BRAF Mutations Are Oncogenic and Targetable Drivers of Colorectal NETs
See article, p. 594.
Oncogenic BRAF mutations were identified at a frequency of 9% in high-grade colorectal NETs.
Two patients with BRAFV600E -positive NETs had sustained responses to BRAF/MEK inhibition.
BRAF mutations define a colorectal NET subset, and BRAFV600E is sensitive to BRAF/MEK inhibitors.
Poorly differentiated neuroendocrine tumors (NET) of the digestive tract are highly heterogeneous and aggressive. The molecular features of NETs are not well characterized, limiting therapeutic options. Klempner and colleagues retrospectively identified recurrent BRAF mutations in 10 of 109 colorectal NETs (9%) from a large database of clinical samples with available genetic profiling information. The majority of these mutations were the V600E substitution, which is prevalent in other tumor types such as melanoma and for which approved targeted therapies exist, including vemurafenib and dabrafenib. BRAF mutations in this data set were mutually exclusive with NRAS mutations, KRAS mutations, and other established oncogenic drivers. Furthermore, two patients with metastatic high-grade colorectal NETs harboring BRAFV600E had dramatic and sustained clinical responses to combination BRAF/MEK inhibitor therapy despite being resistant to standard therapies. In the first case, symptoms resolved 10 days after BRAF/MEK inhibitor therapy was initiated, which coincided with reduced urinary BRAFV600E-positive tumor DNA and dramatic radiographic response. Clinical response for this patient was sustained beyond 7 months. Similarly, BRAF/MEK combination therapy induced symptom relief in a second case within 3 days after therapy initiation, followed by radiographic response sustained beyond 9 months. Taken together, these results provide evidence that mutant BRAF is an oncogenic driver in a subset of aggressive colorectal NETs and that the BRAFV600E mutation can be successfully targeted with combination BRAF/MEK inhibitor therapy to achieve sustained clinical responses in therapy-resistant patients.
Oncogenic EGFR Fusion Proteins May Be Therapeutic Targets in NSCLC
See article, p. 601.
EGFR fusions, most commonly EGFR–RAD51, were identified in patients with metastatic lung cancer.
Patients with EGFR fusion–positive NSCLC had clinical and radiographic responses to EGFR inhibition.
EGFR–RAD51 is oncogenic and can be successfully targeted with EGFR TKIs and antibodies in vitro.
EGFR mutations that lead to constitutive tyrosine kinase activity are predictive biomarkers of sensitivity to EGFR tyrosine kinase inhibitor (TKI) treatment in patients with non–small cell lung cancer (NSCLC). Konduri and colleagues analyzed a large cohort of NSCLC cases with available next-generation sequencing information and identified EGFR fusion events in five patients. In four of the five cases, the EGFR was fused to RAD51. A fusion between EGFR and PURB was identified in the fifth patient. All fusions contained the entire ligand binding domain and kinase domain, but lacked the C-terminus of EGFR. Four of the patients were treated with the EGFR inhibitor erlotinib, resulting in clinical and radiographic responses. Further investigations revealed that the EGFR–RAD51 fusion is both oncogenic in vitro and able to drive downstream signaling through the MAPK and PI3K–AKT pathways, despite the lack of known autophosphorylation sites in the EGFR C-terminus which mediate binding to signaling adaptor proteins. Furthermore, computational modeling suggested that EGFR–RAD51 proteins can form active dimers. Cells expressing the EGFR–RAD51 fusion were sensitive to FDA-approved EGFR inhibitors, including TKIs and the therapeutic antibody cetuximab. Taken together, these results establish that EGFR fusion proteins are oncogenic and are therapeutically targetable in patients with NSCLC.
BRD4 Enhancer Binding Drives Immune Surveillance During Senescence
See article, p. 612.
BRD4 binding at superenhancers drives SASP gene expression during oncogene-induced senescence.
BRD4 is required for SASP-mediated paracrine signaling and immune-mediated clearance of senescent cells.
BET inhibitors may disrupt tumor suppressive functions mediated by BRD4 binding at superenhancers.
In response to certain types of stress, including oncogene activation, cells can undergo senescence, which is marked not only by cell-cycle arrest but also by activation of the senescence-associated secretory phenotype (SASP), which can promote mobilization of immune cells and subsequent immune-mediated clearance of premalignant or malignant senescent cells. Senescence is also associated with broad changes in chromatin organization, prompting Tasdemir, Banito, Roe, Alonso-Curbelo, and colleagues to evaluate the role of enhancer dynamics in oncogene-induced senescence. Genome-wide chromatin profiling of normal human fibroblasts under normal proliferating conditions or following HRAS-driven senescence showed broad changes of the H3K27ac active chromatin mark at enhancers, particularly clusters of enhancers known as superenhancers, during senescence. In line with this observation, senescence-activated superenhancers were enriched for those associated with SASP genes. Robust upregulation and increased binding of the bromo and extraterminal domain (BET) protein BRD4, a transcriptional coactivator known to regulate expression of superenhancer-associated genes, was observed at superenhancer-associated SASP genes in senescent cells. SASP genes were regulated in a BRD4-dependent manner during oncogene- or chemotherapeutic-induced senescence, and genetic or pharmacologic suppression of BRD4 inhibited SASP-mediated cytokine secretion and paracrine senescence signaling in vitro and immune cell-mediated clearance of oncogenic NRAS-expressing cells in vivo. Together, these data show that BRD4-regulated superenhancers, which have primarily been associated with oncogenic functions, can regulate tumor suppressive gene expression programs such as the SASP, and raise the possibility that BET inhibitors in clinical development may have unintended consequences on senescence-mediated immune surveillance.
