A Pharmacometric Analysis of Patient-Reported Outcomes in Breast Cancer Patients Through Item Response Theory

Pharm Res A Pharmacometric Analysis of Patient-Reported Outcomes in Breast Cancer Patients Through Item Response Theory Emilie Schindler 0 1 Lena E. Friberg 0 1 Bertram L. Lum 0 1 Bei Wang 0 1 Angelica Quartino 0 1 Chunze Li 0 1 Sandhya Girish 0 1 Jin Y. Jin 0 1 Mats O. Karlsson 0 1 0 Department of Clinical Pharmacology, Genentech Inc. South San Francisco, California , USA 1 Department of Pharmaceutical Biosciences , Uppsala University Box 591, SE-75124 Uppsala , Sweden 2 Mats O. Karlsson Purpose An item response theory (IRT) pharmacometric framework is presented to characterize Functional Assessment of Cancer Therapy-Breast (FACT-B) data in locally-advanced or metastatic breast cancer patients treated with adotrastuzumab emtansine (T-DM1) or capecitabine-plus-lapatinib. Methods In the IRT model, four latent well-being variables, based on FACT-B general subscales, were used to describe the physical, social/family, emotional and functional well-being. Each breast cancer subscale item was reassigned to one of the other subscales. Longitudinal changes in FACT-B responses and covariate effects were investigated. Results The IRT model could describe both item-level and subscale-level FACT-B data. Non-Asian patients showed better baseline social/family and functional well-being than Asian patients. Moreover, patients with Eastern Cooperative Oncology Group performance status of 0 had better baseline physical and functional well-being. Well-being was described as initially increasing or decreasing before reaching a steadystate, which varied substantially between patients and subscales. T-DM1 exposure was not related to any of the latent variables. Physical well-being worsening was identified in capecitabine-plus-lapatinib-treated patients, whereas TDM1-treated patients typically stayed stable. Conclusion The developed framework provides a thorough description of FACT-B longitudinal data. It acknowledges the multi-dimensional nature of the questionnaire and allows covariate and exposure effects to be evaluated on responses. ado-trastuzumab emtansine; capecitabine; kadcyla; lapatinib; nonlinear mixed effects; NONMEM; T-DM1 INTRODUCTION In the era of targeted therapies, metastatic cancers are becoming chronic-like diseases, for which a goal of treatments is to maintain functioning and improve quality of life. While traditional outcome measures focusing on overall survival and progression-free survival are essential in cancer decision making, there has been growing evidence that patient-reported outcome (PRO) measures convey additional information for assessing the overall burden of cancer, tolerability, and effectiveness of interventions over long treatment durations, where applicable ( 1,2 ). PRO measures are reports about how a patient feels and functions in relation to a disease and its treatment, that come directly from the patient without interpretation of the response by a clinician or any third party (3). PRO measures are usually collected through questionnaires consisting of several items (questions). While some PRO measures focus on single aspects of health-related quality of life (HRQoL) (e.g. the diarrhoea assessment scale), others are designed to evaluate multiple elements of HRQoL (e.g. the European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire (EORTC QLQ-C30), the Functional Assessment of Cancer Therapy-General (FACT-G)), including disease symptoms, drug-related toxicities, physical functioning, and social and psychological well-being. These aspects of HRQoL are important for multiple stakeholders, including sponsors, regulatory agencies, payers, prescribers and patients. Interestingly, patients have been shown to report symptoms earlier and more frequently than clinicians. However, clinician reports of fatigue, nausea, constipation, and performance status were associated with death and emergency room admissions, whereas no association was identified when patients reported those items. Moreover, patient-reported symptoms were more often in agreement with measures of daily health status than clinician-reported symptoms ( 4,5 ). Multi-item PRO questionnaires are therefore likely to provide a comprehensive picture of a patient’s well-being. Multi-item PRO data are traditionally analysed by calculating the sum of the item scores, and criteria based on expert opinion, such as time to symptom worsening, are then derived. Despite quick computing and easy interpretation, the use of these composite scores results in a loss of information (at both longitudinal and item level) and involves technical challenges, including missing data imputation. Alternatively, itemresponse theory (IRT), used extensively in educational testing applications, has gained in popularity and acceptance in PRO research. IRT can help address these practical problems and provide richer and more accurate description of item-level PRO questionnaire data ( 6,7 ). IRT consists of a statistical fram (...truncated)