OBJECTIVES: An indirect comparison of the health-related Quality of Life (HRQoL) data in metastatic colorectal cancer, from the single arm PRECONNECT (NCT3306394) trial of trifluridine/tipiracil (FTD/TPI) with CORRECT (NCT01103323), a RCT comparing regorafenib (REG) with best supportive care (BSC).

METHODS: The area under the curve (AUC) values from PRECONNECT patients were calculated for QLQ-C30 global health status (GHS) and EQ-5D-3L visual analogue scale (VAS). A match-adjusted indirect comparison was performed using the individual patient AUC data in PRECONNECT compared with the published summary statistics for AUC from CORRECT (Yoshino 2015). Potential treatment-effect modifiers and prognostic factors (age, BMI, ECOG PS, gender, KRAS status, number of metastatic treatment lines and months from diagnosis of metastatic disease to randomisation), were chosen to match the PRECONNECT data with CORRECT treatment summaries.

RESULTS: Baseline values for both PRECONNECT and CORRECT trials were similar. PRECONNECT FTD/TPI mean AUC values (62.6 & 65.3) matched to REG patient characteristics for GHS and VAS, respectively, were higher than the published REG (56.3 & 58.5) mean AUC values in CORRECT. PRECONNECT FTD/TPI mean AUC values (63.0 & 65.8) matched to BSC patient characteristics for GHS and VAS, respectively, were higher than the published BSC (57.7 & 59.5) mean AUC values in CORRECT. Mean differences (FTD/TPI vs REG & BSC) for GHS (6.3 & 5.3) and VAS (6.8 & 6.3) were close to the established minimal important difference of 5 for GHS and 7 for VAS. The lower FTD/TPI 95% Confidence Limit for EQ-5D-3L VAS (63.91 & 64.40) was higher than the upper 95% Confidence Limit of both CORRECT treatment groups, REG (62.5) and BSC (63.7).

CONCLUSIONS: The HRQoL was higher (as measured by EQ-5D-3L VAS) and comparable (as measured by QLQ-C30 GHS) for patients treated with FTD/TPI when compared to BSC, and always higher when compared to REG.