Abstract
Background
MET fusions represent emerging therapeutic targets in solid tumors; however, functional interpretation of non-canonical variants remains poorly understood, posing a major challenge for precision oncology.
Methods
We conducted a multicenter, pan-cancer study analyzing 23,299 clinical samples using DNA-based next-generation sequencing (NGS) to profile MET fusions. Transcriptional validation was performed using RNA-based NGS on available samples. Preliminary clinical outcomes were assessed in four patients with advanced malignancies harboring uncommon MET fusions who received MET tyrosine kinase inhibitor therapy.
Results
We identified 116 MET fusions (incidence: 0.5%), with 55.2% (64/116) classified as uncommon fusions. These uncommon fusions were stratified into: Group A (5’-retained, n = 12), Group B (intergenic/exonic breakpoints, n = 19), Group C (rare partners, n = 23), and Group D (dual fusions, n = 10). RNA validation revealed an overall low transcriptional consistency of 43.8% (14/32) for uncommon fusions, versus 100% for canonical fusions (PTPRZ1::MET, CAPZA2::MET). Notably, most 5’-retained fusions were transcriptionally silent, while some intergenic fusions resolved into expressed canonical partners (e.g. PTPRZ1::MET). Therapeutically, all four MET inhibitor-treated patients achieved partial responses, including pediatric diffuse midline gliomas (DMG) (median OS: 11.2 months) and lung adenocarcinoma (median OS: 34 months), demonstrating preliminary clinical activity.
Conclusions
uncommon MET fusions are heterogeneous at genomic and transcriptional levels. DNA-level findings often do not predict functional transcripts, underscoring the necessity of RNA-based confirmation for clinical interpretation. Despite low overall consistency, a subset retains therapeutic potential. We propose a refined diagnostic framework integrating DNA-based stratification and RNA validation to guide the management of MET-altered cancers in precision oncology workflows.
Supplementary Information
The online version contains supplementary material available at 10.1186/s12967-026-07689-y.
Keywords: Biomarker-driven therapies, MET, MET inhibitor, NGS, MET
Introduction
The c-mesenchymal-epithelial transition factor (MET) is a transmembrane receptor tyrosine kinase, that regulates critical signaling pathways governing cell proliferation, migration, and invasion [1, 2]. Dysregulation of MET signaling, through mechanisms such as amplification, exon 14 skipping mutations, and gene rearrangements, is a recognized oncogenic driver across diverse malignancies [2–6]. Guideline-recommended biomarkers for MET amplification and exon 14 skipping in non-small cell lung cancer (NSCLC) are established, while clinical management of MET fusions remains less defined, lacking standardized detection and therapeutic protocols.
The clinical study of MET fusions has been constrained by their rarity and molecular heterogeneity. Unlike amplification or exon 14 skipping, MET fusions involve diverse partner genes and breakpoints, rendering conventional detection METhods like FISH and RT-PCR inadequate for comprehensive screening. The advent of next-generation sequencing (NGS) has begun to illuminate the landscape of MET fusions, reporting incidences that vary by tumor type, such as 0.1%-0.29% in lung cancer [7–9] and 1%-10% in brain cancer [7, 10]. Importantly, emerging evidence from case reports and small series suggests that tumors harboring certain MET fusion, such as PTPRZ1:MET, ST7:MET, CAPZA2:MET, can respond to MET tyrosine kinase inhibitors (TKIs) like crizotinib, capmatinib, tepotinib [8, 9, 11–13]. These findings position MET fusions as promising actionable targets.
However, a significant knowledge gap persists. Current studies primarily focuse on canonical, 3’-retained MET fusions that preserve the intact kinase domain. In clinical practice, DNA-based NGS frequently identifies a spectrum of uncommon or architecturally complex MET rearrangements, including 5’-retained fusions, intergenic rearrangements, exonic breakpoints, and cases with dual fusions [14–18]. The functional significance, transcriptional output, and clinical actionability of these uncommon variants are poorly understood, posing a substantial interpretative challenge for oncologists. This underscores the critical need for large-scale, systematic studies that integrate DNA and RNA sequencing to molecularly characterize these variants and assess their therapeutic relevance.
To address this, we conducted a multicenter, pan-cancer study of 23,299 clinical samples. We aimed to: (1) delineate the genomic spectrum and classify uncommon MET fusions, (2) evaluate their transcriptional reliability through paired RNA sequencing, and (3) report preliminary clinical outcomes of MET inhibitor therapy in patients harboring such fusions. Our work establishes a refined stratification framework to guide the clinical interpretation and management of uncommon MET fusions.
