The ectonucleotidase CD39 identifies tumor-reactive CD8+ T cells predictive of immune checkpoint blockade efficacy in human lung cancer.

Improved identification of anti-tumor T cells is needed to advance cancer immunotherapies. CD39 expression is a promising surrogate of tumor-reactive CD8+ T cells. Here, we comprehensively profiled CD39 expression in human lung cancer. CD39 expression enriched for CD8+ T cells with features of exhaustion, tumor reactivity, and clonal expansion. Flow cytometry of 440 lung cancer biospecimens revealed weak association between CD39+ CD8+ T cells and tumoral features, such as programmed death-ligand 1 (PD-L1), tumor mutation burden, and driver mutations. Immune checkpoint blockade (ICB), but not cytotoxic chemotherapy, increased intratumoral CD39+ CD8+ T cells. Higher baseline frequency of CD39+ CD8+ T cells conferred improved clinical outcomes from ICB therapy. Furthermore, a gene signature of CD39+ CD8+ T cells predicted benefit from ICB, but not chemotherapy, in a phase III clinical trial of non-small cell lung cancer. These findings highlight CD39 as a proxy of tumor-reactive CD8+ T cells in human lung cancer.

Anti-TIGIT antibody improves PD-L1 blockade through myeloid and Treg cells.

Tiragolumab, an anti-TIGIT antibody with an active IgG1κ Fc, demonstrated improved outcomes in the phase 2 CITYSCAPE trial (ClinicalTrials.gov: NCT03563716 ) when combined with atezolizumab (anti-PD-L1) versus atezolizumab alone1. However, there remains little consensus on the mechanism(s) of response with this combination2. Here we find that a high baseline of intratumoural macrophages and regulatory T cells is associated with better outcomes in patients treated with atezolizumab plus tiragolumab but not with atezolizumab alone. Serum sample analysis revealed that macrophage activation is associated with a clinical benefit in patients who received the combination treatment. In mouse tumour models, tiragolumab surrogate antibodies inflamed tumour-associated macrophages, monocytes and dendritic cells through Fcγ receptors (FcγR), in turn driving anti-tumour CD8+ T cells from an exhausted effector-like state to a more memory-like state. These results reveal a mechanism of action through which TIGIT checkpoint inhibitors can remodel immunosuppressive tumour microenvironments, and suggest that FcγR engagement is an important consideration in anti-TIGIT antibody development.

Androgen receptor activity in T cells limits checkpoint blockade efficacy.

Immune checkpoint blockade has revolutionized the field of oncology, inducing durable anti-tumour immunity in solid tumours. In patients with advanced prostate cancer, immunotherapy treatments have largely failed1-5. Androgen deprivation therapy is classically administered in these patients to inhibit tumour cell growth, and we postulated that this therapy also affects tumour-associated T cells. Here we demonstrate that androgen receptor (AR) blockade sensitizes tumour-bearing hosts to effective checkpoint blockade by directly enhancing CD8 T cell function. Inhibition of AR activity in CD8 T cells prevented T cell exhaustion and improved responsiveness to PD-1 targeted therapy via increased IFNγ expression. AR bound directly to Ifng and eviction of AR with a small molecule significantly increased cytokine production in CD8 T cells. Together, our findings establish that T cell intrinsic AR activity represses IFNγ expression and represents a novel mechanism of immunotherapy resistance.

Pregnancy and weaning regulate human maternal liver size and function.

During pregnancy, the rodent liver undergoes hepatocyte proliferation and increases in size, followed by weaning-induced involution via hepatocyte cell death and stromal remodeling, creating a prometastatic niche. These data suggest a mechanism for increased liver metastasis in breast cancer patients with recent childbirth. It is unknown whether the human liver changes in size and function during pregnancy and weaning. In this study, abdominal imaging was obtained in healthy women at early and late pregnancy and postwean. During pregnancy time points, glucose production and utilization and circulating bile acids were measured. Independently of weight gain, most women's livers increased in size with pregnancy, then returned to baseline postwean. Putative roles for bile acids in liver growth and regression were observed. Together, the data support the hypothesis that the human liver is regulated by reproductive state with growth during pregnancy and volume loss postwean. These findings have implications for sex-specific liver diseases and for breast cancer outcomes.

Transcriptional profiling identifies an androgen receptor activity-low, stemness program associated with enzalutamide resistance.

