In This Issue

See article, p. 492

Chimeric antigen receptor (CAR) T-cell therapy has been highly effective in pediatric CD19-expressing acute lymphoblastic leukemia (ALL), and efforts are under way to expand use of CAR T-cell therapy in other pediatric cancers. However, many pediatric patients often undergo aggressive first-line chemotherapy regimens, which could ultimately affect the broader use of CAR T-cell therapy as the quality of collected T cells is a determinant of CAR T-cell efficacy. Das and colleagues enrolled 195 newly diagnosed pediatric patients with 10 different cancers in a clinical trial to prospectively and longitudinally study expansion ability and surface phenotype of T cells isolated at diagnosis and after every cycle of chemotherapy. In vitro expansion ability (a surrogate of successful CAR T-cell manufacture) at diagnosis varied by disease type; T cells from patients with most lymphomas and solid tumors had lower expansion capability compared with standard-risk ALL. T-cell expansion was also differentially affected by multiple cycles of chemotherapy, though this might reflect differences in intensity of treatment regimens in addition to underlying differences in disease biology. The proportion of naïve and stem memory T cells directly correlated with T-cell expansion potential across all disease types, but there were also intrinsic differences in these populations among different cancers, and most chemotherapy regimens caused a decline. These findings suggest that intrinsic and treatment-induced differences in T-cell distribution and proliferative capacity should be considered in efforts to expand use of adoptive cell therapies in pediatric cancers.

See article, p. 500

Cytology and cystoscopy are the current standard procedures for monitoring recurrence in patients with bladder cancer, but the former has a low sensitivity and the latter is invasive. Recent studies have reported the assessment of cell-free DNA (cfDNA) in urine from patients with bladder cancer using an amplicon-based approach. To detect early-stage and recurrent disease in patients with bladder cancer, Dudley and colleagues modified their previously reported ultrasensitive cfDNA detection method, CAPP-Seq, to profile cfDNA in urine samples (uCAPP-Seq) at the time of diagnosis from 54 patients with bladder cancer (with 34 normal controls) and after the earliest post-treatment timepoint from 37 patients with recurrent disease and 27 patients who remained disease-free at least 9 months post-treatment. Overall concordance between mutations in urine tumor (utDNA) and tumors was high, more so for driver mutations; similarly, uCAPP-Seq identified mutations in the PLEKHS1 promoter in 46% of cases, which is comparable to previous findings. Both tumor-naïve uCAPP-Seq (sensitivity = 83%) and tumor-informed uCAPP-Seq (sensitivity = 93%) exhibited higher sensitivity and high specificity compared to standard cytology (sensitivity = 14%) in detecting early-stage bladder cancer, and tumor-naïve uCAPP-Seq profiling of prospectively collected urine samples exhibited approximately 2-fold greater sensitivity for detecting recurrent disease compared to cytology, cystoscopy, or diagnostic FISH. Further, uCAPP-Seq detected 92% of recurrences by a median of 2.7 months before clinical recurrence. Thus, detection of utDNA by uCAPP-Seq may be a noninvasive and more sensitive method to detect and surveil bladder cancer.

See article, p. 510

Clear cell renal cell carcinoma (ccRCC), the most lethal form of RCC, has been treated with first-line tyrosine kinase inhibitors (TKI), but there is a lack of available biomarkers to predict responses. Hakimi, Voss, Kuo, and colleagues evaluated predictors of TKI efficacy in a large cohort of patients with metastatic RCC who were enrolled in the phase III COMPARZ trial, which had compared the efficacy and safety of first-line pazopanib and sunitinib. Integration of somatic mutations, gene expression, and protein expression with clinical outcomes revealed four biologically distinct molecular subtypes associated with differences in response and survival following treatment with TKIs. Characterization of these subtypes identified angiogenesis and macrophage infiltration as critical factors for predicting TKI response. High angiogenesis gene expression was associated with improved outcome of patients treated with TKI and correlated with mutation status of PBRM1 and BAP1, known RCC driver genes. In addition, the group with the worst outcome after TKI treatment exhibited infiltrating PD-L1 positive cells as well as infiltration of M2 macrophages, suggesting the correlation of poor response with immune infiltrates and an immunosuppressed tumor microenvironment. Integrated analysis of angiogenesis and macrophage infiltration proposed low angiogenesis gene expression and high macrophage infiltration as indicators of the worst outcomes, but different interactions with the tumor microenvironment were identified for each TKI. Collectively, these results demonstrate the importance of angiogenesis and macrophage infiltration as predictors of clinical outcome for TKI therapy and can potentially guide future development of personalized treatment of patients with metastatic RCC.