Materials and Methods
Patients and samples
Between Dec. 2018 and Oct. 2023, clinical samples across nine cancer types were collected from six hospital centers (Shanxi Province Cancer Hospital, Shanxi Bethune Hospital Cancer Center, Cancer Hospital of China Medical University, The First Affiliated Hospital, Zhejiang University School of Medicine, Huangdong Sanjiu Brain Hospital, Children’s Hospital of Chongqing Medical University), including bile duct cancer (BDC), brain cancer (BC), breast cancer (BRC), lung cancer (LUC), gastric cancer (GC), hepatocellular carcinoma (HCC), intestine cancer (IC), and ovarian cancer (OC), soft tissue sarcoma (STS). Tumor tissue and peripheral blood sample from each patient were collected and preserved for subsequent testing. Tumor purity was estimated via histopathological examination, more than 10% purity allowed for this study.
DNA-based NGS
Genomic DNA was isolated from formalin-fixed paraffin-embedded (FFPE) sections using the QIAamp DNA FFPE Tissue Kit (QIAGEN) or from fresh-frozen tissues using the QIAamp DNA Kit (QIAGEN), following manufacturer protocols optimized for genotyping and pharmacogenomic applications. DNA concentration was quantified using the Qubit 2.0 FluoroMETer with the Qubit dsDNA High Sensitivity Assay Kit (Thermo Fisher Scientific), ensuring a minimum input of 50 ng DNA with fragment lengths > 500 bp. Genomic DNA was sheared (Covaris M220), and libraries prepared (KAPA HTP Kit) were hybridized with the Onco PanScan panel (targeting 825 genes). Sequencing was performed on Illumina platforms (NovaSeq/Xten) with median depth ≥ 500x per tumor tissue sample and 100x per control.
Raw data underwent adapter trimming and quality filtering using Trimmomatic (Ver 0.36) to remove adaptor sequences and low-quality bases. Reads were aligned to the human reference genome (GRch37) via the Burrows-Wheeler Aligner (v0.7.10) with default paraMETers. Somatic single nucleotide variations (SNVs) and small insertion/deletions (indels) were independently called using muTech (Broad Institute) and Strelka (Illumina), respectively. Structural variations (SVs) were identified using GeneFuse (v0.6.1) with stringent criteria: variant allele frequency (VAF) ≥0.5%) and ≥4 high-confidence supporting reads. All variations were annotated using Oncotator and Ensemble Variant Effect Predictor (VEP) to assess functional impacts. Population-level filtering excluded variants with minor allele frequency (MAF) > 0.001 in the 1000 Genomes Project dataset.
RNA-based NGS
Total RNA was isolated from FFPE sections using the RNeasy FFPE Kit (QIAGEN)or from fresh-frozen tissues using the AllPrep DNA/RNA Mini Kit (QIAGEN), following manufacturer protocols. Due to RNA fragmentation in FFPE samples, RNA integrity was assessed via DV200 index, which quantifies the percentage of RNA fragments > 200 nucleotides [19, 20]. RNA input MET stringent thresholds: ≥150 ng total RNA, a fragment size of 150–350 bp, and DV200 ≥30%.
A two-step reverse transcription (RT) protocol was employed: first-strand cDNA synthesis using the SuperScript VILO cDNA Synthesis Kit (Thermo Fisher Scientific), followed by second-strand synthesis with the Abclonal Second Strand Synthesis Module (Abclonal). Libraries were constructed using the KAPA HyperPrep Kit (Kapa Biosystems) and enriched via hybridization capture with the Fusion-Capture Panel (Genetron Health), targeting 395 cancer-associated genes (including MET). Sequencing was performed on Illumina NovaSeq or Xten platforms. Structural variations were called using a dual threshold: VAF ≥0.5% and ≥4 unique high-quality reads.
Results
MET fusions identified in clinical samples
We analyzed 23,299 clinical samples across nine cancer types (see cohort breakdown in Fig. 1) to comprehensively profile MET fusions. Using DNA-based NGS, we identified 116 tumor samples harboring MET fusions, with an overall incidence of 0.5%. The incidence varied by tumor type (Fig. 2A), being highest in brain cancer (1.2%, 90/7,495), followed by ovarian cancer (0.47%, 1/214) and bile duct cancer (0.42%, 3/711). Lower incidences were observed in gastric cancer (0.39%), breast cancer (0.28%), hepatocellular carcinoma (0.19%), soft tissue sarcoma (0.18%), intestine cancer (0.18%), and lung cancer (0.09%).
Fig. 1.
Workflow of uncommon MET fusion identification
Fig. 2.
