Molecular Profiling of Sinonasal Adenoid Cystic Carcinoma: Canonical and Noncanonical Gene Fusions and Mutation

Authors

Alena Skálová

Martina Bradová

Abbas Agaimy

Jan Laco

Cécile Badoual

Stephan Ihrler

Ivan Damjanov

Niels J. Rupp

Carlos E. Bacchi

Sarina Mueller

Sami Ventelä

Da Zhang

Eva Comperat

Petr Martínek

Radek Šíma

Tomas Vaněček

Petr Grossmann

Petr Steiner

Veronka Hájková

Inka Kovářová

Michal Michal

Ilmo Leivo

Doi

10.1097/PAS.0000000000002349

PMID: 24418857 · DOI: 10.1097/PAS.0000000000002349 · Journal: American Journal of Surgical Pathology (2025)

TL;DR

Skálová et al. molecularly profiled 88 cases of sinonasal adenoid cystic carcinoma (ACYC) drawn from multi-institutional consultation registries, combining the Illumina TruSight Oncology 500 panel (DNA + RNA) with break-apart and dual-fusion FISH for MYB, MYBL1, NFIB, and EWSR1. The cohort was dominated by the canonical MYB::NFIB (49/88, 57%) and MYBL1::NFIB (9/88, 10%) fusions, but NGS uncovered four noncanonical fusions — EWSR1::MYB, ACTB::MYB, ESRRG::DNM3, and ACTN4::MYB — each in a single case, and FISH detected MYB/NFIB/EWSR1 rearrangements in 9 additional fusion-negative tumors. DNA sequencing of 31 cases identified pathogenic mutations in 21/31 (68%) tumors, most frequently in BCOR (19%), NOTCH1 (14%), and EP300 (14%). NOTCH-mutated tumors showed solid/basaloid morphology, frequent metatypical features, and poor outcome (2 of 3 dead of disease at 20 and 64 months), and noncanonical fusions clustered with metatypical AdCC histology.

Cohort & data

  • Sample size: 88 sinonasal AdCC cases (45 men, 41 women, sex unknown in 2; mean age 58.8 y, range 20–86) drawn from the Tumor Registry of the Department of Pathology, Faculty of Medicine in Pilsen / Bioptic Laboratory Ltd. (Czech Republic) and coauthors’ international tumor registries. From 100 originally retrieved cases, 11 HPV-positive cases (reclassified as HPV-associated multiphenotypic carcinoma) and 1 submandibular primary with secondary sinonasal involvement were excluded PMID:24418857.
  • Cancer type: Sinonasal AdCC (ACYC, OncoTree).
  • Primary tumor sites: nasal cavity (49), maxillary sinus (26), sphenoid sinus (8), ethmoid sinus (4), nasopharynx (1) PMID:24418857.
  • Follow-up: Available for 60/88 (68%) patients; mean 62.7 months (range 1–276). 18 patients alive without evidence of disease (30%), 9 alive with disease (15%), 18 dead of disease (30%), 4 dead of other causes (7%), 11 lost to follow-up PMID:24418857.
  • Outcome events: Distant metastases in 14/60 (23%; lung 9, liver 2, brain 1, bone 1, axilla 1); local recurrence in 26/60 (43%) PMID:24418857.
  • Assays / panels: TruSight Oncology 500 kit (Illumina, NovaSeq 6000) for DNA mutations and RNA fusion transcripts (variant filtering via Omnomics, read depth >50, ClinVar-benign excluded); FISH with break-apart probes for MYB, MYBL1, NFIB, EWSR1 plus custom dual-fusion MYB::NFIB and EWSR1::MYB probes; PCR/Sanger HPV genotyping covering types 16/18/31/33/35/45 plus L1/E1 consensus primers CPSGB and GP5+/GP6+. Reference genome hg19 PMID:24418857.
  • Grading: Three solid-component-based grading systems applied (Perzin/Szanto 30% threshold: 49/88 grade III; Spiro 50% threshold: 33 with >50%; van Weert “any solid”: 66/88 high grade). No high-grade transformation/dedifferentiation was identified PMID:24418857.
  • Dataset: Not deposited in cBioPortal; cohort assembled from authors’ consultation registries.

