Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2025 Aug 25.
Published in final edited form as: Clin Cancer Res. 2025 Sep 15;31(18):3922–3931. doi: 10.1158/1078-0432.CCR-25-1071

Molecular characterization of NUT carcinoma: a report from the NUT carcinoma registry

Justin J Kim 1, Sara A Walton 1, Navin R Mahadevan 2,3, Jessica Haradon 1, Francesco Paoloni 1, Paul K Paik 4, Jamie E Chaft 4, Robert Hsu 5, Sarina A Piha-Paul 6, Pasi A Jänne 1,7, David A Barbie 1,7, Lynette M Sholl 2, Steven G DuBois 8, Glenn J Hanna 7,9,10, Geoffrey I Shapiro 7,10, Christopher A French 2,^,+, Jia Luo 1,7,^,+
PMCID: PMC12373430  NIHMSID: NIHMS2100430  PMID: 40704901

Abstract

Purpose:

NUT carcinoma (NC) is an underdiagnosed, poorly differentiated squamous cell cancer with a median survival of 6.7 months. Defined by NUTM1 fusions, NC enhances oncogene transcription, including MYC. We investigated the ability of standard next-generation sequencing (NGS) to identify NUTM1 fusions and describe additional molecular features of NC.

Experimental Design:

This study included 116 NC patients whose tumors underwent broad-panel NGS (>80 genes) of DNA, ctDNA, and/or RNA fusion sequencing between 2013–2024. NGS reports and medical records were manually reviewed.

Results:

Of 116 patients (median age 38, 40.5% female), 84.5% had DNA, 12.1% had ctDNA, and 51.7% had RNA fusion testing. In a subset of 100 patients with DNA/ctDNA testing, 92.9% (n=79/85) had <10 pack-years/never-smoking history, and 58.8% (n=47/80) had a BRD4::NUTM1 fusion. Median TMB was 1.0 mut/Mb (range 0.0–16.0; n=71 known), and 19.7% (n=13/66) had PD-L1 expression ≥1%.

DNA, ctDNA, RNA fusion, NUT IHC, and NUTM1 FISH detected NC fusions in 21.6%, 21.4%, 83.9%, 100.0%, and 91.9% of tests, respectively. Co-occurring pathogenic mutations included oncogenes PIK3CA, RET, FGFR3, and tumor suppressors ATM and BRCA1 (n=1 each). Secondary genes altered in >5% of NCs included LRP1B (10.4%), MLL2/KMT2D (8.0%), and FAT1 (5.5%); common pathways with mutated genes were epigenetic (57.0%), cell cycle (26.0%), and DNA repair (24.0%).

Conclusions:

Standard DNA NGS detects less than a quarter of NUT carcinomas; RNA-based fusion testing, or NUT IHC/NUTM1 FISH, should be routine for suspected NC. NCs are enriched in co-occurring epigenetic, cell cycle, and DNA repair alterations, warranting further evaluation.

TRANSLATIONAL RELEVANCE

NUT carcinoma (NC), a poorly differentiated squamous cell cancer of mostly thoracic/head and neck origin defined by a fusion oncogene, remains a diagnostic and therapeutic challenge due to its under-recognition and aggressive natural history. This study examined standard of care (SOC) next-generation sequencing (NGS) assays for detecting NC-defining NUTM1 fusions and characterizing its molecular profile. We establish the limitations of SOC DNA-based NGS testing, emphasizing the need to prioritize NUT IHC, NUTM1 FISH, or RNA fusion-based NGS assays in suspected cases. Molecular characterization revealed common fusion sites in NUTM1 (exon 3), BRD4 (exon 4), BRD3 (exon 10), and NSD3 (exon 7), with occasional co-occurring secondary pathogenic mutations (e.g., in PIK3CA, FGFR3, and ATM). Most co-occurring mutations of interest were seen in epigenetic regulation, cell cycle control, and DNA repair pathways. These findings warrant immediate change to the diagnostic workflow and provide insight into developing effective therapies for this aggressive cancer.

INTRODUCTION

NUT carcinoma (NC) is an underdiagnosed, poorly differentiated squamous lung or head/neck cancer with a median overall survival of 6.7 months.1–4 NC is defined by specific gene fusions between nuclear protein in testis (NUT, encoded by NUTM1) and an epigenetic reader. The most common fusion partners are bromodomain-containing protein 4 (BRD4::NUTM1), bromodomain-containing protein 3 (BRD3::NUTM1), and nuclear receptor binding SET domain protein 3 (NSD3::NUTM1).5,6 The resulting fusion oncoprotein upregulates the transcription of oncogenes, including MYC.7–10

Diagnosis of NC relies on positive NUT immunohistochemistry (IHC), positive NUTM1 fluorescence in situ hybridization (FISH), or identification of NC-defining NUTM1 fusion oncogenes by next-generation sequencing (NGS).11–13 Due to the wide distribution of fusion sites across introns and exons, standard primers/baits used in DNA-based NGS may miss NUTM1 fusions. Based on the limitations of reliably detecting fusion drivers using DNA-based NGS alone, we hypothesize that NC fusions are often missed in standard-of-care (SOC) clinical testing.

Additionally, DNA-based panel NGS is standard for baseline diagnostics in lung and head/neck cancers. A few case reports14–17 and case series18–20 of NC have examined whether NGS testing yields clinically actionable mutations. Kloker et al. reported 15 patients with NC with DNA-based NGS (median of 708 genes sequenced, range 8–766) and found no actionable mutations.18 Riess et al. examined 31 patients with NC with DNA-based panel NGS (Foundation One, 186–315 genes sequenced) and found no clinically relevant co-occurring alterations, including point mutations, copy number alterations, and other genomic rearrangements.19 Kroening and Luo et al. analyzed 54 patients with NC with DNA- and RNA-based NGS (Caris MI Profile Comprehensive Testing, 700+ sequenced genes) and found co-occurring mutations primarily in epigenetic or cell cycle regulation pathways.20 Overall, these studies are limited by small sample sizes or confined to a single assay.

We sought to characterize the genomic landscape of NC for two reasons. First, the performance of SOC DNA- and RNA fusion-based molecular testing in detecting NC-defining NUTM1 fusions is poorly defined. Second, our understanding of the co-occurring genomic alterations in NC is limited. Assessment of these issues in the largest cohort of patients with NC to date could inform future therapeutic approaches in this difficult-to-treat cancer. Given these knowledge gaps, we leveraged an existing international registry to evaluate SOC NGS testing for NUTM1 fusions and further characterize the molecular features of NC.

