Search

Purpose: To create an in vivo model of PDGFRA D842V-mutant gastrointestinal stromal tumor (GIST) and identify the mechanism of tumor persistence following avapritinib therapy. Experimental Design: We created a patient-derived xenograft (PDX) of PDGFRA D842V-mutant GIST and tested the effects of imatinib, avapritinib, and ML-7, an inhibitor of myosin light-chain kinase (MYLK). Bulk tumor RNA sequencing and oncogenic signaling were evaluated. Apoptosis, survival, and actin cytoskeleton were evaluated in GIST T1 cells and isolated PDX cells in vitro. Human GIST specimens were analyzed for MYLK expression. Results: The PDX was minimally responsive to imatinib but sensitive to avapritinib. Avapritinib therapy increased tumor expression of genes related to the actin cytoskeleton, including MYLK. ML-7 induced apoptosis and disrupted actin filaments in short-term cultures of PDX cells and decreased survival in GIST T1 cells in combination with imatinib or avapritinib. Combined therapy with ML-7 improved the antitumor effects of low-dose avapritinib in vivo. Furthermore, MYLK was expressed in human GIST specimens. Conclusions: MYLK upregulation is a novel mechanism of tumor persistence after tyrosine kinase inhibition. Concomitant MYLK inhibition may enable the use of a lower dose of avapritinib, which is associated with dose-dependent cognitive side effects.

The mammalian Vestigial-like (VGLL) transcriptional cofactor family of proteins VGLL1-4 has recently emerged as an important player in the tumorigenesis of diverse neoplasms. The role of VGLL3 in soft tissue tumors is exemplified by its amplification in myxoinflammatory fibroblastic sarcoma and its rearrangement (fused to CHD7, CHD9, or MAMLD1) in hybrid schwannoma-perineurioma. This study characterizes a distinctive low-grade myogenic neoplasm with a striking predilection for the head and neck, characterized by VGLL3 fusions. The study includes five males and one female patient, aged 30-71 years (median, 56). Three tumors originated in the tongue, with one case each in the nasopharynx, oral cavity, and oropharynx. The VGLL3 fusion partners included TCF12 (n = 3), EP300 (n = 2), and PPARGC1A (n = 1). The tumor size range was 0.8-1.6 cm (all, but one, was1 cm). Histologically, all tumors displayed bland spindle to ovoid cells arranged into vague fascicular and diffuse patterns. Mitotic activity ranged from 1 to 7 per 10 HPFs. Five tumors were muscle-centered and infiltrative, and one was centered beneath nasopharyngeal mucosa. Immunohistochemistry revealed consistent expression of desmin (diffuse in four and patchy in two cases) associated with patchy smooth muscle actin expression (4/6), and focal reactivity for myogenin (5/6) and myoD1 (1/3). All patients were managed surgically; one patient each received adjuvant radio- or chemotherapy. Three patients with follow-up were without disease at 8, 19, and 60 months and one was alive with unknown disease status at 24 months. All VGLL3 fusions were in-frame and involved exon 2, fused with either TCF12 exon 16, EP300 exon 31, or PPARGC1A exon 5, respectively. This series characterizes a distinctive subset of spindle cell rhabdomyosarcoma (RMS) with a predilection for the head and neck in adults, defined by VGLL3 fusions, likely indolent behavior and limited rhabdomyoblastic differentiation. Further delineation of this entity and differentiation from more aggressive molecular subtypes of spindle cell RMS is mandatory to define the most appropriate therapeutic strategy and avoid overtreatment.

To elucidate the mechanisms underlying the divergent clinicopathologic spectrum of EWSR1/FUS::CREB translocation-associated tumors, we performed a comprehensive genomic analysis of fusion transcript variants, recurrent genetic alterations (mutations, copy number alterations), gene expression, and methylation profiles across a large cohort of tumor types. The distribution of the EWSR1/FUS fusion partners-ATF1, CREB1, and CREM-and exon involvement was significantly different across different tumor types. Our targeted sequencing showed that secondary genetic events are associated with tumor type rather than fusion type. Of the 39 cases that underwent targeted NGS testing, 18 (46%) had secondary OncoKB mutations or copy number alterations (29 secondary genetic events in total), of which 15 (52%) were recurrent. Secondary recurrent, but mutually exclusive, TERT promoter and CDKN2A mutations were identified only in clear cell sarcoma (CCS) and associated with worse overall survival. CDKN2A/B homozygous deletions were recurrent in angiomatoid fibrous histiocytoma (AFH) and restricted to metastatic cases. mRNA upregulation of MITF, CDH19, PARVB, and PFKP was found in CCS, compared to AFH, and correlated with a hypomethylated profile. In contrast, S100A4 and XAF1 were differentially upregulated and hypomethylated in AFH but not CCS. Unsupervised clustering of methylation profiles revealed that CREB family translocation-associated tumors form neighboring but tight, distinct clusters. A sarcoma methylation classifier was able to accurately match 100% of CCS cases to the correct methylation class; however, it was suboptimal when applied to other histologies. In conclusion, our comprehensive genomic profiling of EWSR1/FUS::CREB translocation-associated tumors uncovered mostly histotype, rather than fusion-type associated correlations in transcript variants, prognostically significant secondary genetic alterations, and gene expression and methylation patterns.

