OBJECTIVES: Lung cancer remains the leading cause of cancer deaths worldwide. Given its highly symptomatic nature, clinical trials often focused on “composite” endpoints of key, patient-reported symptoms of mNSCLC, including cough, dyspnea, chest pain, arm/shoulder pain, and pain. The analysis of these symptoms has been inconsistently operationalized. The objective of this research was to establish empirical support for the use of individual or composite symptoms in mNSCLC, regardless of treatment.

METHODS: EORTC QLQ-C30 and LC13 data from Phase 2 (GO28625, GO28753, GO28754) and Phase 3 (GO28915, GO29431, GO29436, GO29437, GO29438, GO29537) trials of atezolizumab in mNSCLC was analyzed. Different combinations of symptoms were compared. Spearman correlations were calculated to assess association between items and Kaplan-Meier (K-M) curves of progression-free survival (PFS) by time to confirmed deterioration of symptoms (TTCD; deterioration ≥10 points for ≥2 consecutive timepoints, or deterioration followed by death) to examine the relationship between symptom deterioration and disease progression.

RESULTS: Pooled data from 5573 patients (64.5% male, mean (SD) age 63.2 (9.4) years) showed low-moderate (0.09 - 0.44) Spearman correlations between symptoms at baseline. Only 17.2% (Dyspnea at Rest) to 29.8% (Dyspnea QLQ-C30) of patients had confirmed deterioration (CD) on any individual symptom. Median TTCD for each symptom was 5 or 6 3-week treatment cycles. K-M curves for PFS by TTCD status revealed patients who didn’t experience CD had rapid disease progression; few patients (6.2% Dyspnea at Rest to 10.5% Arm/Shoulder Pain) experienced a late TTCD corresponding to an extended PFS. Similar patterns were observed across individual symptoms as well as K-M curves for various combinations.

CONCLUSIONS: TTCD of pre-specified symptom composites provided no increased efficiency or sensitivity versus individual symptoms in this mNSCLC population, instead individual symptoms should be examined separately. Next steps will explore whether other approaches are more informative when studying symptomatic progression in this setting.