Whole-genome characterization of myoepithelial carcinomas of the soft tissue

Authors

Cyrta J

Rosiene J

Bareja R

Kudman S

Al Zoughbi W

Motanagh S

Wilkes DC

Eng K

Zhang T

Sticca E

Mathew S

Rubin MA

Sboner A

Elemento O

Rubin BP

Imielinski M

Mosquera JM

Doi

10.1101/mcs.a006227

PMID: 36577525 · DOI: 10.1101/mcs.a006227 · Journal: Cold Spring Harbor Molecular Case Studies (2022)

TL;DR

This study presents the first whole-genome sequencing (WGS) characterization of two myoepithelial carcinomas (MECs) of soft tissue. Both tumors displayed very low tumor mutation burden (mean 1.35 mutations/Mb) with no targetable alterations in known cancer genes. WGS revealed complex structural variants including an EWSR1::KLF15 fusion in one case and a novel ASCC2::GGNBP2 fusion with biallelic SMARCB1/NF1 loss via chromoplexy in the other, demonstrating that WGS provides critical biological insights invisible to WES for rare sarcomas.

Cohort & data

  • Two patients with soft tissue MYEC, male, diagnosed at ages 27 and 37
  • 12 total tumor samples (primary + metastatic/recurrent) profiled by WES (mean coverage 98.5x)
  • Primary tumors profiled by WGS (Patient 1: 29x tumor/25x normal; Patient 2: 22x tumor/15x normal)
  • RNA-seq performed on one sample per patient for fusion validation
  • Dataset deposited to cBioPortal: stmyec_wcm_2022
  • Reference genome: GRCh37/hg19

Key findings

  • Both MECs displayed very low tumor mutation burden (mean 1.35 mutations/Mb, range 0.43-2.98), consistent with fusion-driven malignancies
  • Patient 1 (MEC1): homozygous SMARCB1 deletion and NF1 loss, plus a novel ASCC2::GGNBP2 in-frame fusion, all arising from a single >30-way biallelic chromoplexy event involving Chromosomes 17, 19, and 22
  • Patient 2 (MEC2): EWSR1::KLF15 rearrangement (previously reported fusion in MECs), missense CREB1 mutation, and high-level (>10 copy) ELK4 amplification via double minute
  • Patient 2 also harbored a p.E830G POLE missense mutation (predicted probably damaging) in 3 of 6 samples
  • Mutational signature analysis showed COSMIC Signature 3 (HR deficiency-associated) in both tumors (36.8% of SNVs in MEC1, 17.4% in MEC2), though no somatic HR pathway alterations were identified
  • SMARCA2 (BRM) showed heterogeneous “alternating” expression by IHC in both cases, similar to malignant rhabdoid tumors

Genes & alterations

  • SMARCB1 — homozygous deletion with complete loss of protein expression (IHC) in MEC1; part of chromoplexy event on Chr 22
  • NF1 — biallelic loss via chromoplexy in MEC1
  • EWSR1 — EWSR1(exon 8)::KLF15(exon 2) fusion in MEC2; confirmed by RNA-seq and FISH
  • KLF15 — 3-prime fusion partner of EWSR1 in MEC2
  • ASCC2 — 5-prime fusion partner in novel ASCC2(exon 2)::GGNBP2(exon 3) fusion in MEC1; validated by RNA-seq, RT-PCR, and Sanger sequencing
  • GGNBP2 — 3-prime fusion partner encoding zinc finger DNA-binding protein; knockdown causes cell cycle arrest
  • ELK4 — high-level amplification (>10 copies) via double minute in MEC2; ETS family transcription factor previously associated with oncogenic cis-splicing in prostate cancer
  • CREB1 — missense mutation in MEC2; known EWSR1 fusion partner in other soft tissue sarcomas
  • POLE — p.E830G missense mutation in Patient 2
  • SMARCA2 — heterogeneous nuclear expression in both MECs by IHC

Clinical implications

  • No actionable therapeutic targets were identified in either tumor by WES or WGS, including in metastatic/recurrent samples, highlighting the challenge of precision medicine for rare soft tissue sarcomas
  • The low TMB and absence of targetable point mutations support that gene fusions and structural variants are the primary oncogenic drivers in MEC
  • The finding of biallelic chromoplexy causing simultaneous loss of multiple tumor suppressors has implications for understanding tumorigenesis mechanism in MEC and related sarcomas

Limitations & open questions

  • Only two cases were profiled by WGS; larger cohorts are needed to define the full spectrum of structural variants in MEC
  • No functional studies were performed to assess the oncogenic role of the novel ASCC2::GGNBP2 fusion
  • The significance of COSMIC Signature 3 without detectable HR pathway mutations remains unexplained
  • The relationship between MECs with SMARCB1 loss and rhabdoid morphology versus malignant rhabdoid tumors (MRTs) requires further investigation
  • WGS coverage was moderate (22-29x tumor), potentially limiting detection of subclonal events
  • The study does not address whether SMARCB1 loss is a driver or bystander event in MECs harboring 22q rearrangements

Citations from this paper used in the wiki

  • “Soft tissue sarcomas, including Ewing sarcoma, have been previously associated with high burdens of complex structural variants (SVs)” (p. 3)
  • “Both patients displayed low tumor mutational burdens (mean 1.35 mutations/Mb, range 0.43-2.98), and no targetable alterations in cancer genes were detected” (Abstract)
  • “MEC from Patient 1 (MEC1) harbored a homozygous loss of SMARCB1 and NF1, whereas the MEC from Patient 2 (MEC2) showed an in-frame EWSR1(Chr22)::KLF15(Chr3) fusion” (p. 4)
  • “We conclude that a single punctuated >30-way rearrangement targeted both parental alleles in MEC1, explaining the biallelic copy loss of NF1 and SMARCB1 and the ASCC2::GGNBP2 fusion” (p. 9)
  • “WES data were deposited to cBioPortal (study reference: stmyec_wcm_2022)” (p. 13)

This page was processed by crosslinker on 2026-05-05.