Neutrophils Contribute to Intermediate Stages of Invasion–Metastasis
See article, p. 630.
CD11b+/Ly6G+ neutrophils inhibit NK cell–mediated clearance of intraluminal tumor cells.
Neutrophils facilitate extravasation via secretion of IL1β and matrix metalloproteinases.
The tumor-derived cytokine G-CSF acts systemically on different stages of invasion–metastasis.
The role of neutrophils in the cancer invasion–metastasis cascade is unclear, as neutrophils have demonstrated both tumor-promoting and tumor-suppressing effects. Neutrophil-derived matrix metalloproteinases (MMP) promote local invasion and intravasation in the early stages of the invasion–metastasis cascade, whereas some reports suggest that neutrophils that accumulate in the lung prior to metastasis formation have an antimetastatic effect. Spiegel and colleagues evaluated the role of neutrophils in the intermediate stages of the invasion–metastasis cascade, including transport and extravasation. Splenectomy reduced the number of pulmonary metastases in the murine mammary carcinoma 4T1 tumor model with associated neutrophilia, and G-CSF overexpression phenocopied the neutrophilia and metastatic potential of 4T1 cells. Further experiments revealed that G-CSF–induced neutrophils acted to promote metastasis after intravasation but before postextravasation colonization. Similarly, host neutrophils were able to facilitate metastasis to the lungs during the immediate period after intravascular tumor cell introduction. Mechanistically, neutrophils protected cancer cells from natural killer (NK) cell–mediated clearance while still within the lumina of the microvessels and released MMP9 and IL1β, which acted on endothelial cells to induce transendothelial migration of cancer cells into the lung parenchyma. Taken together, these findings describe a transient metastasis-promoting role of systemic neutrophils during the intravascular stage.
In Vivo Genetic Screens Identify Nonmutated Drivers of Melanoma
See article, p. 650.
Functional genetic screens reveal interpatient heterogeneity in essential melanoma growth genes.
KMT2D promotes tumor growth specifically in NRAS-mutant melanomas via enhancer dysregulation.
Essential epigenetic genes driving melanoma can be activated independent of somatic mutations.
Approximately 50% of patients with melanoma harbor BRAF mutations, which can be targeted with BRAF inhibitors, but patients treated with BRAF inhibitors eventually experience disease progression. Therefore, it is essential to identify cancer drivers that may be exploitable in relapsed BRAF-mutant tumors as well as in melanomas lacking mutations in druggable targets. To identify drivers of melanoma, Bossi, Cicalese, and colleagues screened patient-derived xenografts (PDX) from patients with metastatic melanoma with a lentiviral shRNA library targeting 236 epigenetic modulator genes. Depletion of a specific shRNA in tumors suggested an essential role in tumor growth. The screen revealed a high degree of interpatient heterogeneity, with only approximately 15% of depleted shRNAs in common between the different PDX melanoma models. Despite not being mutated in the patient tumors, BAZ1B, SMARCA4, and CHD4 were essential in all PDXs tested, whereas KMT2D was essential only in NRAS-mutant PDXs, indicating that nonmutated genes can be essential in melanoma and suggesting that NRAS-mutant tumors may be specifically vulnerable to KMT2D depletion. Knockdown of KMT2D reduced growth and migration of NRAS-mutant melanoma cells and dysregulated specific enhancers to promote transcription of NRAS-mutant melanoma target genes. Taken together, these findings validate the use of PDX shRNA screens, reveal a high intertumor heterogeneity with activation of multiple nonredundant pathways, and suggest that numerous critical genes in each tumor can potentially be targeted to inhibit melanoma growth.
Select Biomarkers Can Predict Severe Cytokine Release Syndrome
See article, p. 664.
CRS severity correlates with peak levels of 24 cytokines and several standard clinical markers.
Predictive models based on cytokine levels accurately predict severe CRS before symptom onset.
Cytokine profiling may identify patients at high risk of CRS before critical illness develops.
Chimeric antigen receptor (CAR) T-cell therapy has produced remission rates as high as 90% in relapsed/refractory acute lymphoblastic leukemia (ALL) but can produce life-threatening toxicity via cytokine release syndrome (CRS). CRS can be treated with the IL6R inhibitor tocilizumab, but the mechanism by which it alleviates CRS is not well understood. To better understand and potentially predict CRS, Teachey and colleagues evaluated serial biomarker measurements in 51 patients (39 children and 12 adults) treated with engineered T cells targeting CD19 (CTL019). A total of 48 patients (94%) developed CRS. CRS severity correlated with peak levels of 24 cytokines (including IL6, IFNγ, sgp130, and sIL2Rα) and several standard biomarkers (including ferritin and CRP). Only a few cytokines (including IFNγ and sgp130) were differentially elevated during the first 3 days after infusion, before patients became critically ill. Based on these data, predictive models were developed that accurately identified which patients would develop severe CRS and validated in an independent cohort. Patients who developed severe CRS exhibited clinical, laboratory, and cytokine profiles similar to patients with hematophagocytic lymphohistiocytosis. Overall, these findings reveal the cytokine profile underlying CRS, and indicate that in patients receiving CAR T-cell therapy, CRS can be predicted before patients become critically ill based on the levels of several cytokines.
Note: In This Issue is written by Cancer Discovery editorial staff. Readers are encouraged to consult the original articles for full details.