Use of Patient-Reported Outcomes (PROs) to Assess Prognosis, Toxicity, and Response: The Bortezomib Experience in Relapsed and Refractory Multiple Myeloma (MM).

PROs provide patients’ own appraisal of their health status and can be used to measure aspects of health that are generally not captured by ‘traditional’ clinical measures such as physician assessments and test results. This study analyzed PRO data from a multicenter, pivotal phase II trial of bortezomib in 202 patients (pts) with relapsed, refractory MM (

Results: In this population with relapsed, refractory MM, poorer baseline pre-treatment multi-dimensional quality of life (QoL) scores were significantly correlated (range of r= 0.22 to r= 0.77) with fatigue, pain, dyspnea, appetite loss, neurotoxicity, MM disease symptoms, and treatment side effects. Clinical response to bortezomib (Complete Response or Partial Response) was associated with statistically significant (p<0.05) improvements in global QoL, physical, emotional and social functioning, fatigue, dyspnea, constipation, disease symptoms, social support and future perspective, whereas pts with a minimal response, no change or disease progression usually reported stable or worsening scores. Despite concern about peripheral neuropathy (12% for grade 3), FACT/GOG-Ntx change scores were stable in all responder groups. The prognostic value of baseline PRO data relative to clinical information was assessed using Partial Least Square (PLS) regression. First a model predicting death during the study (17 patients died during the study out of 202; 8.42%) identified that at baseline high total serum protein, bone marrow plasma cells and ß2-microglobulin and low platelet count, serum albumin, hemoglobin, Karnofsky score, body surface area and weight were predictive of death. In a separate model, the following baseline PRO scores predicted death during the study: high fatigue, appetite loss, treatment side effects, sleep disturbance, dyspnea and pain scores and low FACIT-Fatigue, physical, role, social, global QoL, cognitive, emotional and FACT/GOG-Ntx scores. The combination of the clinical and PRO PLS components into a hierarchical PLS regression resulted in an increase in the predictive value of the model, suggesting complementary information.

Conclusion: PRO data can serve as independent measures of clinical benefit, toxicity, and even mortality risk, complementing traditional physician assessments and test results. In this trial of relapsed and refractory MM pts, PRO results suggested improvements in QoL for clinical responders, tolerable side effects, and independent prognostic value for mortality.