Changing treatment patterns in patients with stage IV non-small cell lung cancer (NSCLC) from United States community-based oncology practices.

Abstract

13 Background: This study evaluated the shift in treatment patterns in Stage IV NSCLC following the approval of immune-oncology (IO) agents in the US. Methods: A retrospective cohort study was performed using structured data from a US community-based oncology electronic medical record (EMR) database for care received from Jan 2015-May 2018. The study sample included patients with Stage IV NSCLC, ≥18 years of age initiating first-line (1L) treatment with chemotherapy (chemo) or IO agents and classified into 3 groups: chemo alone, IO alone, or chemo+IO. Treatment patterns in 1L and treatment switch patterns in second-line (2L) are reported. A sub-group analysis of patients initiating 1L therapy during the last 6 months of the study period (Dec 2017-May 2018) was conducted to explore changes in 1L treatment patterns in the post-IO approval setting. Chart reviews were done for a subset of patients initiating 1L from Jan 2017-May 2018 to extract information on programmed cell death ligand 1 (PD-L1) testing and evaluate the association of PD-L1 expression levels with receipt of IO therapy. Results: Between Jan 2015-May 2018, 1,969 patients received 1L therapy with a chemo or IO agent. Mean age (SD) was 69.0 (10.1) years, with 44.7% female. The majority of patients (79%, n = 1570) initiated 1L therapy with chemo alone and 21% initiated IO (alone [14%, n = 271] or in combination with chemo [7%, n = 128]). Of 1L patients, 41% (n = 809) were treated with 2L therapy. Of the 1L chemo alone group, 37% (n = 580) received IO in 2L. The use of IO agents in 1L increased from 21% to 48% (n = 147) in the sub-group analysis of 305 patients initiating therapy Dec 2017-May 2018. In the subset of 62 patients whose charts were reviewed, 87% (n = 54) had their tumor tested for PD-L1, of which 37% (n = 20) had high (≥50%) expression values. The majority of high-PD-L1 patients were treated in 1L with IO alone (80%, n = 16), followed by chemo+IO (15%; n = 3), and only 5% (n = 1) received chemo alone. Conclusions: Initially, IO was used as 2L treatment for Stage IV NSCLC, but IO use shifted to 1L setting in the US by the end of 2017. The use of IO therapy alone or with chemotherapy in 1L was more likely in patients with ≥50% PD-L1 expression level.