See article, p. 526

Combination treatment with BRAF and MEK inhibitors is the standard of care for BRAF-mutant melanoma, but patients with BRAF-mutant melanoma eventually relapse, and only limited responses are observed in patients with mutated NRAS or NF1 genes. Maertens and colleagues observed that the combination of BRAF, MEK, and HDAC inhibitors was effective in BRAF-, NRAS-, or NF1-mutant melanomas. The class I HDAC inhibitor entinostat significantly improved the efficacy of BRAF/MEK inhibitors in multiple in vivo models of melanoma with a range of sensitivities to these agents; these effects were HDAC3-dependent. Expression of MGMT was identified as a biomarker for sensitivity to combination MAPK/HDAC inhibition and distinguished melanomas with global reduction of DNA repair gene expression, suggesting that latent defects in DNA repair underlie sensitivity to MAPK/HDAC inhibition. Mechanistically, BRAF/MEK inhibitors suppressed DNA repair genes mostly involved in the homologous recombination (HR) pathway, whereas the addition of entinostat induced the suppression of genes essential to the nonhomologous end-joining (NHEJ) pathway. This coordinate suppression of HR and NHEJ genes was required for cell death induced by the combination therapy and cooperatively triggered the accumulation of excessive DNA damage leading to chemical synthetic lethality; this was due at least in part to inhibition of phosphorylation of the transcription factor ELK1, which regulates DNA repair gene expression. Together, these findings pave the way for the clinical development of a potential therapeutic strategy with a clinically available biomarker for patients with RAS pathway–mutated melanoma.

See article, p. 546

A primary mechanism of immune escape involves altered expression of major histocompatibility complex (MHC) proteins, whose role in antigen presentation to T cells is critical to the success of immunotherapy. Although loss of MHC expression is well documented in various cancers, the precise molecular mechanisms that drive its loss remain poorly understood. Ennishi, Takata, and colleagues took a multiomics approach in 347 cases of diffuse large B-cell lymphoma (DLBCL) to show that transcriptional regulators and several components of both MHC-I and MHC-II complexes were mutated or deleted. In germinal center B cell–like (GCB) DLBCL, loss of MHC-II correlated with altered expression of hundreds of genes, particularly those associated with the centroblast-rich dark zone (DZ), suggesting that MHC-II loss defines a molecular subgroup of GCB-DLBCL derived from DZ B cells. Loss of MHC-II expression in DZ-GCB-DLBCL was associated with poor prognosis and reduced enrichment of CD4⁺ and CD8⁺ T cells, tumor-infiltrating lymphocytes, and regulatory T cells, and reduced cytolytic activity in the tumor microenvironment. In GCB-DLBCL, EZH2 was the most frequently mutated gene in samples where either MHC-I or MHC-II expression was lost. In two mouse models of DLBCL, Ezh2-mutant tumors expressed reduced levels of MHC-I and MHC-II, contained reduced amounts of T-cell infiltrates, and resulted in reduced survival. Inhibition of mutant EZH2 restored MHC expression in DLBCL cell lines by reducing repressive H3K27me3 marks at the promoters of MHC transactivators. These results demonstrate that EZH2 epigenetically regulates MHC expression and that its pharmacologic inhibition can potentially be used to restore patient responsiveness to immunotherapy.

Note: In This Issue is written by Cancer Discovery editorial staff. Readers are encouraged to consult the original articles for full details.