Identification of TGFBR2 mutation as a negative predictor of immunotherapy in NSCLC.

e21002 Abstract Background: Immune adaptive resistance is one of the reasons contributing to the resistance of immune checkpoint inhibitors (ICIs). Previous studies have demonstrated the role of TGFβ-SMAD signaling in the dysfunction of immune response. We hypothesized that transforming growth factor beta receptor 2 (TGFBR2) mutation in non-small cell lung cancer (NSCLC) might be a predictor of immunotherapeutic resistance. Methods: Genomic and transcriptomic data of NSCLC patients from The Cancer Genome Atlas (TCGA) were analyzed by the in silico Pathway Activation Network Decomposition Analysis (iPNADA) to study the association between TGFBR2 mutation and intracellular signaling pathway activation. Clinical and genomic data from public cohorts with patients treated with ICIs including POPLAR/OAK, MSKCC and Van Allen cohorts were obtained and analyzed. An independent China cohort was used to demonstrate the association between TGFBR2 mutation and immunotherapeutic efficacy and we further demonstrated a case with mutated TGFBR2 treated with ICI. Results: TGF-β signaling was decreased and JAK-STAT signaling was increased in patients with mutated TGFBR2 (P < 0.05). The immune checkpoints including CD274, LAG3, TIGHT, CTLA-4, PDCD1 and PDCD1LG2 were significantly increased in patients with mutated TGFRB2 (P < 0.05). The immune dysfunction score tended to be increased in patients with mutated TGFBR2 though no significant difference was observed (P = 0.23). In the POPLAR/OAK cohort, patients with mutated TGFBR2 had shorter progression-free survival (P = 0.004; HR, 2.83; 95% CI, 1.34-6.00) and overall survival (OS) (P = 0.0006; HR, 3.46; 95% CI, 1.63-7.35) than patients with wild-type TGFBR2 when treated with ICIs. While in patients treated with chemotherapy, there was no difference in PFS (P = 0.69; HR, 0.86; 95% CI, 0.41-1.82) and OS (P = 0.61; HR, 1.21; 95% CI, 0.57-2.57) between patients with mutated and wild-type TGFBR2. In the merged MSKCC and Van Allen cohorts, similar result was observed that the OS was shorter in patients with mutated TGFBR2 (P = 0.007; HR, 2.53; 95% CI, 1.25-5.12). At last, we demonstrated a 39-year-old case with TGFBR2 mutation who was treated with 3 mg/kg nivolumab as second-line treatment. The patient progressed after three cycles of nivolumab and the PFS was only 1.3 months. Conclusions: We identified TGFBR2 mutation as a negative predictor for ICIs in NSCLC and the clinical use of ICIs needs to be cautious in those patients, highlighting the importance of genomic profiling in the treatment of ICIs. Combination ICIs and TGF-β signaling inhibitors may overcome the resistance for those patients harboring TGFBR2 mutation and need to be developed in the future.