Diffuse pleural mesotheliomas with genomic near-haploidization: a newly recognized subset with distinct clinical, histologic, and molecular features

Authors

Yang SR

Jayakumaran G

Benhamida J

Febres-Aldana CA

Fanaroff R

Chang J

Gedvilaite E

Villafania LB

Sauter JL

Offin M

Zauderer MG

Ladanyi M

Doi

10.1158/1078-0432.CCR-24-0085

PMID: 38630790 · DOI: 10.1158/1078-0432.CCR-24-0085 · Journal: Clinical Cancer Research (2024)

TL;DR

This study presents the most comprehensive clinico-genomic characterization of diffuse pleural mesotheliomas (DPMs) with genomic near-haploidization (GNH), a rare subtype comprising approximately 5% of DPMs. Using MSK-IMPACT profiling of 290 DPM patients (210 evaluable for LOH analysis via FACETS), the authors identified 10 GNH cases and showed that GNH DPMs are associated with younger age, less occupational asbestos exposure, biphasic histology (80% vs. 15%), high tumor-infiltrating lymphocytes, and recurrent somatic mutations in SETDB1 (present in nearly all and only GNH cases) and TP53. Despite younger age, GNH patients had shorter overall survival (10.9 vs. 25.4 months, p=0.004), and preliminary data suggested potential responsiveness to immune checkpoint blockade.

Cohort & data

  • Dataset: plmeso_msk_2024 — 290 patients with DPM profiled by MSK-IMPACT at Memorial Sloan Kettering Cancer Center (1/2014 – 1/2021).
  • Evaluable cohort: 210 patients (72%) with sufficient tumor purity (>=20%) for LOH analysis via FACETS.
  • Cancer type: PLMESO — diffuse pleural mesothelioma.
  • GNH cases: 10 (4.8%) in the primary cohort; 4 additional in a separate 105-patient validation cohort sequenced with IMPACT505.
  • Additional assays: Whole-exome sequencing (n=3), methylation profiling (Illumina methylationEPIC/850k), SETDB1-specific targeted panels (n=16), germline MSK-IMPACT in 87 patients (41%).
  • Demographics: Majority male (73%), median age 70 years overall; GNH patients younger (median 58 years).

Key findings

  • GNH prevalence: 10/210 evaluable DPMs (4.8%) had GNH (genome-wide LOH >80%), similar to the TCGA-ICGC rate of 3.2%. Half of GNH cases showed genome duplication resulting in >=2n copy states.
  • Younger age: GNH patients were significantly younger at diagnosis (58 vs. 71 years, p=0.008).
  • Asbestos exposure: GNH patients had less frequent self-reported occupational asbestos exposure (20% vs. 58.3%, p=0.02).
  • Biphasic histology: 80% of GNH DPMs were biphasic vs. 15% of non-GNH (p<0.0001). This enrichment persisted after controlling for other covariates.
  • High TILs and necrosis: Among treatment-naive tumors, GNH DPMs had the highest TIL scores (3+: 90% vs. 20.3%, p<0.0001) and tumor necrosis (90% vs. 24.3%, p<0.0001).
  • SETDB1 mutations: Present in 8/10 GNH cases and 0 non-GNH cases (q=0.0035); validated in the 105-patient cohort where all 4 GNH cases harbored SETDB1 mutations. SETDB1 mutations were subclonal relative to TP53 mutations based on variant allele fractions.
  • TP53 alterations: Significantly more common in GNH DPMs (60% vs. 14%, q=0.0035); remained significant after controlling for biphasic histology (q=0.04).
  • NF2 alterations: More frequent in GNH (90% vs. 26.5%, q=0.0007), but not significant after controlling for biphasic histology, suggesting the association is driven by biphasic morphology rather than GNH per se.
  • BAP1, CDKN2A/CDKN2B: These classical DPM driver genes were not enriched in GNH DPMs.
  • Retained heterozygosity: Chromosomes 5p and 7p were recurrently spared from LOH in GNH cases, with genes including TERT, RICTOR, SDHA (5p) and IKZF1, RAC1, PMS2 (7p).
  • Hypermutators: Two rare hypermutators identified — one non-GNH (31.5 muts/Mb, MLH1 truncation, MSI-high) and one GNH (18.4 muts/Mb, non-MSI mechanism).
  • Methylation profiling: No significant differences in DNA methylation between GNH and non-GNH DPMs.
  • Germline analysis: No pathogenic germline variants identified in GNH patients (n=4 tested).
  • Survival: GNH patients had shorter overall survival (10.9 vs. 25.4 months, p=0.004). The poor prognostic impact persisted after controlling for biphasic histology (GNH vs. non-GNH biphasic: 10.9 vs. 19.5 months, p=0.016).

