Hypofractionated 44 Gy in 20 fractions versus conventional chemoradiation in cervical cancer: 2-year results of the randomized HYPOCx-iRex trial with an exploratory pooled analysis of a single-institute EMBRACE-II cohort.
Abstract
Purpose
To evaluate 2-year late gastrointestinal (GI) toxicity, oncologic outcomes, and cost-effectiveness of hypofractionated concurrent chemoradiation with 44 Gy in 20 fractions (44 Gy/20F; HYPO) versus conventional fractionation (45 Gy/25F; CVRT + EMB) for locally advanced cervical cancer (LACC).
Methods
The primary analysis was the randomized HYPOCx-iRex comparison (HYPO n = 21 vs. CVRT n = 19). A pre-planned pooled analysis additionally incorporated a concurrent, nonrandomized single-institute EMBRACE-II cohort (EMB n = 28; total n = 68). Late GI toxicity was graded by CTCAEv5.0. Dose-toxicity analysis used Cox and Probit regression to identify the effective dose associated with 10% toxicity probability (ED10). The indirect ratio of excess dose volume at 60 Gy EQD2 (iRex60) was evaluated as a brachytherapy (BT) conformity index. Locoregional control (LRC) and overall survival (OS) were estimated by Kaplan-Meier. Incremental cost-effectiveness ratios (ICERs) were calculated from provider and societal perspectives.
Results
Median overall treatment time was 39 versus 46 days in HYPO versus CVRT+EMB (p < 0.001). Late GI toxicity was numerically higher in HYPO (2-year Gr ≥2: 23.8% vs. 10.8%, pLR = 0.168; Gr ≥3: 19.1% vs. 5.3%, pLR = 0.066); iRex60 was identified as an independent predictor in multivariable regression (ED10 = 1.6 and 1.85 for Gr ≥2/Gr ≥3; proposed cut-off iRex60 <1.7). Favorable but nonsignificant trends in LRC (95.2% vs. 77.5%; HR = 0.43, pLR = 0.260) and OS (100% vs. 76.8%; HR = 0.22, pLR = 0.107) were observed. The 44Gy/20F regimen demonstrated dominant cost-effectiveness.
Conclusion
Hypofractionated chemoradiation achieved LRC and OS comparable to conventional fractionation with shorter treatment time and lower cost. The higher late GI toxicity observed with hypofractionation is clinically meaningful without statistical significance. Further optimization of brachytherapy conformity and low-to-intermediate dose constraints is proposed before clinical adoption. The iRex60 threshold and the cost-effectiveness findings are hypothesis-generating and need further validation.