We report the first case of a primary renal undifferentiated sarcoma harboring an SS18::POU5F1 gene fusion. The patient was a 38 year-old male diagnosed with a 5 cm renal tumor which invaded the adrenal gland and extended into the renal vein. Microscopically, the neoplasm had a predominantly undifferentiated round cell morphology, with areas of rhabdoid and spindle cell growth. Similar to the previously reported cases with this fusion, by immunohistochemistry the neoplasm expressed S100 protein and epithelial markers (diffuse EMA, focal cytokeratin), suggesting the possibility of a myoepithelial phenotype. This report documents another example of a fusion-positive undifferentiated soft tissue sarcoma occurring as a primary renal neoplasm, adding to the already broad list of such entities. It highlights the crucial role of molecular analysis in establishing a specific diagnosis given the overlapping morphology and immunophenotypes such entities may exhibit.

Uterine PEComas often present a diagnostic challenge as they share morphological and immunohistochemical features with smooth muscle tumors. Herein we evaluated a series of 19 uterine PEComas to compare the degree of melanocytic marker expression with their molecular profile. Patients ranged from 32–77 (median 48) years, with six tumors classified as malignant based on the modified gynecologic-specific prognostic algorithm. All patients with malignant PEComas were alive with disease or dead of disease at last follow-up, while all those of uncertain malignant potential were alive and well (median follow-up, 47 months). Seventeen of 19 (89%) PEComas harbored either a TSC1 or TSC2 alteration. One of the two remaining tumors showed a TFE3 rearrangement, but the other lacked alterations in all genes evaluated. All showed at least focal (usually strong) positivity for HMB-45, with 15/19 (79%) having >50% expression, while the tumor lacking TSC or TFE3 alterations was strongly positive in 10% of cells. Melan-A and MiTF were each positive in 15/19 (79%) tumors, but staining extent and intensity were much more variable than HMB-45. Five of six (83%) malignant PEComas also harbored alterations in TP53, ATRX, or RB1, findings not identified in any tumors of uncertain malignant potential. One malignant PEComa was microsatellite-unstable/mismatch repair protein-deficient. In summary, TSC alterations/TFE3 fusions and diffuse (>50%) HMB-45 expression are characteristic of uterine PEComas. In morphologically ambiguous mesenchymal neoplasms with myomelanocytic differentiation, especially those with metastatic or recurrent disease, next-generation sequencing is recommended to evaluate for TSC alterations; as such, patients can be eligible for targeted therapy.

Purpose: Because the Hedgehog and Notch pathways are often overexpressed in mesenchymal malignancies, we evaluated the efficacy of concurrent inhibition of Notch and Hedgehog signaling using the gamma-secretase inhibitor (GSI) RO4929097 and the smoothened antagonist vismodegib in unresectable or metastatic sarcoma. Patients and Methods: In this investigator-initiated trial, phase Ib used standard 3+3 dose escalation in which patients first received vismodegib once daily for 21 days, followed by the combination of RO4929097 concurrently with vismodegib in 21-day cycles. In phase II, patients were randomized to RO4929097 alone or in combination with vismodegib. Results: Nine patients were treated in phase Ib with no dose-limiting toxicities. RO4929097 at 15 mg daily in combination with 150 mg daily of vismodegib was declared the recommended phase II dose. Most adverse events were grade ≤ 2. In phase II (closed early due to discontinuation of RO4929097 evaluation), 34 patients were randomized to RO4929097 alone and 33 to RO4929097 plus vismodegib. RO4929097 did not interfere with the steady-state concentration of vismodegib, while vismodegib reduced the plasma concentration of RO492909. No patients had an objective response. Neither progression-free nor overall survival differed significantly between treatment arms. Paired tumor biopsies from a subset of patients demonstrated inhibition of cleaved Notch. Conclusions: The combination of RO4929097 plus vismodegib was generally well tolerated. Although accrual to this study was not completed, vismodegib did not meaningfully enhance the clinical efficacy of RO4929097 in an unplanned analysis. GSIs and GSIs plus vismodegib can inhibit intratumoral Notch and downstream phosphorylated Akt signaling.

