Background: Immunotherapy has revolutionized cancer treatment, but most patients are refractory to immunotherapy or acquire resistance, with the underlying mechanisms remaining to be explored. Methods: We characterized the transcriptomes of ~92,000 single cells from 3 pre-treatment and 12 post-treatment patients with non-small cell lung cancer (NSCLC) who received neoadjuvant PD-1 blockade combined with chemotherapy. The 12 post-treatment samples were categorized into two groups based on pathologic response: major pathologic response (MPR; n = 4) and non-MPR (NMPR; n = 8). Results: Distinct therapy-induced cancer cell transcriptomes were associated with clinical response. Cancer cells from MPR patients exhibited a signature of activated antigen presentation via major histocompatibility complex class II (MHC-II). Further, the transcriptional signatures of FCRL4+FCRL5+ memory B cells and CD16+CX3CR1+ monocytes were enriched in MPR patients and are predictors of immunotherapy response. Cancer cells from NMPR patients exhibited overexpression of estrogen metabolism enzymes and elevated serum estradiol. In all patients, therapy promoted expansion and activation of cytotoxic T cells and CD16+ NK cells, reduction of immunosuppressive Tregs, and activation of memory CD8+T cells into an effector phenotype. Tissue-resident macrophages were expanded after therapy, and tumor-associated macrophages (TAMs) were remodeled into a neutral instead of an anti-tumor phenotype. We revealed the heterogeneity of neutrophils during immunotherapy and identified an aged CCL3+ neutrophil subset was decreased in MPR patients. The aged CCL3+ neutrophils were predicted to interact with SPP1+ TAMs through a positive feedback loop to contribute to a poor therapy response. Conclusions: Neoadjuvant PD-1 blockade combined with chemotherapy led to distinct NSCLC tumor microenvironment transcriptomes that correlated with therapy response. Although limited by a small patient sample size subjected to combination therapy, this study provides novel biomarkers to predict therapy response and suggests potential strategies to overcome immunotherapy resistance.

Alterations of gut microbiota have been implicated in multiple diseases including cancer. However, the gut microbiota spectrum in lung cancer remains largely unknown. Here we profiled the gut microbiota composition in a discovery cohort containing 42 early-stage lung cancer patients and 65 healthy individuals through the 16S ribosomal RNA (rRNA) gene sequencing analysis. We found that lung cancer patients displayed a significant shift of microbiota composition in contrast to the healthy populations. To identify the optimal microbiota signature for noninvasive diagnosis purpose, we took advantage of Support-Vector Machine (SVM) and found that the predictive model with 13 operational taxonomic unit (OTU)-based biomarkers achieved a high accuracy in lung cancer prediction (area under curve, AUC = 97.6%). This signature performed reasonably well in the validation cohort (AUC = 76.4%), which contained 34 lung cancer patients and 40 healthy individuals. To facilitate potential clinical practice, we further constructed a ‘patient discrimination index’ (PDI), which largely retained the prediction efficiency in both the discovery cohort (AUC = 92.4%) and the validation cohort (AUC = 67.7%). Together, our study uncovered the microbiota spectrum of lung cancer patients and established the specific gut microbial signature for the potential prediction of the early-stage lung cancer.