Patterns of Failure in Metastatic NSCLC Treated With First Line Pembrolizumab and Use of Local Therapy in Patients With Oligoprogression

• Published in: Clinical lung cancer 2024-01, Vol.25 (1), p.50-60.e6
• Main Authors: Friedes, Cole, Yegya-Raman, Nikhil, Zhang, Siqi, Iocolano, Michelle, Cohen, Roger B., Aggarwal, Charu, Thompson, Jeffrey C., Marmarelis, Melina E., Levin, William P., Cengel, Keith A., Ciunci, Christine A., Singh, Aditi P., D'Avella, Christopher, Davis, Christiana W., Langer, Corey J., Feigenberg, Steven J.
• Format: Article
• Language: English
• Subjects: Antibodies, Monoclonal, Humanized; Carcinoma, Non-Small-Cell Lung - pathology; Humans; LCT; Lung Neoplasms - pathology; NSCLC; Oligoprogression; Pembrolizumab; Retrospective Studies; SBRT
• ISSN: 1525-7304; 1938-0690
• DOI: 10.1016/j.cllc.2023.09.002

•Two hundred ninety-eight patients with mNSCLC were treated with first line pembrolizumab +/− chemotherapy.•One hundred ninety-eight patients experienced progression; failure most often occurred in distant lesions or existing lesions.•Oligoprogression occurred in 39.9% of all cases of progression.•Treating oligoprogression with RT was associated with longer survival than switching systemic therapy.•A model was constructed to predict which patients may develop oligoprogression. The patterns of failure (POF) for metastatic non–small-cell lung cancer (mNSCLC) treated with immunotherapy are not well established. We conducted a retrospective cohort study of mNSCLC that received first-line pembrolizumab with or without chemotherapy at a single academic center from 2015 to 2021. We defined POF with 2 classifications: 1) local, regional, or distant failure, or 2) failure in existing lesions, new lesions, or a combination. Oligoprogression was defined as disease progression (PD) in ≤3 sites of failure. Overall survival (OS) was measured via Kaplan-Meier and modelled with Cox regression. Of 298 patients identified, 198 had PD. Using POF classification 1, most failures were distant (43.9%) or a combination of locoregional and distant (34.4%). For POF classification 2, failures occurred in a combination of new and existing lesions (45.0%), existing lesions alone (33.3%), or in new lesions only (21.7%). Oligoprogression occurred in 39.9% (n = 79) cases. Median OS was higher in the following: PD in existing lesions vs. new or new + existing lesions (28.7 vs. 20.2 vs. 13.9 months, P < .001) and oligoprogression vs. polyprogression (35.1 vs. 12.2 months, P < .001). In oligoprogression, median OS was better for those who received radiation to all sites of PD (62.2 months) than for those who changed systemic therapy (22.9 months, P = .007). On multivariable analysis, radiation for oligoprogression (HR 0.35, 95% CI: 0.20-0.62, P < .001) was associated with improved OS. In mNSCLC treated with pembrolizumab, oligoprogression is relatively common. Randomized data are needed to define the benefits of radiation in oligoprogressive mNSCLC. The patterns of failure and frequency of oligoprogression for mNSCLC on first-line immunotherapy is not well studied. Oligoprogression occurred in 39.9% of all cases and most disease failure occurred in known sites of tumor. Oligoprogression (vs. polyprogression) and the use of radiotherapy (vs. changing systemic therapies) to treat oligoprogres