The androgen receptor (AR) antagonist enzalutamide is one of the principal treatments for men with castration-resistant prostate cancer (CRPC). However, not all patients respond, and resistance mechanisms are largely unknown. We hypothesized that genomic and transcriptional features from metastatic CRPC biopsies prior to treatment would be predictive of de novo treatment resistance. To this end, we conducted a phase II trial of enzalutamide treatment (160 mg/d) in 36 men with metastatic CRPC. Thirty-four patients were evaluable for the primary end point of a prostate-specific antigen (PSA)50 response (PSA decline ≥50% at 12 wk vs. baseline). Nine patients were classified as nonresponders (PSA decline <50%), and 25 patients were classified as responders (PSA decline ≥50%). Failure to achieve a PSA50 was associated with shorter progression-free survival, time on treatment, and overall survival, demonstrating PSA50's utility. Targeted DNA-sequencing was performed on 26 of 36 biopsies, and RNA-sequencing was performed on 25 of 36 biopsies that contained sufficient material. Using computational methods, we measured AR transcriptional function and performed gene set enrichment analysis (GSEA) to identify pathways whose activity state correlated with de novo resistance. TP53 gene alterations were more common in nonresponders, although this did not reach statistical significance (P = 0.055). AR gene alterations and AR expression were similar between groups. Importantly, however, transcriptional measurements demonstrated that specific gene sets-including those linked to low AR transcriptional activity and a stemness program-were activated in nonresponders. Our results suggest that patients whose tumors harbor this program should be considered for clinical trials testing rational agents to overcome de novo enzalutamide resistance.

Increased nuclear but not cytoplasmic activities of CELF1 protein leads to muscle wasting.

mRNA processing is highly regulated during development through changes in RNA-binding protein (RBP) activities. CUG-BP, Elav-like family member 1 (CELF1, also called CUGBP1) is an RBP, the expression of which decreases in skeletal muscle soon after birth. CELF1 regulates multiple nuclear and cytoplasmic RNA processing events. In the nucleus, CELF1 regulates networks of postnatal alternative splicing (AS) transitions, while in the cytoplasm, CELF1 regulates mRNA stability and translation. Stabilization and misregulation of CELF1 has been implicated in human diseases including myotonic dystrophy type 1, Alzheimer's disease and multiple cancers. To understand the contribution of nuclear and cytoplasmic CELF1 activity to normal and pathogenic skeletal muscle biology, we generated transgenic mice for doxycycline-inducible and skeletal muscle-specific expression of active CELF1 mutants engineered to be localized predominantly to either the nucleus or the cytoplasm. Adult mice expressing nuclear, but not cytoplasmic, CELF1 are characterized by strong histopathological defects, muscle loss within 10 days and changes in AS. In contrast, mice expressing cytoplasmic CELF1 display changes in protein levels of targets known to be regulated at the level of translation by CELF1, with minimal changes in AS. These changes are in the absence of overt histopathological changes or muscle loss. RNA-sequencing revealed extensive gene expression and AS changes in mice overexpressing nuclear and naturally localized CELF1 protein, with affected genes involved in cytoskeleton dynamics, membrane dynamics, RNA processing and zinc ion binding. These results support a stronger role for nuclear CELF1 functions as compared to cytoplasmic CELF1 functions in skeletal muscle wasting.

T Cell Transcriptional Profiling and Immunophenotyping Uncover LAG3 as a Potential Significant Target of Immune Modulation in Multiple Myeloma.

