Antibodies that target the immune checkpoint receptor programmed cell death protein 1 (PD-1) have resulted in prolonged and beneficial responses toward a variety of human cancers. However, anti–PD-1 therapy in some patients provides no benefit and/or results in adverse side effects. The factors that determine whether patients will be drug sensitive or resistant are not fully understood; therefore, genomic assessment of exceptional responders can provide important insight into patient response. Here, we identified a patient with endometrial cancer who had an exceptional response to the anti–PD-1 antibody pembrolizumab. Clinical grade targeted genomic profiling of a pretreatment tumor sample from this individual identified a mutation in DNA polymerase epsilon (POLE) that associated with an ultramutator phenotype. Analysis of The Cancer Genome Atlas (TCGA) revealed that the presence of POLE mutation associates with high mutational burden and elevated expression of several immune checkpoint genes. Together, these data suggest that cancers harboring POLE mutations are good candidates for immune checkpoint inhibitor therapy.

Human endogenous retroviruses (hERVs) are remnants of exogenous retroviruses that have integrated into the genome throughout evolution. We developed a computational workflow, hervQuant, which identified more than 3,000 transcriptionally active hERVs within The Cancer Genome Atlas (TCGA) pan-cancer RNA-Seq database. hERV expression was associated with clinical prognosis in several tumor types, most significantly clear cell renal cell carcinoma (ccRCC). We explored two mechanisms by which hERV expression may influence the tumor immune microenvironment in ccRCC: (i) RIG-I–like signaling and (ii) retroviral antigen activation of adaptive immunity. We demonstrated the ability of hERV signatures associated with these immune mechanisms to predict patient survival in ccRCC, independent of clinical staging and molecular subtyping. We identified potential tumor-specific hERV epitopes with evidence of translational activity through the use of a ccRCC ribosome profiling (Ribo-Seq) dataset, validated their ability to bind HLA in vitro, and identified the presence of MHC tetramer–positive T cells against predicted epitopes. hERV sequences identified through this screening approach were significantly more highly expressed in ccRCC tumors responsive to treatment with programmed death receptor 1 (PD-1) inhibition. hervQuant provides insights into the role of hERVs within the tumor immune microenvironment, as well as evidence that hERV expression could serve as a biomarker for patient prognosis and response to immunotherapy.

Response to immune checkpoint therapy can be associated with a high mutation burden, but other mechanisms are also likely to be important. We identified a patient with metastatic gastric cancer with meaningful clinical benefit from treatment with the anti–programmed death–ligand 1 (PD-L1) antibody avelumab. This tumor showed no evidence of high mutation burden or mismatch repair defect but was strongly positive for presence of Epstein-Barr virus (EBV) encoded RNA. Analysis of The Cancer Genome Atlas gastric cancer data (25 EBV+, 80 microsatellite-instable [MSI], 310 microsatellite-stable [MSS]) showed that EBV-positive tumors were MSS. Two-sided Wilcoxon rank-sum tests showed that: 1) EBV-positive tumors had low mutation burden (median = 2.07 vs 3.13 in log₁₀ scale, P < 10^-12) but stronger evidence of immune infiltration (median ImmuneScore 2212 vs 1295, P < 10^-4; log₂ fold-change of CD8A = 1.85, P < 10^-6) compared with MSI tumors, and 2) EBV-positive tumors had higher expression of immune checkpoint pathway (PD-1, CTLA-4 pathway) genes in RNA-seq data (log₂ fold-changes: PD-1 = 1.85, PD-L1 = 1.93, PD-L2 = 1.50, CTLA-4 = 1.31, CD80 = 0.89, CD86 = 1.31, P < 10^-4 each), and higher lymphocytic infiltration by histology (median tumor-infiltrating lymphocyte score = 3 vs 2, P < .001) compared with MSS tumors. These data suggest that EBV-positive low–mutation burden gastric cancers are a subset of MSS gastric cancers that may respond to immune checkpoint therapy.

Although a subset of clear cell renal cell carcinoma (ccRCC) patients respond to immune checkpoint blockade (ICB), predictors of response remain uncertain. We investigated whether abnormal expression of endogenous retroviruses (ERVs) in tumors is associated with local immune checkpoint activation (ICA) and response to ICB. Twenty potentially immunogenic ERVs (πERVs) were identified in ccRCC in The Cancer Genome Atlas data set, and tumors were stratified into 3 groups based on their expression levels. πERV-high ccRCC tumors showed increased immune infiltration, checkpoint pathway upregulation, and higher CD8 ⁺ T cell fraction in infiltrating leukocytes compared with πERV-low ccRCC tumors. Similar results were observed in ER ⁺ /HER2 ⁻ breast, colon, and head and neck squamous cell cancers. ERV expression correlated with expression of genes associated with histone methylation and chromatin regulation, and πERV-high ccRCC was enriched in BAP1 mutant tumors. ERV3-2 expression correlated with ICA in 11 solid cancers, including the 4 named above. In a small retrospective cohort of 24 metastatic ccRCC patients treated with single-agent PD-1/PD-L1 blockade, ERV3-2 expression in tumors was significantly higher in responders compared with nonresponders. Thus, abnormal expression of πERVs is associated with ICA in several solid cancers, including ccRCC, and ERV3-2 expression is associated with response to ICB in ccRCC.

Gene-expression companion diagnostic tests, such as the Oncotype DX test, assess the risk of early stage Estrogen receptor (ER) positive (+) breast cancers, and guide clinicians in the decision of whether or not to use chemotherapy. However, these tests are typically expensive, time consuming, and tissue-destructive. In this paper, we evaluate the ability of computer-extracted nuclear morphology features from routine hematoxylin and eosin (H&E) stained images of 178 early stage ER+ breast cancer patients to predict corresponding risk categories derived using the Oncotype DX test. A total of 216 features corresponding to the nuclear shape and architecture categories from each of the pathologic images were extracted and four feature selection schemes: Ranksum, Principal Component Analysis with Variable Importance on Projection (PCA-VIP), Maximum-Relevance, Minimum Redundancy Mutual Information Difference (MRMR MID), and Maximum-Relevance, Minimum Redundancy - Mutual Information Quotient (MRMR MIQ), were employed to identify the most discriminating features. These features were employed to train 4 machine learning classifiers: Random Forest, Neural Network, Support Vector Machine, and Linear Discriminant Analysis, via 3-fold cross validation. The four sets of risk categories, and the top Area Under the receiver operating characteristic Curve (AUC) machine classifier performances were: 1) Low ODx and Low mBR grade vs. High ODx and High mBR grade (Low-Low vs. High-High) (AUC = 0.83), 2) Low ODx vs. High ODx (AUC = 0.72), 3) Low ODx vs. Intermediate and High ODx (AUC = 0.58), and 4) Low and Intermediate ODx vs. High ODx (AUC = 0.65). Trained models were tested independent validation set of 53 cases which comprised of Low and High ODx risk, and demonstrated per-patient accuracies ranging from 75 to 86%. Our results suggest that computerized image analysis of digitized H&E pathology images of early stage ER+ breast cancer might be able predict the corresponding Oncotype DX risk categories.