Response Shift-Adjusted Treatment Effect on Health-Related Quality of Life in a Randomized Controlled Trial of Taxane Versus S-1 for Metastatic Breast Cancer: Structural Equation Modeling.

Objective: We investigated the quantification of the response shift-adjusted treatment effect on quality-of-life (QOL) data in a randomized controlled trial of taxane versus S-1 for patients with metastatic breast cancer (SELECT-BC).

Methods: This study was a secondary data analysis of a previously published trial. The response shift-adjusted treatment effect on health-related QOL (HRQOL) data measured by the European Organization for Research and Treatment of Cancer Quality of Life Questionnaire Core 30 (EORTC QLQ-C30) was estimated using structural equation modeling techniques in addition to quantifying the "true" treatment effect. Measurement invariances in the values of the common factor loadings, intercepts, and residual variances between before treatment and at the 3-, 6-, and 12-month visits were considered the response shift effects.

Results: In the taxane group, we observed positive recalibration effects for role functioning and positive reprioritization and negative recalibration effects for emotional functioning. The observed change of -4.56 for role functioning comprised +2.26 response shifts and -6.82 "true" change. The observed change of +9.41 for emotional functioning comprised +12.43 response shifts and -1.17 "true" change. In the S-1 group, we observed positive reprioritization and negative recalibration effects for emotional functioning and positive reprioritization effects for social functioning. The observed change of +10.54 for emotional functioning comprised +10.07 response shifts and +0.47 "true" change. The observed change of +2.43 for social functioning comprised +3.50 response shifts and -1.07 "true" change.

Conclusion: Detailed analysis of the response shift effects will improve the evaluation reliability of observed HRQOL data during clinical trials.