692P Discrepancies between clinician toxicity grading and patient-reported symptoms in high-risk early breast cancer: Evidence from the phase III PANTHER study

Background: Despite substantial improvements in breast cancer outcomes with modern systemic therapies, treatment-related adverse effects continue to compromise patient health-related quality of life (HRQoL).Clinician-based toxicity grading may inadequately reflect the symptom burden perceived by patients.We evaluated the level of agreement between clinician-assessed adverse events and patient-reported symptoms during adjuvant chemotherapy in early-stage high-risk breast cancer within a randomized trial. Methods:The phase III PANTHER trial (NCT00798070) compared tailored dose-dense versus standard interval adjuvant epirubicin and cyclophosphamide followed by docetaxel.Adverse events were documented by clinicians using Common Terminology Criteria for Adverse Events (CTCAE) version 3.0, while patient-reported outcomes were collected using the EORTC QLQ-C30 and BR23 questionnaires at midtreatment and treatment completion.Six overlapping symptom domains were analyzed: diarrhea, nausea, vomiting, mucositis, pain, and fatigue.Concordance was assessed using weighted Cohen's kappa statistics at middle and end of treatment.Results: Data from 1566 patients with available CTCAE and HRQoL measurements were included in the analysis.Overall concordance between clinician and patient assessments was consistently low across all evaluated symptoms and timepoints (kappa range: 0.07-0.34).Clinician-assigned CTCAE grades were predominantly lower than patient-reported symptom severity, whereas the opposite occurred infrequently.Fatigue and pain demonstrated the highest rates of this discrepancy, reaching 54% and 38%, respectively.The level of discrepancy between clinician-based scores and patient-reported symptoms increased from mid-treatment to end-oftreatment in both study arms. Conclusions:The observed discordance underscores the limitations of conventional toxicity grading for capturing patient experience.Integrating patient-reported symptom monitoring into oncology workflows may enhance detection of clinically relevant toxicity and improve supportive care strategies.Clinical trial identification: NCT00798070.