Key findings

  • Canonical fusions dominate. 58/88 sinonasal AdCCs carried a canonical MYB-family fusion: 49 MYB::NFIB (57%) and 9 MYBL1::NFIB (10%). No morphologic difference was seen between MYB::NFIB and MYBL1::NFIB tumors — both showed cribriform, tubular, and solid patterns with biphasic luminal/myoepithelial differentiation PMID:24418857.
  • Four noncanonical RNA fusions identified by NGS (each n=1): in-frame EWSR1ex6::MYBex2 (sphenoid sinus, 82 y/o male, metatypical), ACTBex3::MYBex3 (nasal cavity, 62 y/o male), in-frame ACTN4ex18::MYBex2 (nasal cavity, 69 y/o male, metatypical), and ESRRGex3::DNM3ex14 (maxillary sinus, 55 y/o male, metatypical) PMID:24418857.
  • FISH rescues 9 fusion-negative cases. Among 15 cases negative for canonical fusions on RNA-seq, FISH break-apart probes detected MYB rearrangements in 7 cases, NFIB in 1, and EWSR1 in 1. Six AdCCs lacked any fusion or rearrangement, and 11 were unanalyzable PMID:24418857.
  • Mutational landscape (31/88 sequenced). Pathogenic/likely-pathogenic mutations in 21/31 tumors (68%): BCOR 4/21 (19%), NOTCH1 3/21 (14%), EP300 3/21 (14%), SMARCA4 2/21 (9%), RUNX1 2/21 (9%), KDM6A 2/21 (9%), SPEN 2/21 (9%), and 21 additional genes each in 1 case (RIT1, MGA, RB1, PHF6, PTEN, CREBBP, DDX41, CHD2, ROS1, TAF1, CCND1, NF1, PALB2, ACVR1B, ARID1A, PPM1D, LZTR1, GEN1, PDGFRA). A further 24/31 (77%) tumors harbored variants of uncertain pathogenetic significance. No mutation reached a high recurrence rate PMID:24418857.
  • NOTCH-mutated AdCCs have aggressive solid-pattern morphology. All 3 NOTCH1-mutated cases showed predominant solid/basaloid growth with confluent sheets, pronounced nuclear enlargement, comedonecrosis, high mitotic/Ki-67 indices, and perineural/intraneural/bone invasion. 2 of 3 NOTCH-mutated patients died of disease (cases 9 and 21 at 20 and 64 months); the third (case 17) was alive with disease at 73 months on proton therapy PMID:24418857.
  • BCOR-mutated AdCCs also fare poorly. Among 4 BCOR-mutated patients with follow-up data, 2 were alive with disease at 53 and 60 months but both had local recurrences (at 53 and 50 months) with lung metastasis in one PMID:24418857.
  • Noncanonical fusions cluster with metatypical AdCC. 3 of the 4 noncanonical-fusion cases (ACTN4::MYB, ESRRG::DNM3, EWSR1::MYB) displayed metatypical histology — squamous differentiation, tubular epithelial hypereosinophilia with luminal cell prominence, and vacuolated/clear cells; the ACTB::MYB case was high-grade with comedo-like necrosis but more conventional architecture PMID:24418857.
  • Metatypical features in NOTCH-mutated tumors with canonical fusions. 2 of the 3 NOTCH-mutated AdCCs (which all carried canonical MYB::NFIB fusion) also exhibited metatypical features — squamous differentiation, tubular hypereosinophilia, focal sebaceous metaplasia, and foamy vacuolated cells — extending the metatypical phenotype beyond the noncanonical-fusion subset PMID:24418857.
  • SH/ASGSH precursor lesions in 35% of cohort. 31/88 (35%) AdCCs (with both canonical and noncanonical fusions) were associated with seromucinous hamartoma (SH) and/or atypical sinonasal glands arising in SH (ASGSH); NGS of 5 ASGSH lesions revealed BRAF V600E (2 cases), RET R912W (2 cases), and FAT1 P1665L (1 case) supporting a precursor/neoplastic role PMID:24418857.
  • Pathway distribution. SWI/SNF complex (SMARCA4, ARID1A, PBRM1), PI3K-AKT (PTEN), RAS/MAPK (RIT1, NF1), and Wnt pathway genes were each represented at low recurrence; no pathway-defining hotspot emerged PMID:24418857.
  • HPV exclusion. 11/100 retrieved sinonasal “AdCC-like” tumors were HPV-positive and re-classified as HPV-associated multiphenotypic sinonasal carcinoma — a diagnostic mimic that broad PCR-based HPV genotyping helped to exclude PMID:24418857.

