Costimulatory Molecule Expression Profile as a Biomarker to Predict Prognosis and Chemotherapy Response for Patients With Small-cell Lung Cancer

Zhihui Zhang,Peng Wu,Zhaoyang Yang,Chaoqi Zhang,Yuejun Luo,Guochao Zhang,Lide Wang,Qi Xue,Nan Sun,Jie He
DOI: https://doi.org/10.21203/rs.3.rs-801123/v1
2021-08-30
Abstract:Abstract Background : Small cell lung cancer (SCLC) is the deadliest pathological subtype of lung cancer with few precise treatment options. Owing to the paucity of specimens, progress in identifying prognostic and therapeutic biomarkers for SCLC has been stagnant for decades. Considering the costimulatory molecules are essential elements in modulating anti-tumor immune responses and determining therapeutic response, we decided to conduct a comprehensive analyses of costimulatory molecule expression profile with regard to prognosis and chemotherapy response for SCLCs for the first time. Methods : We systematically explored the molecular characteristics and clinical relevance of costimulatory molecules in SCLC. Using a univariate Cox and LASSO Cox model, we then constructed a molecule-based signature (CMS) based on our examination of 208 cases with SCLC from two different cohorts—including an independent cohort that contained 131 cases with qPCR data, to predict the outcome and chemotherapy response for SCLCs. Relationships between the CMS and adjuvant chemotherapy (ACT) benefits and inflammatory landscape were also evaluated. Results : We found T cell activation was restrained in SCLCs, and costimulatory molecules exhibited widespread abnormal genetic alterations and expression. Then, the CMS containing 7 molecules (ICOSLG, EDA2R, TNFRSF25, CD276, RELT, PDCD1, and TNFSF14) was built with a training cohort of 77 cases, which successfully divided patients into high- or low-risk groups with significantly different prognosis and chemotherapy benefit (both P <0.001). The CMS was well validated in the independent cohort with qPCR data. ROC and C-index analysis confirmed the superior predictive performance of the CMS in comparison with other clinicopathological parameters from different cohorts. Importantly, the CMS was confirmed as a significantly independent prognosticator for clinical outcomes and chemotherapy response in SCLCs through multivariate Cox analysis. Further analysis revealed that low-risk patients were characteristic by an activated immune phenotype with distinct expression of immune checkpoints. Conclusions : We firstly uncovered the expression heterogeneity of costimulatory molecules in SCLC, and successfully constructed a novel predictive CMS. The identified signature contributed to more accurate patient stratification and provided robust prognostic value in estimating survival and the clinical response to chemotherapy, allowing optimization of treatment and prognosis management for patients with SCLC.
What problem does this paper attempt to address?