METHODS

Patient Cohort and Characteristics

This retrospective study included patients diagnosed with NC enrolled in the International NC Registry (NC-Registry.org) between 2010–2024. Patients were eligible if they met the following criteria: (1) pathologically confirmed NC via centralized pathology slide review and/or pathology/NGS reports and (2) underwent SOC broad-panel NGS analyzing DNA (>80 genes), circulating tumor DNA (ctDNA) (>80 genes), and/or RNA fusion sequencing. Each patient or their legal guardian provided written informed consent for participation in the registry. All study protocols (Dana-Farber/Harvard Cancer Center protocols 10–228 and 10–088) were approved by the institutional review board, including approval for the inclusion and reporting of patients of all ages. This study was conducted in accordance with the principles of the Declaration of Helsinki. Cases confirmed via review of primary pathology and/or NGS reports were discussed and jointly adjudicated with CAF if the diagnosis of NC was unclear based on review. In some instances, slides were requested for central review to confirm the diagnosis. Cases were excluded if an NC diagnosis could not be made based on review of primary pathology reports.

Additionally, patients were included in a subcohort (DNA cohort) if they underwent DNA NGS testing or ctDNA NGS testing that detected an NC-defining NUTM1 or fusion partner (BRD4, NSD3, BRD3, ZNF592, BRD2) alteration. Another subcohort (exon fusion cohort) included patients with successful NGS testing that detected an NC-defining NUTM1 or fusion partner alteration, with detailed information about where the exon fusion sites were for at least one gene. Primary NGS reports for the DNA cohort and exon fusion cohort were manually reviewed.

Patient Data Abstraction

A comprehensive manual review was conducted of each patient’s primary medical records, NGS reports, and treating physician-completed questionnaires. Baseline demographics and characteristics of NC, including age, sex, race, ethnicity, smoking history, disease stage, primary disease site, diagnostic test results, known NUTM1 fusion partner (if applicable), PD-L1 expression, tumor mutation burden, and survival outcomes were manually abstracted by trained research personnel. Thoracic primary diseases were defined as tumors originating from the thoracic cavity and were most commonly classified as non-small cell lung cancer (NSCLC). Overall survival outcomes were also evaluated as a predefined study objective to compare patients who were diagnosed by IHC and those with NGS first.

Tumor Tissue Characterization

A diagnosis of NC was made by detecting an NC-defining NUTM1 rearrangement using one or more of the following methods: (1) NUT IHC showing NUT staining in more than 50% tumor cells,12,21 and/or (2) identification of NC-defining NUTM1 rearrangements via a validated FISH assay or (3) clinical NGS test (DNA or RNA fusion-based). When possible, the specific NUTM1 fusion type was determined through FISH (BRD4::NUTM1 defined by t(15;19)(p13.1; q14)1,22); either research or a CLIA-certified FISH assay for BRD4::NUTM1 or non-BRD4::NUTM1 fusion at Brigham and Women’s Hospital; or RNA- or DNA-based NGS at a CLIA-certified laboratory.

Genomic Data and Analysis

For each patient, DNA-based and RNA fusion-based NGS testing was conducted based on treating physicians’ recommendations as part of standard-of-care diagnostics. DNA and ctDNA NGS tests were either performed at an academic laboratory or a commercial laboratory. An academic laboratory was defined as a research-focused laboratory affiliated with a university, hospital, or non-profit research institute; a commercial laboratory was defined as a for-profit laboratory that provides diagnostic services, clinical testing, or genetic sequencing as a business.

NGS tumor-only assays reported missense mutations, nonsense mutations, in-frame insertions, in-frame deletions, splice site mutations, frameshift deletions, and copy number alterations. RNA fusion assays reported on select gene fusions. DNA and ctDNA NGS assays testing fewer than 80 genes were excluded to ensure a large panel of cancer-related genes was sequenced for each patient; the cutoff of 80 genes was chosen as it naturally separated the smaller vs larger DNA-based NGS panels used clinically. Only 6 NGS tests were excluded based on this criterion, and none of the associated panels tested for NUTM1 or its fusion partners or found any alterations of clinical interest. Among the patients who received DNA NGS, 26 patients had their tumors tested using matched germline sequencing. For the DNA cohort, we only included ctDNA reports that confirmed the presence of the NUTM1 fusion and excluded variants potentially associated with clonal hematopoiesis (97% [n=97/100] tissue, 3% [n=3/100] ctDNA). To gather genomic data, a manual review of primary NGS reports was conducted, noting reported variants, tumor mutation burden, and genomic exon fusion locations of the NUTM1 gene fusion (if reported). PD-L1 expression was obtained through review of primary clinical pathology reports. PD-L1 testing was performed using the following antibody clones: E1L3N, 22C3, SP263, SP142, and 28–8.

Mutation classification was performed using the OncoKB (RRID: SCR_014782)23,24 and the ClinVar (RRID: SCR_006169)25 databases. The OncoKB tier system categorizes mutations as follows: tier 1 (FDA-recognized biomarkers in specific cancer indications), tier 2 (standard care biomarkers in specific cancer indications), tier 3 (biomarkers with compelling clinical evidence), and tier 4 (biomarkers with compelling biological evidence). The ClinVar classification system categorizes mutations as benign, likely benign, pathogenic, likely pathogenic, or of uncertain significance. Mutations were considered of clinical interest under the following conditions: (1) mutations classified as tiers 1 or 2 in OncoKB were automatically included, and (2) mutations classified as tiers 3 or 4 or not annotated by OncoKB were further evaluated based on additional criteria. For mutations not labeled as OncoKB tier 1 or tier 2, clinical interest was determined if the mutation was labeled as activating or damaging in OncoKB. Additionally, mutations classified as pathogenic, likely pathogenic, or variants of uncertain significance (VUS) in ClinVar were included. Given the limited genomic data available, we included VUSs to avoid excluding potentially meaningful alterations as NC continues to undergo molecular characterization. Conversely, mutations labeled as benign or likely benign in ClinVar were excluded. Furthermore, the final list of genes of clinical interest was manually reviewed and vetted by a molecular pathologist (NRM). Biological functional enrichment analysis was performed using the DAVID Bioinformatics software (RRID: SCR_001881).26–28 Gene Ontology term enrichment analysis was performed using the SRplot online platform (RRID: SCR_025904).29

Statistical Analysis

Survival outcomes were estimated using the Kaplan-Meier method, with overall survival (OS) defined as the time from NC diagnosis until the date of death. Patients who did not experience death at the data cut-off were censored at their last confirmed follow-up date prior to the data lock on September 30, 2024. Differences in survival curves and estimates of univariable hazard ratios were determined using a Cox regression model. A multivariable Cox regression model was also used to adjust hazard ratios based on primary tumor site (thoracic vs. non-thoracic), given its established prognostic significance in NC. Comparisons of categorical variables between groups were evaluated using Fisher’s exact test. All p-values are two-sided, and confidence intervals are at the 95% level, with the significance level defined at p<0.05. All statistical analyses and data visualization were performed using GraphPad Prism 10.4.1 (RRID: SCR_002798).

Data availability statement

The data generated in this study are available within the article and its supplementary data files.