Purpose: This phase Ib trial was designed to evaluate the safety and early efficacy signal of the combination of imatinib and binimetinib in patients with imatinib-resistant advanced gastrointestinal stromal tumors (GISTs). Patients and Methods: This trial used a standard 3 + 3 design to determine the recommended phase II dose (RP2D). Additional patients were enrolled on an expansion cohort at the RP2D enriching for succinate dehydrogenase (SDH)-deficient GISTs to explore potential efficacy. Results: The trial enrolled nine patients in the dose-escalation cohort and 14 in the dose-expansion cohort including six with SDH-deficient GISTs. Imatinib 400 mg daily with binimetinib 45 mg twice daily was established as the RP2D. Dose-limiting toxicity (DLT) was asymptomatic grade 4 creatinine phosphokinase (CPK) elevation. The most common non-DLT grade 3/4 toxicity was asymptomatic CPK elevation (69.6%). Other common ≥grade 2 toxicities included peripheral edema (17.4%), acneiform rash (21.7%), anemia (30.4%), hypophosphatemia (39.1%), and aspartate aminotransferase (AST) increase (17.4%). Two serious adverse events occurred (grade 2 dropped head syndrome and grade 3 central retinal vein occlusion). No unexpected toxicities were observed. Limited clinical activity was observed in KIT-mutant GIST. For SDH-deficient GISTs, one of five had confirmed RECIST1.1 partial response (PR). The median progression-free survival (mPFS) in patients with SDH-deficient GIST was 45.1 months [95% confidence interval (CI), 15.8–not estimable (NE)]; the median overall survival (mOS) was not reached (95% CI, 31.6 months–NE). One patient with a refractory metastatic SDH-deficient GIST had an exceptional pathologic response and durable clinical benefit. Conclusions: The combination of imatinib and binimetinib is safe with manageable toxicity and has encouraging activity in SDH-deficient but not imatinib-refractory KIT/PDGFRA-mutant GISTs. The observed clinical benefits provide a motivation for a larger trial of the combination strategy in SDH-deficient GISTs.

Supplemental Figure 5.Short term culture of cells derived from PDGFRA D842V PDXs are tumor cells.

Normalized Kynurenine Concentrate Over Time

Supplemental Figure 1.

Normalized Tryptophan Concentrate Over Time

Purpose:Epacadostat, an indole 2,3 dioxygenase 1 (IDO1) inhibitor, proposed to shift the tumor microenvironment toward an immune-stimulated state, showed early promise in melanoma but has not been studied in sarcoma. This study combined epacadostat with pembrolizumab, which has modest activity in select sarcoma subtypes.Patients and Methods:This phase II study enrolled patients with advanced sarcoma into five cohorts including (i) undifferentiated pleomorphic sarcoma (UPS)/myxofibrosarcoma, (ii) liposarcoma (LPS), (iii) leiomyosarcoma (LMS), (iv) vascular sarcoma, including angiosarcoma and epithelioid hemangioendothelioma (EHE), and (v) other subtypes. Patients received epacadostat 100 mg twice daily plus pembrolizumab at 200 mg/dose every 3 weeks. The primary endpoint was best objective response rate (ORR), defined as complete response (CR) and partial response (PR), at 24 weeks by RECIST v.1.1.Results:Thirty patients were enrolled [60% male; median age 54 years (range, 24–78)]. The best ORR at 24 weeks was 3.3% [PR, n = 1 (leiomyosarcoma); two-sided 95% CI, 0.1%–17.2%]. The median PFS was 7.6 weeks (two-sided 95% CI, 6.9–26.7). Treatment was well tolerated. Grade 3 treatment-related adverse events occurred in 23% (n = 7) of patients. In paired pre- and post-treatment tumor samples, no association was found between treatment and PD-L1 or IDO1 tumor expression or IDO-pathway–related gene expression by RNA sequencing. No significant changes in serum tryptophan or kynurenine levels were observed after baseline.Conclusions:Combination epacadostat and pembrolizumab was well tolerated and showed limited antitumor activity in sarcoma. Correlative analyses suggested that inadequate IDO1 inhibition was achieved.

Representativeness of study participants.

Supplemental Figure 8.Relative MYLK mRNA expression by qPCR in GIST PDGFRA D842V PDXs treated with vehicle(Veh) or imatinib (IM) for 2 weeks.

Supplemental Figure 3. GIST PDGFRA D842V PDXs are not sensitive to crenolanib.

