Aim: Independent risk factors associated with patient survival when receiving an additional line of therapy (LOT) were evaluated.

Methods: The US Oncology Network of community practices electronic health record database was retrospectively analyzed for mNSCLC patients receiving a second LOT (2L) between 3/1/10 and 12/31/12, with follow-up through 10/31/14. Patients who participated in a clinical trial, received tyrosine kinase inhibitors as first LOT, or had concurrent cancer were excluded. Information on previous therapy, clinical and demographic characteristics of patients, geography, and practice size was collected. Patients without known death dates were censored at last contact date.

Results: Of 2122 patients receiving ≥2L for mNSCLC, 963 received third (3L) and 319 received ≥4 LOT (4L). Median age: 67 years (34–94); 58% male; 54% non-squamous, 25% squamous, and 21% non-specified. Median overall survival (mOS) from start of 2L, 3L, and 4L was 8.9, 7.0, and 7.2 months, respectively. Patients who received 2L vs ≥3L had mOS 13.4 vs 5.0 months (P<0.0001) and patients receiving 3L vs ≥4L had mOS 12.9 vs 4.9 months (P<0.0001). In univariate analysis, a subsequent LOT was a significant, independent prognostic factor for improved survival for 2L to 3L (hazard ratio [HR]=0.55; P<0.0001) and 3L to 4L (HR=0.49; P<0.0001). Receiving a subsequent LOT remained significant in multivariate Cox proportional hazards regressions for 2L to 3L (HR=0.43; P<0.0001) and 3L to 4L (HR=0.46; P<0.0001). In stepwise regression models, Eastern Cooperative Oncology Group performance status (PS) 2+ and bone metastasis (BM) consistently predicted decreased survival from 2L to 3L (HR=1.65; P<0.0001 for PS 2, HR=1.50; P=0.0008 for BM) and 3L to 4L (HR=1.56; P=0.0039 for PS 2, HR=1.50; P=0.0058 for BM).

Conclusions: Receiving a subsequent LOT is associated with improved survival in mNSCLC. PS and BM were consistently associated with shorter survival for each subsequent LOT.