See article, p. 252

  • Targeting FGFR in patients with FGFR2-fusion ICC induces polyclonal secondary FGFR2 mutations.

  • Analysis of cfDNA, tumors, and metastases reveals strong intra- and intertumor heterogeneity.

  • Next-generation FGFR inhibitors may overcome resistance to BGJ398 in patients with ICC.

graphic

FGFR signaling is often activated by genetic alterations in intrahepatic cholangiocarcinomas (ICC), and the FGFR inhibitor BGJ398 has demonstrated clinical activity in patients with FGFR2 fusion–positive ICC. However, drug resistance often develops, prompting Goyal, Saha, and colleagues to investigate the mechanism of resistance. BGJ398 induced tumor regression in 3 of 4 patients with FGFR2 fusion–positive ICC, which was followed by disease progression. Analysis of pre- and post-treatment cell-free circulating tumor DNA (cfDNA) identified FGFR mutations, including V564F, exclusively in the post-treatment cfDNA from all 3 responding patients, suggesting these alterations may be involved in acquired BGJ398 resistance. Whole-exome and RNA sequencing of serial tumor biopsies and multiple metastases obtained by rapid autopsy detected a subset of the FGFR2 mutations seen in cfDNA, and revealed intratumor heterogeneity with multiple FGFR2 resistance mutations arising in different cells or metastatic lesions. Structural modeling indicated that the V564F mutation would block BGJ398 binding to prevent FGFR2 inhibition. Other identified mutations did not directly contact BGJ398, but were predicted to destabilize the inactive kinase conformation required for BGJ398 binding. These findings were confirmed in cells expressing the TEL–FGFR2 fusion, which were sensitive to BGJ398 but rapidly acquired a subpopulation of resistant cells that harbored FGFR2 mutations. In a panel of structurally and functionally diverse FGFR inhibitors, different inhibitors were able to overcome specific resistance mutations, suggesting new treatment possibilities for resistant tumors. In addition to identifying diverse FGFR2 mutations that promote resistance to BGJ398, these findings suggest the possibility of overcoming resistance with structurally distinct FGFR inhibitors in patients with ICC.

See article, p. 264

  • A subset of mutation-associated neoantigens is lost at the onset of resistance to checkpoint blockade.

  • Eliminated neoantigens are immunogenic and their loss is associated with decreased TCR clonality.

  • Neoantigens are lost via elimination of tumor subclones or genetic deletion of truncal alterations.

graphic

Response to immune checkpoint blockade has been linked to an increased burden of nonsynonymous mutations, which may generate tumor-specific neoantigens that can activate antitumor immune responses. However, the efficacy of immune checkpoint inhibitors such as anti–PD-1 and anti–CTLA-4 antibodies is limited by the emergence of acquired resistance. To identify mechanisms underlying resistance to immune checkpoint inhibitors, Anagnostou, Smith, and colleagues analyzed pretreatment and post-relapse tumor samples from four patients with non–small cell lung cancer who developed acquired resistance following treatment with anti–PD-1 or combined anti–PD-1 and anti–CTLA-4. Whole-exome sequencing and computational modeling identified candidate mutation-associated neoantigens (MANA) with the potential to induce immune responses. Of note, a subset of these putative MANAs that were predicted to facilitate higher MHC binding affinity or to alter residues involved in T-cell receptor (TCR) binding were lost in tumors at the time of acquired resistance. Loss of these candidate MANAs occurred via either elimination of neoantigen-harboring tumor subclones or chromosomal deletion of truncal mutations. Stimulation of autologous T cells from each patient with peptides encoded by eliminated MANAs induced clonal expansion of neoantigen-specific T cells, supportive of functional immune responsiveness, and loss of these MANAs was associated with decreased cytotoxic TCR clonality. These results indicate that immune editing of immunogenic tumor neoantigens may contribute to acquired resistance to immune checkpoint blockade and suggest strategies for the development of personalized immunotherapies.

See article, p. 277

  • Diverse ESR1 somatic mutations were characterized in over 900 patients with metastatic breast cancer.

  • ESR1 LBD mutants exhibit differential levels of ER activation in a mutation site–specific manner.

  • AZD9496 is more potent against specific ESR1 mutants compared to the standard antagonist, fulvestrant.

graphic

Recurrent acquired mutations in the estrogen receptor (ER; encoded by ESR1) have been identified in patients with metastatic breast cancer; subsequently, a subset of ESR1 mutations have been shown to promote hormone-independent ER activation, but it is unknown whether all recurrent ESR1 mutations promote resistance to aromatase inhibitors or ER antagonists. To characterize the spectrum of ESR1 mutations present in metastatic breast cancer, Toy and colleagues performed next-generation sequencing of tumors from more than 900 patients with metastatic breast cancer. More than 10 percent of patients harbored somatic ESR1 mutations, all but one of which were located in the ligand binding domain (LBD) and 20 percent of which were uncommon; further, most of the infrequent mutations had not been previously identified. Many of the infrequent ESR1 LBD mutants constitutively induced ligand-independent luciferase activity, transcription, and phosphorylation, and estrogen-independent growth in vitro. Mutations in the tyrosine residue 537, as well as the adjacent residues, exhibited the greatest estrogen-independent activity. Several ESR1 LBD mutants led to reduced sensitivity to selective estrogen receptor degraders (SERD) in vitro, and the ESR1Y537S mutant in particular impaired the response to the SERD fulvestrant in vivo. By contrast, the newly developed SERD AZD9496 was potent against all ESR1 mutants in vitro and in vivo. Together, these results describe the comprehensive characterization of diverse ESR1 mutations and suggest that ER antagonists with superior pharmacokinetic properties may improve outcomes for patients with ER+ metastatic breast cancer.

