Supplementary materials: Adjusting for treatment crossover in the MAVORIC trial: survival in advanced mycosis fungoides and Sezary syndrome

<b>These are peer-reviewed supplementary materials for the article '</b><b>Adjusting for treatment crossover in the </b><b>MAVORIC trial: survival in advanced </b><b>mycosis fungoides and Sezary syndrome</b><b>' published in the</b><b> </b><b><i>Journal of Comparative Effectiveness Research</i></b><b>.</b><b>A. List of variables available for analysis </b><b>B. Methodology to fit survival curves </b><b>C. IPCW extra detail: Weights </b><b>Figure 1: </b>Histogram of weights from the IPCW analysis<b>Figure 2: </b>Plot of weights over time from the IPCW analysis<b>D. ITT population </b><b>RPSFTM</b><b>IPCW</b><b>TSE</b><b>Background: </b>Relative overall survival (OS) estimates reported in the MAVORIC trial are potentially confounded by a high proportion of patients randomized to vorinostat switching to mogamulizumab; furthermore, vorinostat is not used in clinical practice in the UK. <b>Methods:</b> Three methods were considered for crossover adjustment. Survival post-crossover adjustment was compared with data from the Hospital Episode Statistics (HES) to contextualize estimates. <b>Results:</b> Following adjustment, the OS hazard ratio for mogamulizumab versus vorinostat was 0.42 (95% CI: 0.18, 0.98) using the method considered most appropriate based on an assessment of assumptions and comparison with HES. <b>Conclusions:</b> OS of mogamulizumab relative to vorinostat may be underestimated in MAVORIC due to the presence of crossover. The HES database was used to validate this adjustment.