Genes & alterations

  • SETDB1: Somatic loss-of-function mutations in 8/10 GNH cases; absent in non-GNH DPMs. SETDB1 regulates transcriptional repression; its inactivation in the GNH context may rescue expression of monoallelically expressed genes after genome-wide LOH.
  • TP53: Somatic mutations in 60% of GNH vs. 14% of non-GNH DPMs. Variant allele fraction analysis suggests TP53 mutations are early clonal events preceding genome-wide LOH.
  • NF2: Altered in 90% of GNH DPMs, but this association is confounded by biphasic histology enrichment.
  • BAP1: Classical DPM driver, not enriched in GNH subset.
  • CDKN2A/CDKN2B: Classical DPM drivers, not enriched in GNH subset.
  • MLH1: Truncating mutation in one non-GNH hypermutator case (MSI-high).

Clinical implications

  • Prognostic biomarker: GNH identifies an aggressive DPM subset with poor overall survival (10.9 months) independent of histologic subtype, potentially improving risk stratification in younger DPM patients.
  • ICB response signal: Among 3 GNH patients treated with ipilimumab/nivolumab or pembrolizumab, 2 achieved partial response (67%), vs. only 1/44 (2%) in non-GNH patients. A separate clinical trial of durvalumab plus platinum-pemetrexed in 3 GNH patients showed tumor shrinkage in all and OS >=12 months.
  • Diagnostic screening: Biphasic morphology with high TILs, necrosis, and younger age without asbestos exposure should raise suspicion for GNH, warranting allele-specific copy number analysis (e.g., FACETS on NGS data).
  • SETDB1 as a biomarker: SETDB1 mutations are highly specific for GNH DPMs and may serve as a molecular marker for this subset when detected on panels covering this gene (e.g., IMPACT505).

Limitations & open questions

  • Small sample size of GNH cases (n=10 primary, n=4 validation) limits statistical power, particularly for ICB response analysis (n=3).
  • Asbestos exposure data were self-reported and not available for all patients.
  • ICB response assessment was retrospective and clinician-assessed, not using standardized mRECIST criteria.
  • The mechanism by which GNH confers poor prognosis remains unclear; possible widespread hemizygosity of tumor suppressor genes.
  • Whether SETDB1 inactivation is a driver of GNH or a compensatory event after genome-wide LOH remains to be determined.
  • The role of retained heterozygosity at chromosomes 5p and 7p (potentially imprinted genes) needs functional validation.
  • Germline testing was limited to 4 GNH patients; larger studies needed to assess germline predisposition.
  • No enrichment of GNH in peritoneal mesotheliomas was explored in this cohort (cited at 8% in other studies).

Citations from this paper used in the wiki

  • “GNH was detected in 10 cases (4.8%). Compared to non-GNH tumors, GNH DPMs were associated with younger age and less frequent self-reported history of occupational asbestos exposure.” (Abstract)
  • “all but two GNH DPMs harbored mutations in SETDB1, while SETDB1 mutations were not identified in the non-GNH group” (Results, SETDB1 analysis)
  • “patients with GNH DPMs had shorter median overall survival compared to those with non-GNH DPMs (10.9 vs. 25.4 months, Log-rank test, p=0.004)” (Results, Survival)
  • “Within GNH DPMs, a best response of PR was achieved in 2 patients (67%)…Within non-GNH DPMs, PR was seen in only 1 patient (2%)” (Results, ICB response)
  • “alterations in NF2 and TP53 were significantly more common in GNH DPMs (90% vs. 26.5%, q=0.0007 and 60% vs. 14%, q=0.0035, respectively)” (Results, Genomic correlates)

This page was processed by crosslinker on 2026-05-04.