EORTC defines clinically important symptom thresholds for prostate cancer quality-of-life assessment

Five EORTC QLQ-PR25 scales now have proposed thresholds to identify clinically important urinary, bowel, hormonal and sexual problems in prostate cancer.

KEY POINTS

  • This prospective EORTC Quality of Life Group study enrolled 261 patients with prostate cancer across 10 centers in seven countries. Participants completed the EORTC QLQ-PR25 and additional questions assessing whether symptoms caused limitations in daily life, required help or care, or generated worry.
  • The investigators derived thresholds for clinical importance (TCIs) using receiver operating characteristic analysis against patient-reported anchor criteria. Unlike minimally important differences, which measure changes over time, TCIs identify potentially important symptom burden or functional impairment from an individual's score at a single assessment.
  • Clinically important symptom thresholds were established at 35/100 for urinary symptoms (sensitivity 0.81, specificity 0.80), 13/100 for bowel symptoms (0.80, 0.83), and 14/100 for hormonal treatment-related symptoms (0.88, 0.70). Higher scores indicate greater symptom burden on these scales.
  • The threshold for urinary incontinence aid-related problems was 50/100, with sensitivity 0.62 and specificity 0.98. This high specificity makes a positive result relatively informative, but the limited sensitivity means that some patients with clinically important problems would not be flagged.
  • For sexual functioning, the proposed threshold was 63/100 (sensitivity 0.88, specificity 0.48), with lower scores indicating worse function. The low specificity limits its usefulness as a definitive classifier; the threshold is better suited to prompting further clinical discussion.
  • Five of the six questionnaire scales demonstrated acceptable-to-excellent discrimination, with AUC values of 0.74–0.90. No meaningful threshold could be established for sexual activity (AUC 0.48), indicating that activity level did not reliably distinguish patients reporting clinically important problems.
  • Sensitivity analyses did not identify significant threshold differences by age, treatment intention, current treatment status, or comorbidity. However, thresholds were derived and evaluated in the same 261-patient dataset, and clinical importance was defined using patient-reported anchors rather than an independent clinical gold standard.

CLINICAL TAKEAWAY

These proposed thresholds make EORTC QLQ-PR25 scores more interpretable during prostate cancer follow-up and may support systematic screening for urinary, bowel, hormonal and sexual problems. They are intended to flag individual patients requiring further assessment, not to establish toxicity diagnoses or classify group-level mean scores, and external validation is still required.

SOURCE

Radiotherapy and Oncology

Browse more research Suggest a correction