Genes & alterations

  • MYB — Canonical MYB::NFIB fusion in 49/88 (57%); MYB break-apart FISH detected an additional 7 fusion-negative cases. Noncanonical fusion partners ACTB, ACTN4, ESRRG, and EWSR1 fuse MYB exons 2–3 to alternative partner exons, all in-frame PMID:24418857.
  • MYBL1 — Canonical MYBL1::NFIB fusion in 9/88 (10%); represents the “MYBL1-driven” subset that is morphologically indistinguishable from MYB::NFIB tumors PMID:24418857.
  • NFIB — Recipient partner in both canonical fusions; FISH detected NFIB rearrangement in 1 additional fusion-negative case PMID:24418857.
  • EWSR1 — Noncanonical fusion partner with MYB (EWSR1ex6::MYBex2) in 1 metatypical AdCC of the sphenoid sinus; FISH break-apart detected EWSR1 rearrangement in 1 additional fusion-negative case PMID:24418857.
  • ACTB — Noncanonical MYB partner (ACTBex3::MYBex3 in-frame) in 1 nasal cavity AdCC with high-grade features and comedo-like necrosis PMID:24418857.
  • ACTN4 — Noncanonical MYB partner (ACTN4ex18::MYBex2 in-frame) in 1 metatypical nasal cavity AdCC with biphasic basaloid + vacuolated cell populations PMID:24418857.
  • ESRRG / DNM3 — Novel ESRRGex3::DNM3ex14 fusion in 1 maxillary sinus metatypical AdCC infiltrating the maxilla and arising in respiratory epithelial adenomatoid hamartoma; not previously described in AdCC PMID:24418857.
  • BCOR — Most recurrent mutation (4/21 sequenced tumors with pathogenic mutations; 19%). Multiple frameshift/nonsense variants observed including c.4017_4018insT (p.Asp1340Ter), c.1056dup (p.Thr353HisfsTer28), c.1888_1895del (p.Glu630ProfsTer12), c.3669_3673del, c.3892_3898del. Associated with poor outcome PMID:24418857.
  • NOTCH1 — Mutated in 3/21 (14%) — truncating variants p.Ser2467Ter, p.Ser2467HisfsTer8, p.Ala1908SerfsTer72, p.Gln2393Ter, p.Gln2393Ter. Co-occurring with canonical MYB::NFIB fusion in all 3; associated with solid/basaloid + metatypical morphology and poor outcome (2/3 dead of disease) PMID:24418857.
  • EP300 — Mutated in 3/21 (14%) — frameshift c.3355del (p.Asp1119MetfsTer38), missense c.1811G>A (p.Arg604Gln), nonsense c.1927G>T (p.Glu643Ter) PMID:24418857.
  • SMARCA4 — 2/21 (9%) — missense p.Gly1194Trp and p.Glu882Lys PMID:24418857.
  • RUNX1 — 2/21 (9%) — frameshift duplications/deletions PMID:24418857.
  • KDM6A — 2/21 (9%) — frameshift c.2172_2173del + splice-site c.618-619+2del in one case; frameshift c.64del in a second case PMID:24418857.
  • SPEN — 2/21 (9%) — frameshift duplications c.2918dup (p.Ser974LysfsTer48) and c.2906del (p.Pro969LeufsTer5) PMID:24418857.
  • PTEN — Single case: frameshift p.Arg11LysfsTer32 co-occurring with EWSR1::MYB fusion in metatypical AdCC of the sphenoid sinus PMID:24418857.
  • PDGFRA — Single nasal-cavity case (case 2) with missense p.Thr674Ile (AF 8%); this patient received targeted therapy with imatinib (the only patient in the cohort to receive targeted therapy) and was NED at 19 months PMID:24418857.
  • Other single-case mutations: RIT1 p.Met107Val, MGA frameshift, RB1 frameshift, PHF6 nonsense p.Arg347Ter (suspected germline), CREBBP frameshift p.Pro830LeufsTer19, DDX41 nonsense p.Gln41Ter (suspected germline), CHD2 frameshift, ROS1 frameshift (suspected germline), TAF1 splice, CCND1 nonsense p.Cys285Ter (suspected germline), NF1 frameshift, PALB2 frameshift (suspected germline), ACVR1B frameshift, ARID1A in-frame deletion p.Tyr560_Gln561delinsTer, PPM1D frameshift, LZTR1 missense p.Gly404Arg, GEN1 frameshift (suspected germline) PMID:24418857.
  • ASGSH precursor mutations: BRAF V600E (2 cases), RET R912W (2 cases), FAT1 P1665L (1 case) in 5 atypical sinonasal glands arising in seromucinous hamartoma PMID:24418857.
  • Pathway-level genes referenced: NOTCH2 and NOTCH3 discussed as Notch-pathway co-actors in AdCC (literature) and one metatypical case harbored a NOTCH3 mutation PMID:24418857. PBRM1 / ARID1B flagged as SWI/SNF chromatin-remodeling complex partners PMID:24418857.

