Back to the well: Can patients with advanced non-small cell lung cancer benefit from changing PD-1/PD-L1 inhibitors after progression?

e21194 Background: Immune checkpoint inhibitors (ICIs) have transformed the standard of care of non-small cell lung cancer (NSCLC) and are capable of inducing a sustained response in a cohort of patients. For those who progress while on ICIs and do not have a targetable mutation, options are typically reduced to chemotherapeutic regimens which have a higher probability of toxicity. In patients with a suboptimal performance status or those who are opposed to receiving chemotherapy and are not eligible or interested in clinical trials, treatments are unfortunately limited. Patients who progress on one ICI do not receive a different ICI as part of standard care in subsequent lines of therapy. Data behind switching ICIs, particularly to those with a different mechanism of action (ie anti-PD-1 followed by anti-PD-L1 or vice versa), are lacking. We evaluated the efficacy of receiving a 2nd ICI in patients with NSCLC. Methods: A single-center, IRB-approved retrospective analysis was conducted of NSCLC patients treated with two different ICIs from March 2015 to July 2020. ICIs were given either in combination with chemotherapy or as monotherapy. Patient and tumoral characteristics, including PD-L1 status (if available) and sequence of ICIs, were captured. A positive PD-L1 was defined as PD-L1 expression > 0%. Progression-free survival (PFS) of each ICI (defined as PFS1 and PFS2) were calculated in months (m) for each patient. Median PFS2 was compared between groups stratified by a cutoff median PFS1 of 3m, sequence of PD-1 and PD-L1 inhibitors, and PD-L1 positive and negative subsets. Mood’s median test was used to compare medians. Results: 26 patients were included in the final analysis. 19/26 patients had a PFS1 > 3m. For this cohort, the median PFS2 was 2.5m (Range: 0.1-36.2) compared to median PFS2 of 1.6m (Range: 0.6-15.9) in patients with PFS1 < 3m (p=0.1847). 15 patients received a PD-1 inhibitor followed by a PD-L1 inhibitor. In this subgroup, the median PFS2 was 1.2m (Range: 0.2-36.2). Alternatively, 7 patients received a PD-L1 inhibitor followed by a PD-1 inhibitor and the median PFS2 in this cohort was 2.3m (Range: 0.1-15.9) (p=0.6471). PD-L1 data was available for 19 patients. 8/19 patients with positive PD-L1 had a median PFS2 of 3.6m (Range: 0.2-13.3) compared to median PFS2 of 1.3m (Range: 0.1-36.2) in patients with negative PD-L1 (p=0.2599). Conclusions: Treatment with a 2nd ICI can potentially provide a modest benefit in patients with advanced NSCLC, and some may even experience a very prolonged response. PFS1, sequence of PD-1/PD-L1 inhibitors, and PD-L1 expression did not show any significant correlation with PFS2. Using a 2nd ICI should be considered for advanced NSCLC patients with diminished performance status or limited treatment options.