Autologous stem cell transplant (ASCT) is the standard of care for patients with multiple myeloma (MM). The clinical significance of peripheral blood T lymphocyte (PBTL) immunologic changes associated with ASCT is poorly understood. Here we evaluated T cell transcriptional messenger RNA profiles and immunophenotypes to correlate immunologic senescence, exhaustion, and anergy with clinical endpoints in a cohort of patients with MM undergoing ASCT. ASCT induced global transcriptional T cell changes and altered molecular levels of markers of T cell subtypes, T cell activation, and exhaustion. These included reduced CD4/CD8 ratio, skewing toward the Th1 subset, reduced expression of costimulatory receptors CD27 and CD28, heightened T cell activation, and increased expression of immune modulatory molecules LAG3 and PD1. Multicolor flow cytometry experiments confirmed altered circulating CD4 and CD8 subsets and skewing toward differentiated effector cells. Moreover, ASCT promoted an exhausted immunophenotype in CD3+CD4+ subsets and a senescent immunophenotype in CD3+CD8+ subsets. Subset-specific altered expression was also seen for surface molecules with immunomodulatory function. ASCT affected soluble levels of molecules with immunomodulatory function by increasing plasma HVEM and TIM3. High molecular LAG3 level was associated with inferior event-free survival post-ASCT (hazard ratio = 5.44; confidence interval, 1.92 to 15.46; P = .001; adjusted P [controlling for false discovery rate] = .038). Using a comprehensive evaluation of PBTLs on a molecular and phenotypic level, we have identified that ASCT induces global T cell alterations with CD4 and CD8 subset-specific changes. Moreover, LAG3 emerged as an early biomarker of adverse events post-ASCT. These findings will support the development of treatment strategies targeting immune defects in MM to augment or restore T cell responses.

[Multi-objective Identification Method for Influencing Factors of Soil Heavy Metal Content Change].

Identifying the influencing factors of soil heavy metal content changes is the basis for reducing or preventing soil heavy metal pollution. Taking an agricultural experimental field in Changping District of Beijing as an example, the heavy metal content changes in As, Cr, Cu, Ni, Pb, and Zn from 2012 to 2022 were firstly analyzed. Secondly, the influencing factors of the heavy metal content changes were detected based on the geographical detector at the single-target and multi-target levels, respectively. Finally, comparative experiments with the correlation analysis method and existing studies were set up to evaluate the effectiveness of the identification method of influencing factors developed in this study. The results showed that human activity factors have exacerbated the changes in soil heavy metal content in the study area as follows： ① At the single-target level, the land use type was the main influencing factor on the changes in Cr, Cu, and Zn contents, and the annual deposition flux influenced the changes in As. The results of the interaction detection showed that there was an enhancement effect among the factors, and the interaction of the human activity factors dominated for the factor identification. ② The results of the multi-target level detection covered the results of the single-target level detection, which could identify more influencing factors. The land use type affected the changes in Cu, Zn, Cr, Ni, and As, and the changes in As and Zn were influenced by the annual deposition fluxes. ③ The multi-target identification method coupled with geographical detector and principal component analysis could effectively identify the influencing factors of soil heavy metal content changes, which was much more effective than the single soil heavy metal correlation method. The developed multi-target identification method for influencing factors of heavy metal content changes can provide technical support for the regional pollution monitoring and macro-management of soil heavy metals.

Atezolizumab Plus Chemotherapy With or Without Bevacizumab in Advanced Biliary Tract Cancer: Clinical and Biomarker Data From the Randomized Phase II IMbrave151 Trial.

PURPOSE: Biliary tract cancers (BTCs) harbor an immunosuppressed tumor microenvironment and respond poorly to PD-1/PD-L1 inhibitors. Bevacizumab (anti-vascular endothelial growth factor) plus chemotherapy can promote anticancer immunity, augmenting response to PD-L1 inhibition. PATIENTS AND METHODS: This randomized, double-blind, proof-of-concept phase II study enrolled patients (n = 162) with previously untreated advanced BTC (IMbrave151; ClinicalTrials.gov identifier: NCT04677504). Patients were randomly assigned 1:1 to receive cycles of atezolizumab (1,200 mg) plus bevacizumab (15 mg/kg) or atezolizumab plus placebo once every 3 weeks until disease progression or unacceptable toxicity. All patients received cisplatin (25 mg/m2) plus gemcitabine (1,000 mg/m2; cisplatin plus gemcitabine [CisGem]) on days 1 and 8 once every 3 weeks for up to eight cycles. Stratification of patients was by disease status, geographic region, and primary tumor location. The primary end point was progression-free survival (PFS). No formal hypothesis testing was performed. Exploratory correlative biomarker analysis was undertaken using transcriptome analysis (n = 95) and mutation profiling (n = 102) on baseline tumor samples. RESULTS: Between February and September 2021, 162 patients were enrolled. Median PFS was 8.3 months in the bevacizumab arm and 7.9 months in the placebo arm (stratified hazard ratio [HR], 0.67 [95% CI, 0.46 to 0.95]). Median overall survival (OS) was 14.9 and 14.6 months in the bevacizumab and placebo arms, respectively (stratified HR, 0.97 [95% CI, 0.64 to 1.47]). The incidence of grade 3 or 4 adverse events was 74% in both arms. High VEGFA gene expression was associated with improved PFS (HR, 0.44 [95% CI, 0.23 to 0.83]) in the bevacizumab arm versus placebo. CONCLUSION: In unselected patients with advanced BTC, adding bevacizumab to atezolizumab plus CisGem modestly improves PFS but not OS. High VEGFA gene expression may represent a predictive biomarker of benefit from atezolizumab/bevacizumab, warranting further investigation.

