e.g., immunotherapy) rather than radiation dose escalation.