Patient-Reported Symptomatic Adverse Events (PRO-CTCAE) during Proton and Photon Radiation Therapy: Results of a Prospective Registry of Patient-Reported Outcomes in a Large-Volume, Multi-Site Practice

Abstract

While patient-reported outcomes (PROs) such as quality of life are becoming commonplace, the patient self-reporting of symptomatic adverse events is not commonly implemented. In 2015 we instituted a patient outcomes registry to capture PRO questionnaires, clinician-rated toxicities, and clinical outcomes in a large, multi-site practice. Herein, we evaluate the association between patient demographics and treatment with patient-rated (PRO-CTCAE) and clinician-rated (CTCAE) toxicity, as well as the relationship between PRO-CTCAE and CTCAE. Eligible patients treated by RT with a curative intent entered in our institutional registry had PRO-CTCAE and CTCAE at baseline and post-RT between April 2013 and February 2019. Selected questions from PRO-CTCAE, a library of items for creating patient questionnaires to assess treatment side effects, and Common Terminology Criteria for Adverse Events (CTCAE), used by clinicians to clinically grade patient toxicities, were collected from date of diagnosis throughout follow-up. We implemented a baseline adjusted method for PRO-CTCAE and CTCAE; a PRO-CTCAE score of 3 or 4, or a CTCAE grade of 3 or 4 after baseline that was higher than its score or grade at baseline was considered treatment related toxicity (PTOX or CTOX, respectively). Logistic regression was used to investigate predictors of toxicity. Of 2,684 eligible patients, 2,186 (81.4%) had CTCAE and PRO-CTCAE recorded at baseline and at least one post-RT time point. This cohort had the following disease sites (%): Prostate (40.5), Breast (30.2), Esophagus (6.1), Pancreas (5), Colon (4.2), Head & Neck (3.8), Sarcoma (3.8), Lymphoma (3), GYN (1.8), and Brain (1.6) with respective CTOX and PTOX rates of 4.6/43.7, 2.8/7.3, 27.8/21.1, 7.3/11.8, 19.8/30.4, 21.2/39.8, 12.1/25.3, 4.7/6.1, 23.1/41.0, and 6.3/20.6. Disease site was significantly associated with CTOX and PTOX. Univariate logistic regression identified that T ≥3, N ≥1, and total dose received were significantly associated with higher risks of CTOX and PTOX, while hypo-fractionation was significantly associated with lower risks of CTOX and PTOX. PTOX had a strong relationship with older age in most disease sites, while CTOX was not significantly related to age, regardless of disease site. In general, a higher photon dose was associated with significantly more PTOX, while increased proton dose was not associated with PTOX. This was especially true in prostate as photon dose was considered the most predictive factor for increased PTOX. T, N, and total dose were most common predictors of increased CTOX across disease sites. Across patients, PTOX was associated with CTOX (OR 1.98, 95% CI 1.4-2.8). Patient self-reported toxicity by PRO-CTCAE is significantly associated with clinician-reported toxicity by CTCAE, though patient reports were more sensitive to group differences by age and photon dose. Increased proton therapy doses did not increase patient-reported side effects.