NF1-Driven Rhabdomyosarcoma Phenotypes: A Comparative Clinical and Molecular Study of NF1-Mutant Rhabdomyosarcoma and NF1-Associated Malignant Triton Tumor.

PURPOSE: Alterations of the NF1 tumor suppressor gene is the second most frequent genetic event in embryonal rhabdomyosarcoma (ERMS), but its associations with clinicopathologic features, outcome, or coexisting molecular events are not well defined. Additionally, NF1 alterations, mostly in the setting of neurofibromatosis type I (NF1), drive the pathogenesis of most malignant peripheral nerve sheath tumor with divergent RMS differentiation (also known as malignant triton tumor [MTT]). Distinguishing between these entities can be challenging because of their pathologic overlap. This study aims to comprehensively analyze the clinicopathologic and molecular spectrum of NF1-mutant RMS compared with NF1-associated MTT for a better understanding of their pathogenesis. METHODS: We investigated the clinicopathologic and molecular landscape of a cohort of 22 NF1-mutant RMS and a control group of 13 NF1-associated MTT. Cases were tested on a matched tumor-normal hybridization capture-based targeted DNA next-generation sequencing. RESULTS: Among the RMS group, all except one were ERMS, with a median age of 17 years while for MTT the mean age was 39 years. Three MTTs were misdiagnosed as ERMS, having clinical impact in one. The most frequent coexisting alteration in ERMS was TP53 abnormality (36%), being mutually exclusive from NRAS mutations (14%). MTT showed coexisting CDKN2A/B and PRC2 complex alterations in 38% cases and loss of H3K27me3 expression. Patients with NF1-mutant RMS exhibited a 70% 5-year survival rate, in contrast to MTT with a 33% 5-year survival. All metastatic NF1-mutant ERMS were associated with TP53 alterations. CONCLUSION: Patients with NF1-mutant ERMS lacking TP53 alterations may benefit from dose-reduction chemotherapy. On the basis of the diagnostic challenges and significant treatment and prognostic differences, molecular profiling of challenging tumors with rhabdomyoblastic differentiation is recommended.

FGFR1 fusions as a novel molecular driver in rhabdomyosarcoma.

The wide application of RNA sequencing in clinical practice has allowed the discovery of novel fusion genes, which have contributed to a refined molecular classification of rhabdomyosarcoma (RMS). Most fusions in RMS result in aberrant transcription factors, such as PAX3/7::FOXO1 in alveolar RMS (ARMS) and fusions involving VGLL2 or NCOA2 in infantile spindle cell RMS. However, recurrent fusions driving oncogenic kinase activation have not been reported in RMS. Triggered by an index case of an unclassified RMS (overlapping features between ARMS and sclerosing RMS) with a novel FGFR1::ANK1 fusion, we reviewed our molecular files for cases harboring FGFR1-related fusions. One additional case with an FGFR1::TACC1 fusion was identified in a tumor resembling embryonal RMS (ERMS) with anaplasia, but with no pathogenic variants in TP53 or DICER1 on germline testing. Both cases occurred in males, aged 7 and 24, and in the pelvis. The 2nd case also harbored additional alterations, including somatic TP53 and TET2 mutations. Two additional RMS cases (one unclassified, one ERMS) with FGFR1 overexpression but lacking FGFR1 fusions were identified by RNA sequencing. These two cases and the FGFR1::TACC1-positive case clustered together with the ERMS group by RNAseq. This is the first report of RMS harboring recurrent FGFR1 fusions. However, it remains unclear if FGFR1 fusions define a novel subset of RMS or alternatively, whether this alteration can sporadically drive the pathogenesis of known RMS subtypes, such as ERMS. Additional larger series with integrated genomic and epigenetic datasets are needed for better subclassification, as the resulting oncogenic kinase activation underscores the potential for targeted therapy.

Sequential genomic analysis using a multisample/multiplatform approach to better define rhabdomyosarcoma progression and relapse.

The genomic spectrum of rhabdomyosarcoma (RMS) progression from primary to relapse is not fully understood. In this pilot study, we explore the sensitivity of various targeted and whole-genome NGS platforms in order to assess the best genomic approach of using liquid biopsy in future prospective clinical trials. Moreover, we investigate 35 paired primary/relapsed RMS from two contributing institutions, 18 fusion-positive (FP-RMS) and 17 fusion-negative RMS (FN-RMS) by either targeted DNA or whole exome sequencing (WES). In 10 cases, circulating tumor DNA (ctDNA) from multiple timepoints through clinical care and progression was analyzed for feasibility of liquid biopsy in monitoring treatment response/relapse. ctDNA alterations were evaluated using a targeted 36-gene custom RMS panel at high coverage for single-nucleotide variation and fusion detection, and a shallow whole-genome sequencing for copy number variation. FP-RMS have a stable genome with relapse, with common secondary alterations CDKN2A/B, MYCN, and CDK4 present at diagnosis and impacting survival. FP-RMS lacking major secondary events at baseline acquire recurrent MYCN and AKT1 alterations. FN-RMS acquire a higher number of new alterations, most commonly SMARCA2 missense mutations. ctDNA analyses detect pathognomonic variants in all RMS patients within our collection at diagnosis, regardless of type of alterations, and confirmed at relapse in 86% of FP-RMS and 100% FN-RMS. Moreover, a higher number of fusion reads is detected with increased disease burden and at relapse in patients following a fatal outcome. These results underscore patterns of tumor progression and provide rationale for using liquid biopsy to monitor treatment response.

Extremity Rhabdomyosarcoma-An Integrated Clinicopathologic and Genomic Study to Improve Risk Stratification.

PURPOSE: Extremity rhabdomyosarcoma (RMS) is associated with a very poor outcome compared with other sites, mainly because of its high incidence of alveolar histology and regional lymph node involvement. To better define prognostic markers in this clinical subset, we investigated our experience of 61 patients with extremity RMS treated at our tertiary cancer center for the past 2 decades. PATIENTS AND METHODS: The patients had a median age of 8 years at diagnosis, equal gender distribution, and two-thirds occurred in the lower extremity. Most (85%) patients had FOXO1 fusion-positive alveolar RMS (ARMS), with 70% having a PAX3::FOXO1 transcript. Remaining were seven patients with fusion-negative embryonal RMS (ERMS) and two with MYOD1-mutant spindle cell/sclerosing RMS (SRMS). In 40% of the patients, material was available for DNA-based targeted sequencing using MSK-IMPACT cancer gene panel. RESULTS: One-third of patients presented with localized disease at diagnosis while the remaining had regional nodal (18%) or distant metastases (51%). Metastatic disease, high-risk group, and age 10 years or older significantly affected the overall survival (OS; hazard ratio [HR], 2.68 [P = .004], 2.78 [P = .010] and 2.26 [P = .034], respectively). Although the presence of metastatic disease had a dismal impact on 5-year EFS and OS (19% and 29%, respectively), nodal involvement had a comparatively lower impact on 5-year EFS and 5-year OS (43% and 66%, respectively). PAX3::FOXO1 ARMS had worse prognosis and afflicted older children compared with PAX7::FOXO1 (HR = 3.45, P = .016). The most common events in the ARMS group included MED12 alterations, CDK4 amplifications, and CDKN2A deletions (8%-17%). The latter two abnormalities were mutually exclusive, enriched for acral and high-risk lesions, and correlated with poor outcome on OS (P = .02). CONCLUSION: Our data provide rationale for considering the integration of molecular abnormalities to refine risk stratification in extremity RMS.