Immunomodulatory effects and improved outcomes with cisplatin- versus carboplatin-based chemotherapy plus atezolizumab in urothelial cancer.

In metastatic urothelial cancer (mUC), cisplatin versus carboplatin leads to durable disease control in a subset of patients. The IMvigor130 trial reveals more favorable effects with atezolizumab combined with gemcitabine and cisplatin (GemCis) versus gemcitabine and carboplatin (GemCarbo). This study investigates the immunomodulatory effects of cisplatin as a potential explanation for these observations. Our findings indicate that improved outcomes with GemCis versus GemCarbo are primarily observed in patients with pretreatment tumors exhibiting features of restrained adaptive immunity. In addition, GemCis versus GemCarbo ± atezolizumab induces transcriptional changes in circulating immune cells, including upregulation of antigen presentation and T cell activation programs. In vitro experiments demonstrate that cisplatin, compared with carboplatin, exerts direct immunomodulatory effects on cancer cells, promoting dendritic cell activation and antigen-specific T cell killing. These results underscore the key role of immune modulation in cisplatin's efficacy in mUC and highlight the importance of specific chemotherapy backbones in immunotherapy combination regimens.

Wheat hypersensitive-induced reaction genes TaHIR1 and TaHIR3 are involved in response to stripe rust fungus infection and abiotic stresses.

KEY MESSAGE : TaHIR1 and TaHIR3 play positive roles in resistance to the stripe rust fungus via inducing HR and regulating defense-related genes, but are negatively regulated by various abiotic stimuli. Plant hypersensitive-induced reaction (HIR) genes are known to be associated with the hypersensitive response and disease defense. In wheat, two HIR genes, TaHIR1 and TaHIR3, have been identified and found to be up-regulated after infection with the stripe rust fungus. Here, we further determined their roles in defense against abiotic stresses and the stripe rust pathogen, Puccinia striiformis f. sp. tritici. TaHIR1 and TaHIR3 proteins were localized in the plasma membrane of tobacco cells. The expression of TaHIR1 and TaHIR3 was reduced by the environmental stimuli, including low temperature, drought, and high salinity stresses. In addition, the expression of TaHIR1 and TaHIR3 was down-regulated by exogenously applied ethrel and abscisic acid, whereas expression was not affected by treatments with salicylic acid and methyl jasmonate. Furthermore, barley stripe mosaic virus-induced gene silencing of TaHIR1 and TaHIR3 reduced resistance in wheat cultivar Suwon11 against an avirulent stripe rust pathotype CYR23 and area of necrotic cells neighboring the infection sites, and altered the expression levels of defense-related genes. These results suggest that TaHIR1 and TaHIR3 function positively in the incompatible interaction of wheat-stripe rust fungus, but exhibit negative transcriptional response to abiotic stresses.

Agrobacterium tumefaciens-mediated transformation of Valsa mali: an efficient tool for random insertion mutagenesis.

Valsa mali is a causal agent of apple and pear trees canker disease, which is a destructive disease that causes serious economic losses in eastern Asia, especially in China. The lack of an efficient transformation system for Valsa mali retards its investigation, which poses difficulties to control the disease. In this research, a transformation system for this pathogen was established for the first time using A. tumefaciens-mediated transformation (ATMT), with the optimal transformation conditions as follows: 10(6)/mL conidia suspension, cocultivation temperature 22°C, cocultivation time 72 hours, and 200 µ M acetosyringone (AS) in the inductive medium. The average transformation efficiency was 1015.00 ± 37.35 transformants per 10(6) recipient conidia. Thirty transformants were randomly selected for further confirmation and the results showed the presence of T-DNA in all hygromycin B resistant transformants and also revealed random and single gene integration with genetic stability. Compared with wild-type strain, those transformants exhibited various differences in morphology, conidia production, and conidia germination ability. In addition, pathogenicity assays revealed that 14 transformants had mitigated pathogenicity, while one had enhanced infection ability. The results suggest that ATMT of V. mali is a useful tool to gain novel insight into this economically important pathogen at molecular levels.

