Non–small cell lung cancer (NSCLC) accounts for approximately 85% of lung cancer, the leading cause of death from cancer in the United States and worldwide.¹ Biomarkers that can reliably estimate disease prognosis and patient survival would have tremendous value in guiding the management of lung cancer.² For example, most patients with stage I NSCLC currently do not receive adjuvant systemic treatment after local therapy because several large randomized studies^3-5 have failed to show a survival benefit among unselected patients, for whom the toxic effects associated with chemotherapy may outweigh the potential benefit for many patients. Thus, identification of the subset of patients at highest risk for recurrence and death who have the greatest need for additional systemic therapy is needed.

Radiographic imaging is routinely used to evaluate treatment response in solid tumors. Current imaging response metrics do not reliably predict the underlying biological response. Here, we present a multi-task deep learning approach that allows simultaneous tumor segmentation and response prediction. We design two Siamese subnetworks that are joined at multiple layers, which enables integration of multi-scale feature representations and in-depth comparison of pre-treatment and post-treatment images. The network is trained using 2568 magnetic resonance imaging scans of 321 rectal cancer patients for predicting pathologic complete response after neoadjuvant chemoradiotherapy. In multi-institution validation, the imaging-based model achieves AUC of 0.95 (95% confidence interval: 0.91–0.98) and 0.92 (0.87–0.96) in two independent cohorts of 160 and 141 patients, respectively. When combined with blood-based tumor markers, the integrated model further improves prediction accuracy with AUC 0.97 (0.93–0.99). Our approach to capturing dynamic information in longitudinal images may be broadly used for screening, treatment response evaluation, disease monitoring, and surveillance.

We discovered and validated three intratumoral subregions with distinct MR perfusion imaging parameters. In addition to clinical-pathologic and genomic factors, intratumoral spatial heterogeneity c... Purpose To characterize intratumoral spatial heterogeneity at perfusion magnetic resonance (MR) imaging and investigate intratumoral heterogeneity as a predictor of recurrence-free survival (RFS) i...

Purpose: To identify immune subtypes and investigate the immune landscape of squamous cell carcinomas (SCC), which share common etiology and histologic features. Experimental Design: Based on the immune gene expression profiles of 1,368 patients with SCC in the Cancer Genome Atlas (TCGA), we used consensus clustering to identify robust clusters of patients and assessed their reproducibility in an independent pan-SCC cohort of 938 patients. We further applied graph structure learning-based dimensionality reduction to the immune profiles to visualize the distribution of individual patients. Results: We identified and independently validated six reproducible immune subtypes associated with distinct molecular characteristics and clinical outcomes. An immune-cold subtype had the least amount of lymphocyte infiltration and a high level of aneuploidy, and these patients had the worst prognosis. By contrast, an immune-hot subtype demonstrated the highest infiltration of CD8⁺ T cells, activated NK cells, and elevated IFNγ response. Accordingly, these patients had the best prognosis. A third subtype was dominated by M2-polarized macrophages with potent immune-suppressive factors such as TGFβ signaling and reactive stroma, and these patients had relatively inferior prognosis. Other subtypes showed more diverse immunologic features with intermediate prognoses. Finally, our analysis revealed a complex immune landscape consisting of both discrete clusters and continuous spectrum. Conclusions: This study provides a conceptual framework to understand the tumor immune microenvironment of SCCs. Future work is needed to evaluate its relevance in the design of combination treatment strategies and guiding optimal selection of patients for immunotherapy.