RESULTS

Clinical Characteristics

Between January 1, 2010, and September 30, 2024, a total of 270 patients with NC consented to participate in the NC registry (NC-Registry.org) (Supplementary Fig. S1). Of these, we identified 116 patients with NC tumors or peripheral blood sent for at least one SOC NGS-based test and for whom an original NGS report was available for extraction; this included 84.5% (n=98/116) DNA, 12.1% (n=14/116) ctDNA, and 51.7% (n=60/116) RNA fusion tests (Figure 1A). Of the 116 patients, 29 were included in a prior report that did not discuss NGS reports.2 (ct)DNA-based tests were included if the assay tested >80 cancer-associated genes. The NGS tests encompassed 30 unique assays, including commercial DNA assays, academic DNA assays (e.g., BWH/DFCI Oncopanel, an in-house CLIA validated DNA NGS assay testing 447 cancer-associated genes),30 ctDNA assays, and RNA fusion assays (Supplementary Table S1). Of the 23 patients diagnosed solely with an IHC test, 95.7% (n=22/23) of the histopathological diagnoses were carcinomas and the remaining patient had a histopathology report consistent with a limited sample (Supplementary Fig. S2).

Figure 1: NUTM1 fusion detection performance among assays in the overall cohort.

Figure 1:

Open in a new tab

(A) Distribution of NGS assays performed on the entire cohort (B) Survival since the diagnostic test that led to NC stratified by NUT IHC versus an NGS assay (C) Proportion of unique NGS assays that tested for NC-defining NUTM1 gene fusions grouped by type of assay and academic or commercial assay (D) Detection of NC NUTM1 gene fusions in the overall cohort grouped by type of assay and academic or commercial assay (E) Proportion of assays that tested for (e.g., included NUTM1 and/or the most common fusion partners BRD4, BRD3, and NSD3) and detected NC grouped by type of assay and NUT fusion partner. The size of the circle represents the number of tests. The size of the arc within each circle represents the percentage detected. Categories with n<5 were not illustrated with a circle (F) Proportion of unique NGS assays that tested for NC NUTM1 gene fusions by year of testing, pre-2000 vs post-2020 (G) Proportion of NC NUTM1 gene fusions detected by DNA or RNA fusion NGS by year of testing, pre-2000 vs post-2020

In this cohort of 116 patients, the median age was 38 years (range 7–76), and 40.5% (n=47/116) were female (Table 1). Patients’ self-reported race was 72.5% (n=66/91) White, 13.2% (n=12/91) Asian, 7.7% (n=7/91) Black/African American, and 6.6% (n=6/91) Other. Ethnicity was reported in 83 patients, of whom 3.6% (n=3/83) self-identified as Hispanic. Greater than half of the patients had a thoracic primary (62.9%, n=73/116), a BRD4::NUTM1 fusion (59.1%, n=55/93), and metastatic disease at diagnosis (57.8%, n=67/116). A majority (78.8%, n=82/104) reported no cigarette smoking history (median pack-years: 0, range: 0–20). PD-L1 expression ≥1% (by tumor proportion score [TPS] or combined positive score [CPS]) was observed in 21.9% (n=16/73) of cases (range: 0–70%). The median tumor mutation burden was 1.0 mt/Mb (range 0.0–16.0, n=73 known), and no cases of microsatellite instability were detected by NGS. When comparing the overall survival in those diagnosed via NGS testing (DNA, ctDNA, or RNA fusion) versus NUT IHC, 1-year survival was 25% for NGS-diagnosed patients and 43% for those diagnosed via IHC. The unadjusted hazard ratio was 1.42 (95% CI 0.82–2.46, p=0.16) (Figure 1B and Supplementary Fig. S3). After adjusting for primary tumor site (thoracic vs. non-thoracic) using a multivariable Cox regression model, the association remained non-significant (adjusted HR=1.27; 95% CI 0.53–2.76, p=0.57).

Table 1:

Characteristics of patients with NUT carcinoma

Characteristic Overall cohort (n=116) DNA cohort (n=100) Exon fusion cohort (n=46)
Age, median (range) — years 38 (7–76) 38 (9–76) 39 (7–73)
Sex — no. (%)
 Female 47 (40.5%) 40 (40.0%) 18 (39.1%)
 Male 69 (59.5%) 60 (60.0%) 28 (60.9%)
Race, self-reported — no./total no. (%)
 White 66/91 (72.5%) 59/81 (72.8%) 24/36 (66.7%)
 Black/African American 7/91 (7.7%) 6/81 (7.4%) 5/36 (13.9%)
 Asian 12/91 (13.2%) 10/81 (12.3%) 5/36 (13.9%)
 Other 6/91 (6.6%) 6/81 (7.4%) 2/36 (5.6%)
 Unknown 25 19 10
Ethnicity, self-reported — no./total no. (%)
 Hispanic 3/83 (3.6%) 3/75 (4.0%) 2/33 (6.1%)
 Not Hispanic 80/83 (96.4%) 72/75 (96.0%) 31/33 (93.9%)
 Unknown 33 25 13
Cigarette Smoking History — no./total no. (%)
 Never 82/104 (78.8%) 72/90 (80.0%) 26/38 (68.4%)
 Former or Current 22/104 (21.2%) 18/90 (20.0%) 12/38 (31.6%)
  Pack-years, median (range) 0 (0–20) 0 (0–20) 0 (0–15)
 Unknown 12 10 8
Disease stage
 Metastatic 67/116 (57.8%) 56/100 (56.0%) 26/46 (56.5%)
 Non-metastatic 49/116 (42.2%) 44/100 (44.0%) 20/46 (43.5%)
Site of Primary Disease — no./total no. (%)
 Thoracic 73/116 (62.9%) 63/100 (63.0%) 32/46 (69.6%)
 Non-thoracic 43/116 (37.1%) 37/100 (37.0%) 14/46 (30.4%)
  Head and Neck 38/43 (88.4%) 33/37 (89.2%) 12/14 (85.7%)
  Other 5/43 (11.6%) 4/37 (10.8%) 2/14 (14.3%)
NUTM1 - fusion partner — no./total no. (%)
 BRD4 55/93 (59.1%) 47/80 (58.8%) 22/46 (47.8%)
 Non-BRD4 38/93 (40.9%) 33/80 (41.3%) 24/46 (52.2%)
  BRD3 18/36 (50.0%) 14/31 (45.2%) 13/24 (54.2%)
  NSD3 16/36 (44.4%) 16/31 (51.6%) 9/24 (37.5%)
  BRD2 1/36 (2.8%) 1/31 (3.2%) 1/24 (4.2%)
  ZNF592 1/36 (2.8%) N/A 1/24 (4.2%)
  Non-BRD4, Fusion Unknown 2 2 N/A
 Unknown 23 20 N/A
PD-L1 Score
 Tumor Proportion Score (%), median (range) 0 (0–70) 0 (0–70) 0 (0–70)
  Tumor Proportion Score (%) < 1 — no./total no. (%) 50/60 (83.3%) 47/55 (85.5%) 22/28 (78.6%)
  Tumor Proportion Score (%) ≥ 1 — no./total no. (%) 10/60 (16.7%) 8/55 (14.5%) 6/28 (21.4%)
 Combined Proportion Score (%), median (range) 0 (0–30) 0 (0–30) 5 (0–20)
  Combined Proportion Score (%) < 1 — no./total no. (%) 7/13 (53.8%) 6/11 (54.5%) 2/5 (40.0%)
  Combined Proportion Score (%) ≥ 1 — no./total no. (%) 6/13 (46.2%) 5/11 (45.5%) 3/5 (60.0%)
 Unknown 43 34 13
Tumor Mutation Burden (TMB)
 TMB (mut/Mb), median (range) 1.0 (0–16) 1.0 (0–16) 1.6 (0–16)
 Unknown 43 29 19
Open in a new tab