Tetramer-aided sorting and single-cell RNA sequencing facilitate transcriptional profiling of antigen-specific CD8+ T cells.

BACKGROUND: Recent advances in single-cell technologies and an improved understanding of tumor antigens have empowered researchers to investigate tumor antigen-specific CD8+ T cells at the single-cell level. Peptide-MHC I tetramers are often utilized to enrich antigen-specific CD8+ T cells, which however, introduces the undesired risk of altering their clonal distribution or their transcriptional state. This study addresses the feasibility of utilizing tetramers to enrich antigen-specific CD8+ T cells for single-cell analysis. METHODS: HLA-A*02:01-restricted human cytomegalovirus (CMV) pp65 peptide-specific CD8+ T cells were used as a model for analyzing antigen-specific CD8+ T cells. Single-cell RNA sequencing and TCR sequencing were performed to compare the frequency and gene expression profile of pp65-specific TCR clones between tetramer-sorted, unstimulated- and tetramer-stimulated total CD8+ T cells. RESULTS: The relative frequency of pp65-specific TCR clones and their transcriptional profile remained largely unchanged following tetramer-based sorting. In contrast, tetramer-mediated stimulation of CD8+ T cells resulted in significant gene expression changes in pp65-specific CD8+ T cells. An Antigen-Specific Response (ASR) gene signature was derived from tetramer-stimulated pp65-specific CD8+ T cells. The ASR signature had a predictive value and was significantly associated with progression free survival in lung cancer patients treated with anti-PD-L1, anti-VEGF, chemotherapy combination (NCT02366143). The predictive power of the ASR signature was independent of the conventional CD8 effector signature. CONCLUSIONS: Our findings validate the approach of enriching antigen-specific CD8+ T cells through tetramer-aided Fluorescence-Activated Cell Sorting (FACS) sorting for single-cell analysis and also identifies an ASR gene signature that has value in predicting response to cancer immunotherapy.

Supervised learning of high-confidence phenotypic subpopulations from single-cell data.

Accurately identifying phenotype-relevant cell subsets from heterogeneous cell populations is crucial for delineating the underlying mechanisms driving biological or clinical phenotypes. Here, by deploying a learning with rejection strategy, we developed a novel supervised learning framework called PENCIL to identify subpopulations associated with categorical or continuous phenotypes from single-cell data. By embedding a feature selection function into this flexible framework, for the first time, we were able to select informative features and identify cell subpopulations simultaneously, which enables the accurate identification of phenotypic subpopulations otherwise missed by methods incapable of concurrent gene selection. Furthermore, the regression mode of PENCIL presents a novel ability for supervised phenotypic trajectory learning of subpopulations from single-cell data. We conducted comprehensive simulations to evaluate PENCILs versatility in simultaneous gene selection, subpopulation identification and phenotypic trajectory prediction. PENCIL is fast and scalable to analyze 1 million cells within 1 hour. Using the classification mode, PENCIL detected T-cell subpopulations associated with melanoma immunotherapy outcomes. Moreover, when applied to scRNA-seq of a mantle cell lymphoma patient with drug treatment across multiple time points, the regression mode of PENCIL revealed a transcriptional treatment response trajectory. Collectively, our work introduces a scalable and flexible infrastructure to accurately identify phenotype-associated subpopulations from single-cell data.

Single-cell analysis reveals clonally expanded tumor-associated CD57+ CD8 T cells are enriched in the periphery of patients with metastatic urothelial cancer responding to PD-L1 blockade.

