Immunogenomic Landscape of Neuroendocrine Prostate Cancer.

PURPOSE: Patients with neuroendocrine prostate cancer (NEPC) are often managed with immunotherapy regimens extrapolated from small-cell lung cancer (SCLC). We sought to evaluate the tumor immune landscape of NEPC compared with other prostate cancer types and SCLC. EXPERIMENTAL DESIGN: In this retrospective study, a cohort of 170 patients with 230 RNA-sequencing and 104 matched whole-exome sequencing data were analyzed. Differences in immune and stromal constituents, frequency of genomic alterations, and associations with outcomes were evaluated. RESULTS: In our cohort, 36% of the prostate tumors were identified as CD8+ T-cell inflamed, whereas the remaining 64% were T-cell depleted. T-cell-inflamed tumors were enriched in anti-inflammatory M2 macrophages and exhausted T cells and associated with shorter overall survival relative to T-cell-depleted tumors (HR, 2.62; P < 0.05). Among all prostate cancer types in the cohort, NEPC was identified to be the most immune depleted, wherein only 9 out of the 36 total NEPC tumors were classified as T-cell inflamed. These inflamed NEPC cases were enriched in IFN gamma signaling and PD-1 signaling compared with other NEPC tumors. Comparison of NEPC with SCLC revealed that NEPC had poor immune content and less mutations compared with SCLC, but expression of checkpoint genes PD-L1 and CTLA-4 was comparable between NEPC and SCLC. CONCLUSIONS: NEPC is characterized by a relatively immune-depleted tumor immune microenvironment compared with other primary and metastatic prostate adenocarcinoma except in a minority of cases. These findings may inform development of immunotherapy strategies for patients with advanced prostate cancer.

NuCLS: A scalable crowdsourcing approach and dataset for nucleus classification and segmentation in breast cancer.

BACKGROUND: Deep learning enables accurate high-resolution mapping of cells and tissue structures that can serve as the foundation of interpretable machine-learning models for computational pathology. However, generating adequate labels for these structures is a critical barrier, given the time and effort required from pathologists. RESULTS: This article describes a novel collaborative framework for engaging crowds of medical students and pathologists to produce quality labels for cell nuclei. We used this approach to produce the NuCLS dataset, containing >220,000 annotations of cell nuclei in breast cancers. This builds on prior work labeling tissue regions to produce an integrated tissue region- and cell-level annotation dataset for training that is the largest such resource for multi-scale analysis of breast cancer histology. This article presents data and analysis results for single and multi-rater annotations from both non-experts and pathologists. We present a novel workflow that uses algorithmic suggestions to collect accurate segmentation data without the need for laborious manual tracing of nuclei. Our results indicate that even noisy algorithmic suggestions do not adversely affect pathologist accuracy and can help non-experts improve annotation quality. We also present a new approach for inferring truth from multiple raters and show that non-experts can produce accurate annotations for visually distinctive classes. CONCLUSIONS: This study is the most extensive systematic exploration of the large-scale use of wisdom-of-the-crowd approaches to generate data for computational pathology applications.