Performance of Clinical NGS Assays in Detecting NC

A total of 20 unique DNA, 4 ctDNA, and 15 RNA fusion assays were performed on the NC tumors/plasma in the overall cohort (Supplementary Table S1). 40.0% (n=8/20) of DNA assays, 25.0% (n=1/4) of ctDNA assays, and 80.0% (n=12/15) of RNA fusion assays tested for NUTM1 or all three of the most common fusion partners that define NC (BRD4, NSD3, and BRD3) (Figure 1C). There were no significant differences in coverage of NC-defining genes when comparing academic and commercial assays (p>0.99 for both DNA and RNA fusion).

Overall, 62.9% (n=73/116) of patients had NC-defining NUTM1 fusions detected by at least one of these NGS tests. The detection of fusions diagnostic of NC (NUTM1 or NC fusion partners [BRD4, NSD3, BRD3, ZNF592, BRD2]) varied across testing modalities; the rate of detection was 21.6% (n=22/102) for DNA tests, 21.4% (n=3/14) for ctDNA tests, and 83.9% (n=52/62) for RNA fusion tests, with no significant differences observed between academic and commercial tests (DNA, p>0.99 and RNA, p=0.49) (Figure 1D).

Among assays specifically testing for the NUTM1 fusion in the overall cohort (NUTM1 or all three most common fusion partners [BRD4, BRD3, or NSD3]), detection rates were 34.9% (n=22/63) for DNA tests, 60.0% (n=3/5) for ctDNA tests, and 89.7% (n=52/58) for RNA fusion tests (Figure 1E). When comparing tests performed using an assay that tests for NC-defining NUTM1 fusions, RNA fusion assays were much more likely to detect NUTM1 gene fusions than DNA assays (p<0.001, Fisher’s exact). Similarly, both NUT IHC (100.0%, n=99/99) and NUTM1 FISH (91.9%, n=34/37) assays performed significantly better than DNA NGS (p<0.001, Fisher’s exact, for both). Overall, NUT IHC demonstrated the highest performance for diagnosis, followed by NUTM1 FISH and RNA fusion assays, with DNA NGS showing the lowest performance. Additionally, we also compared detection rates from assays that tested for the NUTM1 fusion for all patients who received an IHC test. Of the 99 patients who received an IHC test, detection rates were 100.0% (n=99/99) for IHC tests, 32.1% (n=18/56) for DNA tests, 88.0% (n=44/50) for RNA tests, and 60.0% (n=3/5) for ctDNA tests (Supplementary Table S2).

We examined whether NC detection rates have improved over time. Prior to 2020 (2015–2019), 28.6% (n=2/7) of DNA assays and 66.7% (n=2/3) of RNA assays tested for NUTM1 or common fusion partners (Figure 1F). In this period, NUTM1 fusions were identified in 31.0% (n=13/42) of DNA tests and 75.0% (n=12/16) of RNA fusion tests (Figure 1G). Among assays specifically testing for NUTM1 fusions, detection rates were 44.8% (n=13/29) for DNA tests and 85.7% (n=12/14) for RNA fusion tests. Post-2020 (2020–2024), more assays tested NUTM1 and common partners – 44.4% (n=8/18) for DNA assays and 80.0% (n=12/15) for RNA assays (Figure 1F). DNA tests had a detection rate of 16.1% (n=9/56) and RNA fusion tests had a detection rate of 88.6% (n=39/44) (Figure 1G). Among assays testing for NUTM1 fusions, detection rates were 27.3% (n=9/33) for DNA tests and 92.9% (n=39/42) for RNA fusion tests.

Molecular Features of NC

Next, we examined the DNA cohort, a subset of 100 patients with NC who underwent panel ctDNA/DNA NGS testing. The DNA cohort demonstrated similar demographic and clinical characteristics as the overall cohort when looking at factors like age, sex, race, ethnicity, cigarette smoking history, disease stage, site of primary disease, NUTM1 fusion partner, PD-L1 expression, and tumor mutation burden (Table 1, Figure 2A, Figure 2B, and Supplementary Fig. S4).

Figure 2: Molecular features of NUT carcinoma.

Figure 2:

Open in a new tab

(A) Distribution of TMB in cases that had DNA NGS testing (B) Distribution of PD-L1 expression in cases that had DNA NGS testing (C) Oncoprint of the most common co-occurring gene mutations with clinical interest from DNA NGS tests (D) Prevalence of co-occurring mutations in patients with NC classified by epigenetic, cell cycle, and DNA repair functions (E) Distribution of the exon variants for NUTM1 and the fusion partner genes in the exon fusion cohort (F) Distribution of NUT exon fusion variants in the exon fusion cohort

Molecular analysis identified multiple secondary alterations with potential clinical interest (Supplementary Table S3). Tier 1/2 mutations for other cancer indications, as assigned by the OncoKB classification, included oncogenic alterations in PIK3CA (activating), RET (activating), and FGFR3 (activating) (n=1 each), along with mutations in tumor suppressors ATM (inactivating), BARD1 (inactivating), BRCA1 (inactivating), and TSC1 (n=1 each). Frequent somatically altered genes included LRP1B (10.4%, n=5/48), KMT2D/MLL2 (8.0%, n=7/88), and FAT1 (5.5%, n=3/54) (Figure 2C). Of these relevant mutations, all were found in samples prior to starting cancer therapies. No significant recurrent copy number or large-scale chromosomal loss/gain alterations were seen. Pathway analysis revealed enrichment in epigenetic regulation (57.0%, n=57/100), cell cycle control (26.0%, n=26/100), and DNA repair (24.0%, n=24/100) genes (Figure 2D). Gene ontology analysis found significant enrichment in categories of kinase activity, transcriptional regulation, and DNA repair mechanisms (Supplementary Fig. S5).