Supplemental Figure 7.

Supplemental Figure 4.

Purpose:To create an in vivo model of PDGFRA D842V-mutant gastrointestinal stromal tumor (GIST) and identify the mechanism of tumor persistence following avapritinib therapy.Experimental Design:We created a patient-derived xenograft (PDX) of PDGFRA D842V-mutant GIST and tested the effects of imatinib, avapritinib, and ML-7, an inhibitor of myosin light-chain kinase (MYLK). Bulk tumor RNA sequencing and oncogenic signaling were evaluated. Apoptosis, survival, and actin cytoskeleton were evaluated in GIST T1 cells and isolated PDX cells in vitro. Human GIST specimens were analyzed for MYLK expression.Results:The PDX was minimally responsive to imatinib but sensitive to avapritinib. Avapritinib therapy increased tumor expression of genes related to the actin cytoskeleton, including MYLK. ML-7 induced apoptosis and disrupted actin filaments in short-term cultures of PDX cells and decreased survival in GIST T1 cells in combination with imatinib or avapritinib. Combined therapy with ML-7 improved the antitumor effects of low-dose avapritinib in vivo. Furthermore, MYLK was expressed in human GIST specimens.Conclusions:MYLK upregulation is a novel mechanism of tumor persistence after tyrosine kinase inhibition. Concomitant MYLK inhibition may enable the use of a lower dose of avapritinib, which is associated with dose-dependent cognitive side effects.

Supplemental Figure 6.Cell survival of GIST T1 cells treated with imatinib or avapritinib for 72h in complete media.Each data point represents an average of 10 wells.

Supplementary Data from PRC2-Inactivating Mutations in Cancer Enhance Cytotoxic Response to DNMT1-Targeted Therapy via Enhanced Viral Mimicry

Polycomb repressive complex 2 (PRC2) has oncogenic and tumor-suppressive roles in cancer. There is clinical success of targeting this complex in PRC2-dependent cancers, but an unmet therapeutic need exists in PRC2-loss cancer. PRC2-inactivating mutations are a hallmark feature of high-grade malignant peripheral nerve sheath tumor (MPNST), an aggressive sarcoma with poor prognosis and no effective targeted therapy. Through RNAi screening in MPNST, we found that PRC2 inactivation increases sensitivity to genetic or small-molecule inhibition of DNA methyltransferase 1 (DNMT1), which results in enhanced cytotoxicity and antitumor response. Mechanistically, PRC2 inactivation amplifies DNMT inhibitor–mediated expression of retrotransposons, subsequent viral mimicry response, and robust cell death in part through a protein kinase R (PKR)–dependent double-stranded RNA sensor. Collectively, our observations posit DNA methylation as a safeguard against antitumorigenic cell-fate decisions in PRC2-loss cancer to promote cancer pathogenesis, which can be therapeutically exploited by DNMT1-targeted therapy.Significance:PRC2 inactivation drives oncogenesis in various cancers, but therapeutically targeting PRC2 loss has remained challenging. Here we show that PRC2-inactivating mutations set up a tumor context–specific liability for therapeutic intervention via DNMT1 inhibitors, which leads to innate immune signaling mediated by sensing of derepressed retrotransposons and accompanied by enhanced cytotoxicity.See related commentary by Guil and Esteller, p. 2020.This article is highlighted in the In This Issue feature, p. 2007

AFS98 depletion

Figure S1. The relative mRNA expression of DKK4 and CTNNB1 in human KIT+ cells.

Human GIST characteristics

Primary untreated KIT+ vs. Metastatic resistant KIT+, pathways with P< 0.05.

Supplementary Figure Legends

PDF file - 576K

Figure S1. Expression of forkhead family transcription factors in GIST; S2. Motif analysis of the FOXF1 cistrome and validation of FOXF1 binding sites in GIST 48 cells; S3. FOXF1 regulates ETV1, KIT, and ETV1- and KIT-dependent transcriptome as well as cell cycle regulated genes; S4. Representative ChIP-seq and ATAC-seq profiles with siRNAmediated downregulation of ETV1 and FOXF1 in GIST48 cells; S5. FOXF1 regulates KIT expression and signaling; S6. FOXF1 regulates cell cycle progression and is required for growth and proliferation of GIST cells in vitro; S7. Generation of the Foxf1f-HA-Foxf1 allele; S8. Foxf1 expression in mouse GI tract and Foxf1 ablation reduces mutant Kit-mediated hyperplasia of mouse large intestine.

