| Traditionally, data to describe the association of patient-reported outcome (PRO) and survival endpoints in malignant pleural mesothelioma (MPM) have been examined separately. However, PROs should be associated with established measures such as progression-free survival (PFS) to support the criterion validity framework of the instrument used. Thus far, no known research has used methods to directly link longitudinal PRO assessments with PFS. The purpose of this study was to conduct joint modeling of latent growth curve models (LGCMs) for longitudinal PROs and survival models for PFS to estimate their association in 243 patients with MPM. Post-hoc analyses were conducted on data from a phase III randomized controlled trial of best supportive care (BSC) versus pemetrexed-plus-BSC. PROs were collected using the Lung Cancer Symptom Scale (LCSS). LGCMs were constructed for the nine LCSS items including a treatment covariate PFS was then regressed onto the growth factors of each LCSS item. Statistically significant associations were found between PFS and latent growth factors (intercepts and slopes) of appetite loss, cough, dyspnea, symptom distress, and interference with activity level. PFS was significantly associated with the intercept of pain, and the slope of global quality of life from the LGCMs no growth factors were associated with fatigue or treatment. The joint hemoptysis model did not converge therefore the association between PFS and hemoptysis cannot be assessed. This study has significant social change implications as these techniques will provide further insight into the meaningfulness and significance of PROs in relation to survival, and can also help clinicians and patients make more informed and cost-effective treatment decisions in the evaluation of oncology treatment medication. |