The exon fusion cohort included 46 patients with DNA, ctDNA, or RNA fusion tests that reported the exact exon where the fusion occurred on at least one gene. The vast majority of exon fusion sites were upstream of NUTM1 exon 3 (93.2%, n=41/44) (Figure 2E and Supplementary Fig. S6). For the fusion partner gene, more than half of the exon fusion sites in BRD4 were in exon 11 (59.1%, n=13/22), a majority of the exon fusion sites in BRD3 were in exon 10 (75.0%, n=9/12), and all of the exon fusion sites in NSD3 were in exon 7 (100.0%, n=8/8). There were 11 different fusion transcripts spanning NUTM1 and the fusion partner gene (Figure 2F). The most common fusion transcripts of NUTM1 and its fusion partner were BRD4 exon 11::NUTM1 exon 3 (23.9%, n=11/46), BRD3 exon 10::NUTM1 exon 3 (19.6%, n=9/46), and NSD3 exon 7::NUTM1 exon 3 (17.4%, n=8/46). The two transactivation domains in the NUTM1 gene, two bromodomains, and extra-terminal domain in the BRD family genes were retained in all cases.

DISCUSSION

This study represents the largest comprehensive genomic characterization of NC to date. We found that NUT IHC, NUTM1 FISH, and RNA fusion-based NGS assays are preferred over DNA-based NGS assays for detecting NC-defining NUTM1 fusions. Over the last decade, detection rates of NUTM1 fusions decreased for DNA tests and increased for RNA fusion tests, while the coverage of NUTM1 and common partners increased for both DNA and RNA fusion assays. Co-occurring mutations in NC are enriched in epigenetic regulation, cell cycle control, and DNA repair pathways. Moreover, while the exon fusion site of NUTM1 was consistently upstream of exon 3 (93%, n=41/44), there was a wide variety of exon fusion sites for the NUTM1 partner fusion genes, with 11 different fusion transcripts identified.

We confirmed that RNA fusion assays are much more sensitive than DNA assays for detecting NC. This is consistent with prior studies demonstrating that RNA sequencing improves the detection of gene fusions compared to DNA-based methods. For example, in NSCLC diagnosis, driver gene fusions such as NTRK1/2/3 and MET exon 14 skipping variants are more likely to be detected by RNA-based NGS than by DNA-based NGS.31–33 This discrepancy is explained by the technical challenges DNA-based NGS faces in capturing variable fusion sites distributed across a variety of introns and exons. The performance of DNA and ctDNA tests was comparable, consistent with prior NGS data for patients with NSCLC.34 Recent preclinical data suggest that NC is a subtype of squamous cell cancer arising within squamous epithelium and transcriptionally co-classifying with squamous cancer, which should prompt professional societies to recategorize NC as a squamous NSCLC and squamous head and neck cancer.35,36 Our findings suggest that NGS approaches for NC diagnostics should mirror those in the other fusion cancers originating in the lung and head and neck, where RNA fusion testing and DNA testing are performed simultaneously in suspected cases.31–33 Additionally, if NC is suspected, physicians should prioritize NUT IHC, NUTM1 FISH, or RNA fusion-based NGS assays over DNA-based NGS.

When comparing NGS data from prior years (2015–2019) to more recent years (2020–2024), we found that DNA and RNA fusion assays both improved coverage of NUTM1 and common partners (28.6% [n=2/7] to 44.4% [n=8/18] and 66.7% [n=2/3] to 80.0% [n=12/15], respectively). Interestingly, DNA tests showed a decreased detection rate and RNA fusion tests showed an increased detection rate of NC-defining NUTM1 fusions over time (31.0% [n=13/42] to 16.1% [n=9/56] and 75.0% [n=12/16] to 88.6% [n=39/44], respectively). This suggests that while there is increased awareness of including NC genes for both types of NGS assays that are being developed, the most commonly used DNA assays are not being updated with new clinical knowledge. In contrast, RNA fusion assays show improved performance in detecting NUTM1 fusions. This again highlights the reduced sensitivity of DNA assays for NC detection and the need for continuous updating of NGS assays to ensure the identification of NUTM1 fusions and emerging actionable alterations.

Interestingly, patients first diagnosed with NC using an NGS test (DNA, ctDNA, or RNA fusion) had lower 1-year survival than those diagnosed with NUT IHC (25% vs 43%, respectively). This suggests that proactively testing for NC with a focused NUT IHC test may improve survival outcomes compared to relying on incidental NGS testing. Moreover, IHC testing has higher accuracy and quicker turnaround time, leading to faster treatment initiation compared to NGS testing.

On the molecular level, NCs had no microsatellite instability and, on average, low TMB. However, a significant subset (~20%, n=16/73) express PD-L1, consistent with case series in the literature.18–20 These findings suggest that a subset of these tumors may respond to existing immunotherapy-based treatments, as noted in some reports; however, NC is generally considered an immunologically cold tumor, and the benefit of immunotherapy in this setting remains uncertain.36 NC has also been associated with a lack of additional actionable mutations in the literature. Recently, Kroening and Luo et al. reported DNA alterations in 54 patients with NC, which they categorized as primarily involving epigenetic/histone modification and cell cycle regulation pathways. Our study examined DNA alterations in 100 patients with NC and similarly found enrichment in common pathways related to epigenetic regulation, cell cycle control, and DNA repair. Although more than half of the cases did not include a second cancer-associated gene with a significant mutation, we identified several co-occurring alterations of potential biological interest including those in PIK3CA, RET, FGFR3, ATM, BARD1, BRCA1, and TSC1. Ultimately, patient-derived organoids or xenograft models may help determine whether targeting these co-mutations offers any therapeutic benefit.

The genes with the greatest number of co-existing mutations were LRP1B (10.4%, n=5/48), MLL2/KMT2D (8.0%, n=7/88), and FAT1 (5.5%, n=3/54). The tumor suppressor gene LDL receptor-related protein 1B (LRP1B) encodes an LDL receptor protein. Alterations of LRP1B in NC may lead to dysregulation of lipid metabolism, cell migration, and cell survival. The tumor suppressor gene histone-lysine N-methyltransferase 2D (MLL2/KMT2D) encodes a histone methyltransferase involved in chromatin remodeling and epigenetic regulation. Alterations of MLL2/KMT2D in NC may contribute to genomic instability and disruption of normal chromatin remodeling, adding to the existing transcriptional dysregulation in NC. The tumor suppressor gene FAT atypical cadherin 1 (FAT1) encodes a type 1 transmembrane cadherin involved in cell adhesion and communication. FAT1 mutations in NC may contribute to cellular dedifferentiation and metastasis. Alterations in these genes may represent potential therapeutic targets in NC, warranting further preclinical exploration of these genes and pathways involved in future studies. When comparing the co-occurring mutations observed in this NC cohort with those reported in lung squamous cell carcinoma patients with minimal smoking history, several overlapping genes were identified. Reuss et al. examined 2891 tumor samples from patients with squamous NSCLC, of which 2.2% (n=63/2891) had a never-smoking history.37 In this study, all samples were molecularly profiled with DNA-based NGS (Caris Life Sciences using the NextSeq [592 sequenced genes] or NovaSeq 6000 platform [719 sequenced genes]). Among the patients with a never-smoking history, pathogenic genomic alterations in KMT2D were found in 10.0% (n=6/60) of patients. Additionally, Díaz-Gay et al. analyzed 871 patients with lung cancer and a never-smoking history.38 In this cohort, all 871 matched tumor and germline samples underwent DNA-based whole genome sequencing. Out of the 871 patients, 3.6% (n=31/871) had squamous cell carcinoma histology. Among the patients with a never-smoking history and squamous cell carcinoma lung cancer, LRP1B was notably mutated in 16.1% (n=5/31) of patients. Overall, alterations in KMT2D and LRP1B were enriched in both our NC cohort and found in lung squamous cell carcinoma patients with minimal smoking history, suggesting potential shared genomic features.

