OBJECTIVES: The NSCLC-SAQ is a 7-item patient-reported outcome (PRO) instrument covering five symptom concepts of NSCLC: cough, pain, shortness of breath (SOB), fatigue, and appetite. The objective of this study was to assess the psychometric properties and derive MWPC thresholds for the NSCLC-SAQ SOB domain and total score (TS).

METHODS: The analysis was conducted on the ITT population using an interim blinded data from the phase 3 EVOKE-01 trial (NCT05089734) in metastatic NSCLC. Patients completed the NSCLC-SAQ and other PROs prior to dosing at every 21-day cycle. Reliability, validity, and sensitivity to change were assessed. MWPC thresholds for improvement and deterioration were determined using anchor-based analyses supported with distribution-based analyses.

RESULTS: Data from 556 patients were included in the analyses. The NSCLC-SAQ SOB domain and TS exhibited high completion rates, ranging from 76% to 100% across visits for both scores. Test-retest reliability was confirmed with intraclass correlation coefficients exceeding the acceptable threshold of 0.70 (0.80 and 0.81 for the SOB domain and TS, respectively). Internal consistency reliability for the TS was also demonstrated (Cronbach’s alpha =0.74). Both the SOB domain and TS demonstrated convergent/divergent and known-group validity by showing the strengths of correlations with external PROs (EQ-5D, QLQ-C30, and Patient Global Impression of Severity [PGIS]) that were consistent with pre-specified hypotheses and good differentiation of known groups. The ability to detect changes in other relevant measures over time was also shown for the SOB domain and TS, with longitudinal correlations exceeding the acceptable threshold of 0.30. MWPC thresholds of ±1 (on a scale range of 0–4) and ±2 to 4 (on a scale range of 0–20) were proposed for SOB and TS, respectively.

CONCLUSIONS: Both the NSCLC-SAQ SOB domain and TS demonstrated robust psychometric properties, making them suitable for assessing treatment effects in patients with advanced/metastatic NSCLC.