See article, p. 288

  • OTX2 is a pioneer transcription factor that activates Group 3 medulloblastoma enhancers.

  • OTX2 cooperates with NEUROD1 to promote activation of enhancers including the NEK2 enhancer.

  • OTX2 or its target genes may be potential therapeutic targets in patients with Group 3 medulloblastoma.

graphic

Medulloblastoma has been divided into at least four molecular subgroups: WNT, SHH, Group 3, and Group 4. Group 3 tumors are the most aggressive, with the poorest outcomes and limited treatment options. To investigate gene-regulatory mechanisms that drive Group 3 medulloblastoma, Boulay and colleagues used chromatin immunoprecipitation sequencing and RNA sequencing to identify enhancers that were consistently activated in Group 3 medulloblastoma tumors. The majority of these enhancers were also active in medulloblastoma cell lines, and motif enrichment analysis showed the majority were bound by the homeobox transcription factor OTX2. OTX2-bound sites with the highest activity were also enriched for the transcription factor NEUROD1, and both proteins contributed to increased levels of acetylated histone 3 lysine 27 (H2K27ac), a mark of active enhancers. Further, OTX2 interacted with chromatin modifying complexes including EP300, MLL, and BAF and was able to function as a pioneer factor to induce chromatin opening and deposition of activating marks at medulloblastoma enhancers. In contrast, while NEUROD1 co-occupied and promoted activation of many OTX2-bound enhancers, it was less potent as a pioneer factor by itself. When OTX2 was depleted, changes in H3K27ac at OTX2 sites resulted in reduced expression of nearby genes including NEK2, which was associated with several OTX2-bound enhancers. Depletion or inhibition of NEK2, a mitotic kinase, reduced medulloblastoma cell viability. Altogether, these results identify OTX2 as a key regulator of the chromatin landscape of Group 3 medulloblastoma, and implicate OTX2 targets, including NEK2, as putative therapeutic targets in patients with medulloblastoma.

See article, p. 302

  • The MEK inhibitor trametinib induces an epigenetic upregulation of RTKs that blunts sensitivity.

  • Targeting MEK promotes genome-wide enhancer formation mediated by BRD4 and P-TEFb to upregulate RTKs.

  • P-TEFb inhibition may prevent the development of resistance to MEK inhibitors in patients with TNBC.

graphic

The large majority of basal-like triple-negative breast cancers (TNBC) harbor genomic amplifications that activate EGFR–KRAS–BRAF signaling. Treatment with the MEK 1/2 inhibitor trametinib induces initial growth arrest followed by activation of alternate kinase pathways to overcome inhibition. To investigate this adaptive response to trametinib treatment, Zawistowski, Bevill, and colleagues performed a seven-day window-of-opportunity trial in 6 patients with TNBC. RNA sequencing of pre- and post-treatment tumor samples found that up to 26% of tyrosine kinases were upregulated by trametinib, indicative of an adaptive bypass response whereby tyrosine kinase upregulation might contribute to MEK inhibitor resistance. In multiple preclinical models, trametinib induced epigenetic changes to activate target promoters and enhancers including the receptor tyrosine kinase (RTK) DDR1, which displayed increased levels of the activating histone marks H3K27ac and H3K4me1, and increased occupancy of BRD4, p300, and MED1. Further, MEK inhibition induced genome-wide enhancer formation that was blocked by inhibition of BRD4. BRD4 inhibitors synergized with trametinib to reduce the growth of TNBC cell lines and resensitized a resistant cell line to trametinib. Moreover, BRD4 inhibition reduced TNBC growth in vivo. BRD4 interacts with components of the P-TEFb transcriptional elongation regulatory complex, and, similar to BRD4 inhibition, inhibition of P-TEFb complex components potentiated trametinib-mediated growth suppression. Altogether, these findings indicate that epigenetic activation of RTKs may promote resistance to trametinib, and suggest that P-TEFb inhibitors may potentially bypass this resistance mechanism in patients with TNBC.

See article, p. 322

  • CREBBP opposes BCL6 function at GC enhancers to express key signaling and differentiation genes.

  • CREBBP haploinsufficiency cooperates with BCL2 dysregulation to promote B-cell lymphomagenesis.

  • CREBBP haploinsufficiency supports the possible development of therapies modulating acetylation.

graphic

Mutations frequently inactivate the CREBBP acetyltransferase in germinal center (GC)–derived cancers including diffuse large B-cell lymphoma (DLBCL) and follicular lymphoma (FL), but the role of CREBBP inactivation in lymphomagenesis has not been elucidated. Zhang and colleagues found that CREBBP primarily bound to enhancer regions in GC B cells to promote expression of genes essential for GC B-cell physiology. CREBBP co-occupied many genomic sites bound by BCL6 in GC B cells, consistent with the previously described role of BCL6 as a CREBBP substrate whose transcriptional repressor activity is impaired when acetylated. CREBBP and BCL6 had opposing functions on gene expression, suggesting that CREBBP may restrain pathogenic transcriptional repression by BCL6. Loss of CREBBP disrupted terminal B-cell differentiation and enhanced the proliferation of B cells. However, Crebbp deficiency was not sufficient to induce lymphoma in vivo, but Crebbp haploinsufficiency along with BCL2 overexpression (to mimic the BCL2 translocation common in FL and DLBCL) increased the incidence of B-cell malignancies that resembled human FL. Further, gene expression analysis of human DLBCLs revealed an underexpression of CREBBP signature genes, suggesting that these findings are relevant to human disease. The identification of CREBBP as a haploinsufficient tumor suppressor gene elucidates a mechanism by which CREBBP mutations may contribute to GC B-cell lymphoma genesis, and suggests the potential for the development of therapies that modulate acetylation in patients with GC-derived cancers.

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