NGS revealed 11 different fusion transcripts, capturing the heterogeneity of exon fusion sites in the NUTM1 fusion partner gene that results in the oncoprotein. All transcripts included the major functional domains of both proteins. This wide variability further supports the need for diagnostic assays to have sufficient coverage for these fusions.

This study has several limitations. Our cohort was defined by patients with NC with a clinical molecular diagnostics report. Additionally, we did not have access to raw sequencing data. Our analysis was limited by the variability in genes sequenced in each individual’s tumor and the data listed in primary reports. However, we did include important criteria to enhance quality, i.e., only including DNA assays testing >80 genes and only ctDNA tests where the NC fusion was identified. Additionally, all variants called in the NGS reports were reviewed and further curated by a molecular pathologist. Testing bias and survival bias may have influenced the study, as not all patients with NC undergo NGS testing. Lastly, the retrospective nature of our study limits our ability to calculate the sensitivity and specificity of these tests because we do not know how many cases were missed. Although these factors should be considered, our sample size includes all patients with primary reports from an international registry, which is notable considering the rarity of this tumor and the detailed molecular data available.

In conclusion, our findings highlight the superiority of RNA fusion-based NGS assays over DNA-based NGS for diagnosing NUT carcinoma (NC). Molecular analyses reveal the enrichment of mutations in pathways related to epigenetic regulation, cell cycle control, and DNA repair, alongside the identification of several actionable mutations and 11 different NUTM1 fusion transcripts. Overall, the results of our study emphasize the need to prioritize NUT IHC, NUTM1 FISH, or RNA fusion NGS testing for diagnosing NC. Additionally, they highlight the importance of continuous NGS assay refinement and pave the way for future functional studies of co-occurring mutations to improve NC diagnosis and the development of effective treatment for this aggressive cancer.

Supplementary Material

1
2
3
4
5
6
7
8
9

ACKNOWLEDGMENTS

This work was supported by the NIH/NCI (R01CA124633–16 [to C.A. French] and the K12TR004381 [to J. Luo] through Harvard Catalyst, and the Harvard Clinical and Translational Science Center [National Center for Advancing Translational Sciences, National Institutes of Health]), DF/HCC Lung Spore Career Enhancement Program (to J. Luo), 2024–25 American-Italian Cancer Foundation Post-Doctoral Research Fellowship (to F. Paoloni), Lowe Center for Thoracic Oncology, Friends of Jay Dion, the Ryan Richards Foundation, McDevitt Strong, the Max Vincze Foundation, the Victor Family Foundation, the Fortisure Foundation for NUT Carcinoma Research, and the Alexandra Hallock Memorial Fund.

Footnotes

Conflict of Interest Disclosures:

N.R. Mahadevan reports ownership of stocks in AstraZeneca, BioNTech and Roche. P. K. Paik reports personal fees from EMD Serono, AstraZeneca, Janssen, Takeda, Bicara, Mirati, Novartis, and Crown Bioscience outside the submitted work. J. E. Chaft reports grants and personal fees from AstraZeneca, Merck, and Genentech/Roche, grants from Bristol Myers Squibb and BeiGene, and personal fees from Guardant Health, Boehringer Ingelheim, Janssen, Eli Lilly and Company, and Sanofi-Regeneron outside the submitted work. R. Hsu reports consulting for Takeda, EMD Serono, and MJH Life Sciences, honorarium from DAVA Oncology and The Dedham Group, advisory board participation with Oncohost, and research funding to institution from Bristol Myers Squibb Foundation outside the scope of the submitted work. S. A. Piha-Paul reports other support from AbbVie, Inc., ABM Therapeutics, Inc., Acepodia, Inc., Alkermes, Aminex Therapeutics, BioMarin Pharmaceutical, Inc., Boehringer Ingelheim, Bristol Myers Squibb, Cerulean Pharma, Inc., Chugai Pharmaceutical Co., Ltd., Cullinan Oncology, Inc., Curis, Inc., Cyclacel Pharmaceuticals, Daiichi Sankyo, Inc., Eli Lilly, ENB Therapeutics, Epigenetix Inc., Five Prime Therapeutics, F-Star Beta, Limited, Gene Quantum Healthcare, Genmab A/S, Gilead Sciences, Inc., GlaxoSmithKline, Helix BioPharma Corp., Hengrui Pharmaceuticals, Co., Ltd., HiberCell, Inc., Immunomedics, Inc., Incyte Corp., Innovent Biologics, Co., Ltd., Jacobio Pharmaceuticals Co., Ltd., Jazz Pharmaceuticals, Jiangsu Simcere Pharmecutical Co., Ltd., Johnson & Johnson, Loxo Oncology, Inc., Lytix Biopharma AS, Medimmune, LLC, Medivation, Inc., Merck, Sharp and Dohme Corp., Nectin Therapeutics, Ltd., Novartis Pharmaceuticals, NRG Oncology, Nurix, OncoNano Medicine, Inc., Pieris Pharmaceuticals, Inc., Pfizer, Phanes Therapeutics, Principia Biopharma, Inc., ProfoundBio US Co., Puma Biotechnology, Inc., Purinomia Biotech, Inc., Rapt Therapeutics, Inc., Replimune, Roche/Blueprint, Seattle Genetics, Shasqi, Inc., Silverback Therapeutics, Strand Therapeutics, Inc., Synologic Therapeutics, Taiho Oncology, Tallac Therapeutics, Inc., Tesaro, Inc., Theradex Oncology, Toragen Therapeutics, Inc., TransThera Bio, Xencor, Inc., ZielBio, Inc., and F-Star Therapeutics, Limited and grants from NCI/NIH P30CA016672 - Core Grant (CCSG Shared Resources) outside the submitted work; in addition, S. Piha-Paul reports work as a consultant for CRC Oncology and Lilly USA, LLC. S.A. Olwill reports patents for WO2021089588A1 and WO2016177762A1 pending. P. A. Jänne has consulting fees from AstraZeneca, BoehringerIngelheim, Pfizer, Roche/Genentech, Takeda Oncology, ACEA Biosciences, Eli Lilly and Company, Araxes Pharma, Ignyta, Mirati Therapeutics, Novartis, LOXO Oncology, Daiichi Sankyo, Sanofi Oncology, Voronoi, SFJ Pharmaceuticals, Takeda Oncology, Transcenta, Silicon Therapeutics, Syndax, Nuvalent, Bayer, Esai, Biocartis, Allorion Therapeutics, Accutar Biotech and Abbvie, Monte Rosa, Scorpion Therapeutics, Merus, Frontier Medicines, Hongyun Biotechnology and Duality; post-marketing royalties from DFCI owned intellectual property on EGFR mutations licensed to Lab Corp; sponsored research agreements with AstraZeneca, Daichi-Sankyo, PUMA, Boehringer Ingelheim, Eli Lilly and Company, Revolution Medicines and Astellas Pharmaceuticals; stock ownership in Gatekeeper Pharmaceuticals. D. A. Barbie reports personal fees from N-of-One/Qiagen and other support from Xsphera Biosciences outside the submitted work. L. M. Sholl reports receiving research and consulting income to institution from Genentech; consulting income to institution from Eli Lilly; and research support from Bristol-Myers Squibb. S. G. DuBois reported receiving personal fees from Bayer, Amgen, EMD Merck Serono, InhibRx, and Jazz Pharmaceuticals outside the submitted work. G. J. Hanna reports personal fees and nonfinancial support from Naveris and Adela outside the submitted work. G. I. Shapiro reports grants and personal fees from Merck KGaA/EMD Serono; grants from Tango Therapeutics, Bristol Myers Squibb, Pfizer, Eli Lilly, and Merck & Co.; and personal fees from Concarlo Therapeutics, Circle Pharma, Schrodinger, FoRx Therapeutics, and Xinthera outside the submitted work. In addition, he has patents entitled, “Dosage regimen for sapacitabine and seliciclib,” and “Compositions and methods for predicting response and resistance to CDK4/6 inhibition.” C. A. French reports receiving research funding from Boehringer-Ingelheim, and reports receiving consultant fees from Boehringer-Ingelheim. J. Luo reports research support to her institution from Erasca, Genentech, Kronos Bio, Novartis, and Revolution Medicines; honoraria from Targeted Oncology, Physicians’ Education Resource, VJ Oncology, Cancer GRACE, and Community Cancer Education, Inc.; advisory board participation from AstraZeneca and Amgen; and personal fees from Erasca and Genentech; has a patent filed by Memorial Sloan Kettering related to multimodal features to predict response to immunotherapy (PCT/US2023/115872). The remaining authors have no disclosures to report.

