Search

Patient-reported outcome (PRO) data are critical in understanding treatments from the patient perspective in cancer clinical trials. The potential benefits and methodological approaches to the collection of PRO data after treatment discontinuation (eg, because of progressive disease or unacceptable drug toxicity) are less clear. The purpose of this article is to describe the Food and Drug Administration's Oncology Center of Excellence and the Critical Path Institute cosponsored 2-hour virtual roundtable, held in 2020, to discuss this specific issue. We summarize key points from this discussion with 16 stakeholders representing academia, clinical practice, patients, international regulatory agencies, health technology assessment bodies/payers, industry, and PRO instrument development. Stakeholders recognized that any PRO data collection after treatment discontinuation should have clearly defined objectives to ensure that data can be analyzed and reported. Data collection after discontinuation without a justification for its use wastes patients' time and effort and is unethical. • All roundtable participants agreed that it is important to understand long-term effects, from the patients' perspective, of treatments under investigation in clinical trials, which includes the period after the study treatment has ended. This information is of critical importance for regulatory, payer, clinical, and patient decision making. • To ensure the most effective use of patient-reported outcome (PRO) data collected in the period after study treatment has ended, clinical trial sponsors need to set a research question before designing the collection of the PRO data needed to answer that research question. • The patient advocates who attended this roundtable agreed that patients will complete PRO measures in this period if it is made clear why their input is valuable. [ABSTRACT FROM AUTHOR]

Objective: Although patient-reported symptoms and side effects are increasingly measured in cancer clinical trials, an appropriate assessment frequency has not yet been established. To determine whether differences in assessment frequency affect the apparent incidence and severity of patient-reported symptoms using two well-established patient-reported outcome measures used within the same clinical trial. Methods: We examined patient-reported outcome results from AURA3 (NCT02151981), a randomized open-label study comparing Tagrisso (osimertinib) with platinum-based chemotherapy in patients with previously treated estimated glomerular filtration rate/T790M mutation-positive metastatic non-small cell lung cancer. The outcome of interest was the proportion of patients in each arm that reported worsening of nausea, vomiting, fatigue, diarrhea, constipation, and appetite loss from baseline measured using the patient-reported outcome—common terminology criteria for adverse event (weekly) or European Organization for Research and Treatment of Cancer Quality of Life Questionnaire Core 30 (every 6 weeks). Results: Similar trends were observed for all six symptoms investigated. Using nausea in the chemotherapy arm as an example, 76% of patients reported any worsening from baseline based on weekly patient-reported outcome—common terminology criteria for adverse event assessments. When using an every 6-week assessment of Quality of Life Questionnaire Core 30 nausea and restricting analysis to an every 6-week assessment for patient-reported outcome—common terminology criteria for adverse event nausea, the proportion of chemotherapy arm patients reporting any worsening of nausea was 40% for both measures. Across the six patient-reported symptomatic adverse events, we observed differential proportions when comparing frequent versus sparse assessment. Conclusion: This analysis demonstrates that more frequent assessment of patient-reported symptomatic adverse events will lead to improved detection, and therefore a more complete understanding of the tolerability of experimental anti-cancer therapies. [ABSTRACT FROM AUTHOR]

Objectives: How frequently patient-reported outcome (PRO) data are collected in commercial cancer clinical trials after treatment discontinuation and the quality of that data are poorly understood. We reviewed treatment discontinuation follow-up PRO data collection to learn about trials collecting these data and understand data quality. The review included 4 cancer types representing potential for long- (prostate cancer), medium-/long- (breast cancer), and short-term (pancreatic cancer and hepatocellular carcinoma) follow-up owing to disease trajectory.Methods: We reviewed registration trials in US Food and Drug Administration databases between January 2010 and January 2019. Protocols were reviewed to determine whether PROs were collected and, if so, whether these included the follow-up phase. Clinical study reports were reviewed when follow-up PROs were collected to determine completion rates. Results were summarized using descriptive analyses.Results: Of the 46 trials containing PRO data, 46% had at least 1 follow-up PRO assessment. Follow-up schedules of assessment were wide ranging; the first assessment occurred between 30 days and 6 months after stopping treatment with follow-up for as long as 3 years. PRO completion rates were reported in 57% of 21 trials; at the first follow-up assessment, completion rates for the treatment arm ranged from 38% to 91% and from 41% to 100% in the control arm.Conclusions: The quality of the follow-up PRO data, based on completion rates, was variable, as was the duration of follow-up. A clear research objective should be developed for follow-up PRO data, accounting for patient burden. If PRO data are collected, monitoring should be implemented to improve completion because poor completion limits data use in the benefit-risk assessment. [ABSTRACT FROM AUTHOR]

With the advent of patient-focused drug development, the US Food and Drug Administration (FDA) has redoubled its efforts to review patient-reported outcome (PRO) data in cancer trials submitted as part of a drug's marketing application. This Review aims to characterise the statistical analysis of PRO data from pivotal lung cancer trials submitted to support FDA drug approval between January, 2008, and December, 2017. For each trial and PRO instrument identified, we evaluated prespecified PRO concepts, statistical analysis, missing data and sensitivity analysis, instrument completion, and clinical relevance. Of the 37 pivotal lung cancer trials used to support FDA drug approval, 25 (68%) trials included PRO measures. The most common prespecified PRO concepts were cough, dyspnoea, and chest pain. At the trial level, the most common statistical analyses were descriptive (24 trials [96%]), followed by time-to-event analyses (19 trials [76%]), longitudinal analyses (12 trials [48%]), and basic inferential tests or general linear models (10 trials [40%]). Our findings indicate a wide variation in the analytic techniques and data presentation methods used, with very few trials reporting clear PRO research objectives and sensitivity analyses for PRO results. Our work further supports the need for focused research objectives to justify and to guide the analytic strategy of PROs to facilitate the interpretation of patient experience. [ABSTRACT FROM AUTHOR]