Clinical implications

  • No effective systemic therapy for sinonasal AdCC. Surgery (radical resection in 30/47 with available data, 64%) with adjuvant radiotherapy or proton therapy remained the mainstay; chemotherapy was used in selected cases. Only 1 patient (case 2, PDGFRA p.Thr674Ile) received targeted therapy with imatinib and was NED at 19 months PMID:24418857.
  • NOTCH pathway as a candidate target. The authors highlight Notch signaling as “a potentially targetable pathway” and explicitly raise brontictuzumab (anti-NOTCH1 monoclonal antibody) as a candidate intervention for the NOTCH-mutated subgroup, citing the demonstration by Ferrarotto and colleagues that activating NOTCH1 mutations identify an AdCC subset with bone/liver metastatic propensity and potential response to Notch1 inhibitors PMID:24418857.
  • Prognostic stratification by fusion + mutation class. NOTCH1 and BCOR mutations were associated with high-grade solid/metatypical histology, recurrence, and reduced survival in this cohort, consistent with prior reports that NOTCH alterations associate with poor survival in AdCC PMID:24418857.
  • Diagnostic implication for sinonasal AdCC workup. RNA-seq panels (TruSight Oncology 500) miss ~10% of AdCCs detectable only by FISH; the authors recommend pairing NGS with break-apart FISH for MYB, MYBL1, NFIB, and EWSR1 when the diagnosis is uncertain — especially in metatypical AdCC where canonical fusions may be absent and noncanonical partners (EWSR1, ACTB, ACTN4, ESRRG) emerge PMID:24418857.
  • HPV testing as mandatory exclusion step. 11% of initially retrieved cases were reclassified after HPV genotyping; HPV-associated multiphenotypic sinonasal carcinoma is a key mimic to exclude before calling sinonasal AdCC PMID:24418857.