REFERENCES

  • 1.French CA, Kutok JL, Faquin WC, et al. Midline carcinoma of children and young adults with NUT rearrangement. J Clin Oncol Off J Am Soc Clin Oncol. 2004;22(20):4135–4139. doi: 10.1200/JCO.2004.02.107 [DOI] [PubMed] [Google Scholar]
  • 2.Chau NG, Ma C, Danga K, et al. An Anatomical Site and Genetic-Based Prognostic Model for Patients With Nuclear Protein in Testis (NUT) Midline Carcinoma: Analysis of 124 Patients. JNCI Cancer Spectr. 2019;4(2):pkz094. doi: 10.1093/jncics/pkz094 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Bauer DE, Mitchell CM, Strait KM, et al. Clinicopathologic features and long-term outcomes of NUT midline carcinoma. Clin Cancer Res Off J Am Assoc Cancer Res. 2012;18(20):5773–5779. doi: 10.1158/1078-0432.CCR-12-1153 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Lauer UM, Hinterleitner M, Horger M, Ohnesorge PV, Zender L. NUT Carcinoma—An Underdiagnosed Malignancy. Front Oncol. 2022;12. doi: 10.3389/fonc.2022.914031 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.French CA. Demystified molecular pathology of NUT midline carcinomas. J Clin Pathol. 2010;63(6):492–496. doi: 10.1136/jcp.2007.052902 [DOI] [PubMed] [Google Scholar]
  • 6.French CA, Rahman S, Walsh EM, et al. NSD3-NUT fusion oncoprotein in NUT midline carcinoma: implications for a novel oncogenic mechanism. Cancer Discov. 2014;4(8):928–941. doi: 10.1158/2159-8290.CD-14-0014 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.French CA. Pathogenesis of NUT midline carcinoma. Annu Rev Pathol. 2012;7:247–265. doi: 10.1146/annurev-pathol-011811-132438 [DOI] [PubMed] [Google Scholar]
  • 8.Grayson AR, Walsh EM, Cameron MJ, et al. MYC, a downstream target of BRD-NUT, is necessary and sufficient for the blockade of differentiation in NUT midline carcinoma. Oncogene. 2014;33(13):1736–1742. doi: 10.1038/onc.2013.126 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Eagen KP, French CA. Supercharging BRD4 with NUT in carcinoma. Oncogene. 2021;40(8):1396–1408. doi: 10.1038/s41388-020-01625-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.French CA, Cheng ML, Hanna GJ, et al. Report of the First International Symposium on NUT Carcinoma. Clin Cancer Res Off J Am Assoc Cancer Res. 2022;28(12):2493–2505. doi: 10.1158/1078-0432.CCR-22-0591 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.French CA. NUT Carcinoma: Clinicopathologic features, pathogenesis, and treatment. Pathol Int. 2018;68(11):583–595. doi: 10.1111/pin.12727 [DOI] [PubMed] [Google Scholar]
  • 12.Haack H, Johnson LA, Fry CJ, et al. Diagnosis of NUT Midline Carcinoma Using a NUT-Specific Monoclonal Antibody. Am J Surg Pathol. 2009;33(7):984–991. doi: 10.1097/PAS.0b013e318198d666 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Mao N, Liao Z, Wu J, et al. Diagnosis of NUT carcinoma of lung origin by next-generation sequencing: case report and review of the literature. Cancer Biol Ther. 2019;20(2):150–156. doi: 10.1080/15384047.2018.1523852 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Chen M, Li S, Jiang L. Clinicopathological molecular characterizations of sinonasal NUT carcinoma: a report of two cases and a literature review. Front Oncol. 2023;13:1296862. doi: 10.3389/fonc.2023.1296862 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Chen M, Zhao S, Liang Z, Wang W, Zhou P, Jiang L. NUT carcinoma of the parotid gland: report of two cases, one with a rare ZNF532-NUTM1 fusion. Virchows Arch Int J Pathol. 2022;480(4):887–897. doi: 10.1007/s00428-021-03253-9 [DOI] [PubMed] [Google Scholar]
  • 16.Liu Y, Li YY, Ke XX, Lu Y. The primary pulmonary NUT carcinomas and some uncommon somatic mutations identified by next-generation sequencing: a case report. AME Case Rep. 2020;4:24. doi: 10.21037/acr-19-168 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Zhang Y, Han K, Dong X, et al. Case Report and Literature Review: Primary Pulmonary NUT-Midline Carcinoma. Front Oncol. 2021;11:700781. doi: 10.3389/fonc.2021.700781 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Kloker LD, Sidiras M, Flaadt T, et al. Clinical management of NUT carcinoma (NC) in Germany: Analysis of survival, therapy response, tumor markers and tumor genome sequencing in 35 adult patients. Lung Cancer Amst Neth. 2024;189:107496. doi: 10.1016/j.lungcan.2024.107496 [DOI] [PubMed] [Google Scholar]
  • 19.Riess JW, Rahman S, Kian W, et al. Genomic profiling of solid tumors harboring BRD4-NUT and response to immune checkpoint inhibitors. Transl Oncol. 2021;14(10):101184. doi: 10.1016/j.tranon.2021.101184 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Kroening G, Luo J, Evans MG, et al. Multiomic Characterization and Molecular Profiling of Nuclear Protein in Testis Carcinoma. JCO Precis Oncol. 2024;8:e2400334. doi: 10.1200/PO.24.00334 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.WHO Classification of Tumours Editorial Board, ed. Head and Neck Tumours: WHO Classification of Tumours. Vol 9. 5th ed. International Agency for Research on Cancer; 2024. [Google Scholar]
