Jordan Kardos, Paola Correa, Sam Kim, Min Wang, Yang Pan, Li Li, Scott Donald Patterson, Shahed Iqbal
Background:Non-small cell lung cancer (NSCLC) is the most common type of lung cancer and a leading cause of cancer death. Standard of care (SOC) treatment for advanced NSCLC patients include immune checkpoint inhibitors (ICI) and/or chemotherapy (CT). However, the majority of patients do not respond to SOC therapy. Mechanisms of resistance and how they differ across treatments are not fully understood. We characterized molecular markers of resistance for advanced NSCLC patients with ICI monotherapy, CT, or combination ICI-CT.
Methods:Advanced stage III or IV NSCLC patients with adenocarcinoma histology who received at least 21 days of ICI monotherapy, CT, or combination ICI-CT were identified from a de-identified, multimodal real-world database from Tempus Labs, Inc. All patients had a tissue biopsy taken within 100 days prior to the specified therapy start and all samples underwent next generation sequencing with the Tempus xT assays (DNA-seq of 596-648 genes at 500x coverage and whole transcriptome RNA-seq). Resistance to therapy was defined as having time to next treatment < 90 days, or disease progression or death within 120 days of therapy initiation. We characterized the clinical and molecular characteristics of patients and compared differential expression of genes by resistance status for each cohort.
Results:A total of 365 patients were included in the study (ICI = 121, CT = 122, ICI-CT = 122). Across cohorts, 30% of patients were identified as having resistance to treatment (ICI = 35%, CT = 23%, ICI-CT = 32%). 53% of patients were > 65 years (median [range] = 67 [25- 88]), and 53% were males. Most patients were white (63%) and had a history of tobacco use (85% current or former smoker). At the time of biopsy, most patients were diagnosed with metastatic NSCLC (Stage IV = 73%) and had received no previous lines of therapy (92%). The alteration frequencies of known NSCLC oncogenic driver mutations including TP53 (65%), KRAS (42%), STK11 (14%), KEAP1 (11%), and EGFR (9%) were broadly consistent with previous characterizations of NSCLC. Validated biomarkers of response to ICI treatment such as Tumor Mutation Burden, IFN-γ signaling, PD-1, and PD-L1 expression were negatively associated with resistance to ICI treatment. Across the ICI, ICI-CT, and CT cohorts, 253, 153, and 178 genes, respectively, were significantly upregulated (FDR < 0.05, LogFC > 1) in patients with resistance to treatment. Of the 550 distinct genes correlated to resistance across treatments, 95% were unique to each cohort, and only 32 genes (5%) were shared across more than one treatment cohort.
Conclusions:Distinct molecular characteristics were associated with resistance to ICI and/or CT treatment. This suggests that mechanisms of resistance to ICI monotherapy, ICI-CT, or CT could be separate and unique. Further research can provide valuable insights for novel therapies or combinations.
VIEW THE PUBLICATION