P01.22 Outcomes in NSCLC Patients with Early Termination of Immune Checkpoint Inhibitors Due to Toxicities

Abstract

Immune checkpoint inhibitors (ICIs) have improved outcomes of patients with advanced non-small cell carcinoma (NSCLC). There is no consensus on the maximum duration of ICIs in metastatic disease. Immune related adverse events (IRAEs) with ICIs, although rare, can have serious and life-long consequences. We aim to describe the real-world data on outcomes of NSCLC patients who stopped ICIs earlier than proposed duration due to IRAEs. Patients with stage III or IV NSCLC treated with ICIs at Monash Health (Melbourne, Australia) from 2015 to 2019 were retrospectively reviewed. Patients enrolled to clinical trials were excluded. Baseline clinicopathological factors and anti-cancer treatment details were collected. Immunohistochemistry for PD-L1 staining was performed and scored according to standardised practice. IRAEs and reason for ICI cessation was recorded. Overall survival (OS) from the start of ICI was calculated using Kaplan-Meier method. A total of 148 patients received ICI therapy with 97 male (66%) and 51 female (34%) patients. 91% were smokers and ECOG 0 or 1 was found in 95%. PD-L1 expression was known in 53/148 patients; where 19% were scored as low, 43% as intermediate and 20% as high. Stage at commencement of ICI was III in 36/148 (24%), where 32/36 (89%) received adjuvant durvalumab after curative intent chemoradiotherapy, and metastatic in 112/148 (76%), where 99 (88%) received nivolumab and 13 (12%) received pembrolizumab, after first line chemotherapy. Median doses of ICIs were 17 in stage III, and 7 in stage IV cohorts. ICIs were ceased in 93/148 (63%) cases due to disease progression, while 38/148 (26%) stopped due to IRAEs; 17 (12%) in stage III and 21 (14%) in stage IV cohort. Severe IRAEs (grade 3 or higher) were pneumonitis (7/148, 5%), colitis (4/148, 3%), endocrinopathy (2/148, 1%), arthritis (2/148, 1%), rash (2/148, 1%) cardiac (2/148, 1%), liver (1/148, 1%), optic neuritis (1/148, 1%), thrombocytopenia (1/148, 1%; grade 5, death). Median OS (mOS) was 29.7 months (range 0.57 – 40.5) in stage III disease and 9.4 months (range 0.37 – 58.1) in stage IV disease. mOS in patients stopping ICI due to disease progression was 8.16 months, while it was 32.43 months in patients stopping earlier due to toxicities, (N/R in patients receiving durvalumab; and 32.43 months in those receiving nivolumab or pembrolizumab). Only 2/112 (2%) patients with stage IV disease completed a minimum of 2 years of ICI treatment, while only 8/36 (22%) completed the recommended 12 months of durvalumab in the stage III cohort. PD-L1 status did not predict survival (p=0.99). Significant number of patients terminated ICIs due to toxicities without progression of their disease. However, this early termination did not result in inferior survival. Exploring strategies involving early stopping of ICIs at maximum tumour response and restarting at progression may have useful clinical value in preventing serious IRAEs.