Limitations & open questions

  • Partial mutational profiling. Only 31/88 cases (35%) underwent DNA mutation analysis; 57 cases had insufficient material or other constraints. Mutation frequencies are therefore reported on a non-random subset and may not generalize to the full cohort PMID:24418857.
  • Low recurrence rates of individual mutations. Authors explicitly note that “none of the mutations observed in this study had a high recurrence rate in our sinonasal AdCCs, and consequently they do not appear to play essential roles in the pathogenetic mechanism of AdCC” — leaving open whether the long tail of singleton mutations represents true co-drivers, passenger noise, or context-specific contributors PMID:24418857.
  • Suspected germline variants not formally confirmed. Several mutations (e.g., GEN1, BCOR p.Ala1224ArgfsTer19, DDX41 p.Gln41Ter, PHF6 p.Arg347Ter, ROS1, CCND1 p.Cys285Ter, PALB2) were flagged as “suspected germline” based on allele fraction but were not validated against matched normal tissue PMID:24418857.
  • Retrospective multi-institutional cohort. Cases were collected from consultation files spanning multiple countries (Czech Republic, Germany, France, USA, Brazil, Austria, Switzerland, Finland) with non-uniform treatment protocols, follow-up cadence, and tissue quality. Follow-up was available for only 60/88 (68%) patients.
  • No matched normal sequencing. TruSight Oncology 500 tumor-only sequencing relied on ClinVar-benign exclusion and visual review for somatic-variant calling; germline contamination and rare variants of unknown significance remain a concern.
  • Single noncanonical-fusion cases. Each noncanonical fusion (EWSR1::MYB, ACTB::MYB, ACTN4::MYB, ESRRG::DNM3) was observed in only 1 case. Generalizing the metatypical-histology association from n=3 of n=4 noncanonical-fusion tumors will require larger cohorts.
  • Unanalyzable fraction. 11 cases were unanalyzable by either NGS or FISH due to tissue quality; 6 cases lacked any detectable fusion or rearrangement and remain mechanistically unexplained.
  • Targetable-pathway claims are aspirational. The mention of brontictuzumab and Notch1 inhibition is grounded in prior literature (Ferrarotto et al.) rather than in any prospective intervention reported here; the imatinib-treated PDGFRA-mutant case is a single observation at 19 months follow-up.
  • No expression or copy-number profiling. The study did not include whole-transcriptome RNA-seq, SNP-array, or methylation profiling that could clarify subtype boundaries; classification rests on the targeted DNA+RNA TSO500 panel plus FISH.

Citations from this paper used in the wiki

  • “Thus, a cohort of 88 cases of sinonasal AdCC was included in this study for further characterization.” (Materials and Methods, Case Selection)
  • “Sinonasal AdCC was predominantly characterized by canonical MYB:NFIB (49 cases)… and MYBL1:NFIB (9 cases)… fusions. In addition, rearrangements in MYB (8 cases), NFIB (1 case), and EWSR1 (1 case) genes were detected in 9 cases using FISH.” (Results, Molecular Findings)
  • “NGS analysis revealed novel noncanonical fusion transcripts, including ACTB:MYB; ACTN4:MYB; ESRRG:DNM3, and EWSR1:MYB, each in 1 case.” (Results, Molecular Findings)
  • “Mutations in genes with established roles in oncogenesis were identified in 21/31 tumors (68%), including BCOR (4/21; 19%), NOTCH1 (3/21; 14%), EP300 (3/21; 14%), SMARCA4 (2/21; 9%), RUNX1 (2/21; 9%), KDM6A (2/21; 9%), SPEN (2/21; 9%)…” (Results, Molecular Findings)
  • “The clinical outcome of patients with NOTCH and BCOR mutated AdCCs was notably poor.” (Results)
  • “Patients with such aberrations might potentially benefit from anti-NOTCH drugs, such as Brontictuzumab.” (Discussion)
  • “Our cohort of primary sinonasal AdCCs included 31/88 (35%) cases (both with canonical and noncanonical fusions) associated with growth patterns of SH and/or ASGSH… The possible precursor/neoplastic nature of ASGSH was supported by various mutations revealed by NGS in 5 cases, including BRAF Val600Glu (2 cases), RET Arg912Trp (2 cases), and FAT1 Pro1665Leu (1 case).” (Discussion)
  • “None of the mutations observed in this study had a high recurrence rate in our sinonasal AdCCs, and consequently they do not appear to play essential roles in the pathogenetic mechanism of AdCC.” (Discussion)
  • “One patient received targeted therapy with Imatinib.” (Results, Demographic and Clinicopathologic Findings)
  • “Mutation analysis and fusion-transcript detection were performed using the TruSight Oncology 500 Kit (Illumina, San Diego, CA).” (Methods, Molecular Studies)

This page was processed by crosslinker on 2026-05-09.