Repeated measures analyses of dermatitis symptom evolution in breast cancer patients receiving radiotherapy in a phase 3 randomized trial of mometasone furoate vs placebo (N06C4 [alliance])

Purpose: Radiotherapy-related dermatological toxicities over time have not been well quantified. We examined during and immediately following radiation therapy skin toxicities over time in a randomized study of mometasone furoate vs placebo during breast radiotherapy. Material and methods: Patients with breast cancer undergoing radiotherapy to the breast or chest wall were randomized. Symptoms related to skin toxicity were addressed weekly using provider-reported Common Terminology Criteria for Adverse Events (CTCAE v3.0) and 4 patient-reported outcomes (PRO) surveys. We applied repeated measures and risk analysis methodologies. Results: One hundred seventy-six patients were enrolled. By CTCAE, significant differences favoring mometasone were detected over time in all toxicities except skin striae, atrophy, and infection. Statistically significant differences between arms at baseline but not over time occurred for all Linear Analog Self-Assessment. Statistically significant differences occurred for all symptoms in the temporal profile of symptoms as measured by PRO surveys (all P < .001). Conclusions: The use of longitudinal methods enhanced the ability of PRO tools to detect differences between study arms. Our results strengthened the conclusions of the original report that mometasone reduced acute skin toxicities. PRO surveys can accurately assess patients’ experiences of symptom type and intensity over time and should be included in future clinical trials. Summary: For radiotherapy-related dermatological toxicity, we hypothesized that clinically significant differences over time, if any, can be found by repeated measures. We examined the acute skin toxicities in a randomized study of mometasone vs placebo during breast radiotherapy. For secondary end points, we showed that longitudinal methods enhanced the detection of differences between study arms and strengthened the conclusions from the original report. Frequent patient-reported outcome surveys over time should be included in future clinical trials.