Self-evaluation subjective toxicity according to the Common Toxicity Criteria in breast cancer patients undergoing adjuvant chemotherapy: final results of an Italian prospective study

Abstract

Background: For patients ( pts) enrolled in cancer clinical trials, chemotherapy-related side effects (CSE) are collected by an interview and by assigning a grade of severity according to conventional systems like the Common Toxicity Criteria (CTC) for adverse events. This process is time consuming and it is hardly feasible in the clinical practice. We tried to evaluated a CTC v4.02-based, self-administered questionnaire to collect 10 common CSE in pts undergoing standard adjuvant chemotherapy (ACT) for operable breast cancer. Methods: Our questionnaire was administered to 604 pts from 11 Italian sites, after the 1st and the 3rd cycle of ACT. For each item (nausea, vomiting, constipation, anorexia, taste alterations, diarrhea, fatigue, pain, neuropathy and dyspnea) the CTC v4.02 definitions of severity grade were translated into Italian and rephrased into statements. Pts were asked to choose the statement that best represented the worst CSE experienced after ACT. At each time-point, information on CSE was extracted from the medical charts to compare pts versus physicians (MD) reported side effects. Results: Overall 1177 questionnaires were collected, 596 after cycle 1 and 581 after cycle 3 of ACT. A median of 82% of the fields was completely filled-in. 594 and 573 pts-questionnaires had a corresponding MD-questionnaire. The frequency of CSE (any grade) was systematically higher in pts than in MD questionnaires at both time-points, resulting in low concordance rates (concordance regarding the first time-point is summarized in Table 1, Cohen's K statistics are reported in the last column). We found that the magnitude of the discrepancy in the frequency of CSE was linearly correlated with the number of pts enrolled at each specific site. Conclusion: Self-evaluation of adjuvant CSE according to the CTC system is feasible in the clinical practice and potentially time saving. Whether the observed discrepancy is due to pts over- or MD underreporting CSE needs to be clarified, but our results suggest a potential effect of local workloads on this phenomenon. (Table Presented).