S1. De novo motifs in GIST-Specific ETV1 enhancers; S2. De novo motifs in Prostate Specific ETV1 enhancers; S3. De novo motifs in Shared ETV1 enhancers; S4. De novo motifs in Shared ETV1 promoters; S5: Expression of Forkhead Transcription Factors in GIST and non-GIST tumors from GENT dataset; S6: Custom Gene Sets Used for GSEA; S7: GSEA results of gene sets enriched among downregulated genes in GIST48 transfected with siFOXF1 compared with siScrambled; S8: GSEA results of gene sets enriched among downregulated genes in GIST48 treated with siETV1

Supplementary Figure 1: Etv1 is required for intramuscular and myenteric ICC development. Supplementary Figure 2: Etv1 ablation with tamoxifen decreases Etv1 and Kit expression in Kit^delta558/+;Etv1^flox/flox;Rosa26^CreERT2/CreERT2 mice. Supplementary Figure 3: Etv1 positively regulates Kit expression in murine stomach. Supplementary Figure 4: ETV1 regulates KIT expression in GIST48 cells. Supplementary Figure 5: MAP kinase pathway inhibition decreases ETV1 target gene expression. Supplementary Figure 6: Dual lineage inhibition inhibits tumor initiation in vivo.

PDF file - 171KB, Supplementary Fig. S2. NUT is translocated to the NSD3 locus in 1221 cell line as demonstrated by FISH on metaphase spread.

Supplementary Table 2: Primary untreated (Prim/UT) KIT+ ( n = 8 ) vs. Metastatic resistant (Met/Res) KIT+ ( n = 4). Significant genes with a False Discovery Rate of < 0.05 and a fold change > 2.0

Supplementary Table 2: Gene sets enriched among genes DOWNregulated by tamoxifen treatment (Etv1 ablation) in murine GIST-like tumors.

Supplementary Table 3: Gene sets enriched among genes UPregulated by tamoxifen treatment (Etv1 ablation) in murine GIST-like tumors.

PDF file - 304KB, Supplementary Fig. S4. Among several ET-interacting proteins, only NSD3 is required for the blockade of differentiation in NMC. (A) Quantitative RT-PCR of NSD3, JMJD6, GLTSCR1, and ATAD5 levels 24 h following transfection of siRNAs. (B) Immunoblots of TC-797 lysates 120h following transfection with siRNAs.

Tyrosine kinase inhibition of gastrointestinal stromal tumors (GIST) is effective but typically culminates in resistance and is rarely curative. Immunotherapy has potential application to GIST, as we previously showed that T-cell checkpoint blockade increases the antitumor effects of imatinib. Here, we showed that ligation of CD40 using an agonistic antibody (anti-CD40) activated tumor-associated macrophages (TAMs) in vivo in a knock-in mouse model of GIST harboring a germline mutation in Kit exon 11. Activated TAMs had greater TNFα production and NFκB signaling and directly inhibited tumor cells in vitro. Anti-CD40 required concomitant therapy with imatinib for efficacy and depended on TAMs, and to a lesser extent CD8+ T cells, but not on CD4+ T cells or B cells. In an analysis of 50 human GIST specimens by flow cytometry, we found that CD40 was expressed on human TAMs and tumor cells yet was downregulated after response to imatinib. CD40 ligation did not have a direct inhibitory effect on human GIST cells. Our findings provide the rationale for combining anti-CD40 and tyrosine kinase inhibition to treat human GIST. Cancer Immunol Res; 6(4); 434–47. ©2018 AACR.

PDF file - 213KB, Supplementary Fig. S3. Wild type NSD3 is required for the blockade of differentiation in BRD4-NUT-expressing NMC cells, 8645 and PER-403.

Supplementary Table 1: Differentially expressed genes (p1.5) between oil and tamoxifen treated tumors to conditionally ablate Etv1.

PDF file - 960K

PDF file - 140KB, Supplementary Fig. S1. Karyotype of the 1221 NSD3-NUT-positive cell line.

Supplementary Figure 1 from EWS-CREB1: A Recurrent Variant Fusion in Clear Cell Sarcoma—Association with Gastrointestinal Location and Absence of Melanocytic Differentiation

This file contains the details about methods not described in the article file

Supplemental Table 1. Adverse events observed and adjudicated as possibly, probably or definitely related to therapy (n = 19).

PDF file - 281K, Model of liposarcoma initiation and progression.

PDF file - 73K, Multivariate analysis of chromosome 19q loss and clinical factors in local recurrence.

Combination of homozygous BrafV600E activation and Trp53 loss induced dramatic ICC hyperplasia in stomach, cecum and large intestine, but not in small intestine.

PDF file - 205K, Progression associated copy number alterations.