  • 22.French CA, Miyoshi I, Kubonishi I, Grier HE, Perez-Atayde AR, Fletcher JA. BRD4-NUT fusion oncogene: a novel mechanism in aggressive carcinoma. Cancer Res. 2003;63(2):304–307. [PubMed] [Google Scholar]
  • 23.Chakravarty D, Gao J, Phillips SM, et al. OncoKB: A Precision Oncology Knowledge Base. JCO Precis Oncol. 2017;2017:PO.17.00011. doi: 10.1200/PO.17.00011 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Suehnholz SP, Nissan MH, Zhang H, et al. Quantifying the Expanding Landscape of Clinical Actionability for Patients with Cancer. Cancer Discov. 2024;14(1):49–65. doi: 10.1158/2159-8290.CD-23-0467 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Landrum MJ, Lee JM, Riley GR, et al. ClinVar: public archive of relationships among sequence variation and human phenotype. Nucleic Acids Res. 2014;42(Database issue):D980–985. doi: 10.1093/nar/gkt1113 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Huang DW, Sherman BT, Lempicki RA. Bioinformatics enrichment tools: paths toward the comprehensive functional analysis of large gene lists. Nucleic Acids Res. 2009;37(1):1–13. doi: 10.1093/nar/gkn923 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Huang DW, Sherman BT, Lempicki RA. Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources. Nat Protoc. 2009;4(1):44–57. doi: 10.1038/nprot.2008.211 [DOI] [PubMed] [Google Scholar]
  • 28.Sherman BT, Hao M, Qiu J, et al. DAVID: a web server for functional enrichment analysis and functional annotation of gene lists (2021 update). Nucleic Acids Res. 2022;50(W1):W216–W221. doi: 10.1093/nar/gkac194 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Tang D, Chen M, Huang X, et al. SRplot: A free online platform for data visualization and graphing. PLOS ONE. 2023;18(11):e0294236. doi: 10.1371/journal.pone.0294236 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Sholl LM, Do K, Shivdasani P, et al. Institutional implementation of clinical tumor profiling on an unselected cancer population. JCI Insight. 2016;1(19):e87062. doi: 10.1172/jci.insight.87062 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Bruno R, Fontanini G. Next Generation Sequencing for Gene Fusion Analysis in Lung Cancer: A Literature Review. Diagnostics. 2020;10(8):521. doi: 10.3390/diagnostics10080521 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Owen D, Ben-Shachar R, Feliciano J, et al. Actionable Structural Variant Detection via RNA-NGS and DNA-NGS in Patients With Advanced Non–Small Cell Lung Cancer. JAMA Netw Open. 2024;7(11):e2442970. doi: 10.1001/jamanetworkopen.2024.42970 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Davies KD, Lomboy A, Lawrence CA, et al. DNA-Based versus RNA-Based Detection of MET Exon 14 Skipping Events in Lung Cancer. J Thorac Oncol Off Publ Int Assoc Study Lung Cancer. 2019;14(4):737–741. doi: 10.1016/j.jtho.2018.12.020 [DOI] [PubMed] [Google Scholar]
  • 34.Thompson JC, Yee SS, Troxel AB, et al. Detection of Therapeutically Targetable Driver and Resistance Mutations in Lung Cancer Patients by Next-Generation Sequencing of Cell-Free Circulating Tumor DNA. Clin Cancer Res Off J Am Assoc Cancer Res. 2016;22(23):5772–5782. doi: 10.1158/1078-0432.CCR-16-1231 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Durall RT, Huang J, Wojenski L, et al. The BRD4-NUT Fusion Alone Drives Malignant Transformation of NUT Carcinoma. Cancer Res. 2023;83(23):3846–3860. doi: 10.1158/0008-5472.CAN-23-2545 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Luo J, Bishop JA, DuBois SG, et al. Hiding in plain sight: NUT carcinoma is an unrecognized subtype of squamous cell carcinoma of the lungs and head and neck. Nat Rev Clin Oncol. Published online February 3, 2025. doi: 10.1038/s41571-025-00986-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Reuss JE, Zaemes J, Gandhi N, et al. Comprehensive molecular profiling of squamous non-small cell lung cancer reveals high incidence of actionable genomic alterations among patients with no history of smoking. Lung Cancer. 2025;200. doi: 10.1016/j.lungcan.2025.108101 [DOI] [PubMed] [Google Scholar]
  • 38.Díaz-Gay M, Zhang T, Hoang PH, et al. The mutagenic forces shaping the genomic landscape of lung cancer in never smokers. Published online May 17, 2024:2024.05.15.24307318. doi: 10.1101/2024.05.15.24307318 [DOI] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1
NIHMS2100430-supplement-1.pdf (116.8KB, pdf)
2
NIHMS2100430-supplement-2.pdf (99.9KB, pdf)
3
NIHMS2100430-supplement-3.pdf (90.2KB, pdf)
4
NIHMS2100430-supplement-4.pdf (113.5KB, pdf)
5
NIHMS2100430-supplement-5.pdf (139.3KB, pdf)
6
NIHMS2100430-supplement-6.pdf (77.2KB, pdf)
7
NIHMS2100430-supplement-7.docx (37.8KB, docx)
8
NIHMS2100430-supplement-8.docx (32.3KB, docx)
9
NIHMS2100430-supplement-9.docx (51.9KB, docx)

Data Availability Statement

The data generated in this study are available within the article and its supplementary data files.

RESOURCES