BACKGROUND: A growing body of evidence suggests that T-cell responses against neoantigens are critical regulators of response to immune checkpoint blockade. We previously showed that circulating neoantigen-specific CD8 T cells in patients with lung cancer responding to anti-Programmed death-ligand 1 (PD-L1) (atezolizumab) exhibit a unique phenotype with high expression of CD57, CD244, and KLRG1. Here, we extended our analysis on neoantigen-specific CD8 T cells to patients with metastatic urothelial cancer (mUC) and further profiled total CD8 T cells to identify blood-based predictive biomarkers of response to atezolizumab. METHODS: We identified tumor neoantigens from 20 patients with mUC and profiled their peripheral CD8 T cells using highly multiplexed combinatorial tetramer staining. Another set of patients with mUC treated with atezolizumab (n=30) or chemotherapy (n=40) were selected to profile peripheral CD8 T cells by mass cytometry. Using single-cell transcriptional analysis (single-cell RNA sequencing (scRNA-seq)), together with CITE-seq (cellular indexing of transcriptomes and epitopes by sequencing) and paired T-cell receptor (TCR) sequencing, we further characterized peripheral CD8 T cells in a subset of patients (n=16). RESULTS: High frequency of CD57 was observed in neoantigen-specific CD8 T cells in patients with mUC responding to atezolizumab. Extending these findings to bulk CD8 T cells, we found higher frequency of CD57 expressing CD8 T cells before treatment in patients responding to atezolizumab (n=20, p<0.01) but not to chemotherapy. These findings were corroborated in a validation cohort (n=30, p<0.01) and notably were independent of known biomarkers of response. scRNA-seq analysis identified a clonally expanded cluster enriched within CD57+ CD8 T cells in responding patients characterized by higher expression of genes associated with activation, cytotoxicity, and tissue-resident memory markers. Furthermore, compared with CD57- CD8 T cells, TCRs of CD57+ CD8 T cells showed increased overlap with the TCR repertoire of tumor-infiltrating T cells. CONCLUSIONS: Collectively, we show high frequencies of CD57 among neoantigen-specific and bulk CD8 T cells in patients responding to atezolizumab. The TCR repertoire overlap between peripheral CD57+ CD8 T cells and tumor-infiltrating lymphocytes suggest that accumulation of peripheral CD57+ CD8 T cells is reflective of an ongoing antitumor T-cell response. Our findings provide evidence and rationale for using circulating CD8 T cells expressing CD57 as a readily accessible blood-based biomarker for selecting patients with mUC for atezolizumab therapy.

Identifying phenotype-associated subpopulations by integrating bulk and single-cell sequencing data.

Single-cell RNA sequencing (scRNA-seq) distinguishes cell types, states and lineages within the context of heterogeneous tissues. However, current single-cell data cannot directly link cell clusters with specific phenotypes. Here we present Scissor, a method that identifies cell subpopulations from single-cell data that are associated with a given phenotype. Scissor integrates phenotype-associated bulk expression data and single-cell data by first quantifying the similarity between each single cell and each bulk sample. It then optimizes a regression model on the correlation matrix with the sample phenotype to identify relevant subpopulations. Applied to a lung cancer scRNA-seq dataset, Scissor identified subsets of cells associated with worse survival and with TP53 mutations. In melanoma, Scissor discerned a T cell subpopulation with low PDCD1/CTLA4 and high TCF7 expression associated with an immunotherapy response. Beyond cancer, Scissor was effective in interpreting facioscapulohumeral muscular dystrophy and Alzheimer's disease datasets. Scissor identifies biologically and clinically relevant cell subpopulations from single-cell assays by leveraging phenotype and bulk-omics datasets.

Transcriptional profiling of matched patient biopsies clarifies molecular determinants of enzalutamide-induced lineage plasticity.

The androgen receptor (AR) signaling inhibitor enzalutamide (enza) is one of the principal treatments for metastatic castration-resistant prostate cancer (CRPC). Several emergent enza clinical resistance mechanisms have been described, including lineage plasticity in which the tumors manifest reduced dependency on the AR. To improve our understanding of enza resistance, herein we analyze the transcriptomes of matched biopsies from men with metastatic CRPC obtained prior to treatment and at progression (n = 21). RNA-sequencing analysis demonstrates that enza does not induce marked, sustained changes in the tumor transcriptome in most patients. However, three patients' progression biopsies show evidence of lineage plasticity. The transcription factor E2F1 and pathways linked to tumor stemness are highly activated in baseline biopsies from patients whose tumors undergo lineage plasticity. We find a gene signature enriched in these baseline biopsies that is strongly associated with poor survival in independent patient cohorts and with risk of castration-induced lineage plasticity in patient-derived xenograft models, suggesting that tumors harboring this gene expression program may be at particular risk for resistance mediated by lineage plasticity and poor outcomes.