Uncommon MET fusion classification and distribution. (A) incidences of MET fusions in LUC, BC, IC, STS, GC, BDC, HCC, BRC, and OC. (B) numbers of common and uncommon MET fusions in all nine cancer types. Classification of uncommon MET fusion across four subgroups A, B, C, and D in (C) and distribution in different cancer types (D). Group A: single 5’-retained MET fusions; Group B: intergenic/exonic breakpoints; Group C: rare/novel partners; Group D: dual MET fusions. LUC: lung cancer; BC: brain cancer; IC: intestine cancer; STS: soft tissue sarcoma; GC: gastric cancer; BDC: bile duct cancer; HCC: hepatocellular carcinoma; BRC: breast cancer; and OC: ovarian cancer
Classification of common and uncommon MET fusion and specificity in different cancer types
MET fusions were classified as “common” if they preserved the predicted intact TKD and detected in > 3 samples. This group (n = 52) consisted predominantly of PTPRZ1::MET (n = 24), ST7::MET (n = 16), and CAPZA2::MET (n = 12). The remaining 64 fusions (55.2%) were classified as “uncommon.” The proportion of uncommon fusions was notably high in lung (87.5%, 7/8), intestine (100%, 3/3), soft tissue sarcoma (75%, 3/4), gastric (80%, 4/5), and bile duct cancers (66.7%, 2/3) (Fig. 2B, C).
To systematically interpret this heterogeneity, we stratified uncommon fusions into four subgroups: Group A (single 5’-retained MET fusions, n = 12); Group B (breakpoints in intergenic regions or exons, potentially causing frameshifts, n = 19); Group C (rare or novel fusion partners, n = 23); and Group D (samples harboring dual MET fusions, n = 10) (Fig. 2C). While common fusions were highly enriched in brain cancer (92.3%, 48/52), uncommon fusions showed a broader distribution across cancer types (Fig. 2D).
Genomic characteristics of common and uncommon MET fusions
All common MET fusions resulted from intrachromosomal rearrangements. In contrast, 20.3% (13/64) of uncommon fusions involved interchromosomal rearrangements, with the highest proportion in Group C (26.1%) (Fig. 3A, B).
Fig. 3.
Genomic rearrangement patterns and structural features of uncommon MET fusions. (A) Distribution of intrachromosomal and interchromosomal rearrangements in common and uncommon MET fusions. (B) Rearrangement patterns in subgroup of uncommon MET fusions. (C) enrichment of break points in four subgroups of uncommon MET fusions. (D) Circus plot summarizing distribution of uncommon MET fusion breakpoints across different subgroups A, B, C, and D
Analysis of MET breakpoints revealed that 100% (52/52) of common fusions had breakpoints in MET intron 1. For uncommon fusions (74 breakpoints analyzed, counting dual fusions in Group D separately), 70.3% (52/74) also occurred in intron 1, but with varying frequencies across subgroups (Group A: 58.3%; B: 52.6%; C: 91.3%; D: 70.0%). Notably, 13.5% (10/74) of breakpoints were located within MET exons, and 5.4% (4/74) were in intron 14, potentially leading to exon 14 skipping (Fig. 3C, D and Supplementary Table S1).
Transcriptional validation of MET fusions
To assess the functional output of fusions, we performed RNA-based NGS on available samples. For common fusions, complete consistency between DNA and RNA findings was 100% for PTPRZ1:MET (11/11) and CAPZA2:MET (8/8), but only 66.7% for ST7:MET (4/6). In one soft tissue sarcoma case, DNA-identified ST7:MET was expressed as CAPZA2:MET at the RNA level (partial consistency), and one brain cancer case showed no fusion transcript (complete inconsistency) (Table 1).
Table 1.
Analyses of 52 common MET fusions identified by DNA and RNA-based NGS
| Classification | Cancer type | DNA NGS | RNA NGS | Consistency |
|---|---|---|---|---|
|
PTPRZ1::MET (n = 24) |
BC | PTPRZ1::MET (int1::int1) (n = 8) | PTPRZ1::MET (ex1::ex2) (n = 3) | 100% (11/11) |
| Failed (n = 3) | ||||
| N/A (n = 2) | ||||
| BC | PTPRZ1::MET (int2::int1) (n = 9) | PTPRZ1::MET (ex2:::ex2) (n = 7) | ||
| N/A (n = 2) | ||||
| BC | PTPRZ1::MET (int3::int1) (n = 4) | PTPRZ1::MET (ex3::ex2) (n = 1) | ||
| N/A (n = 3) | ||||
| BC | PTPRZ1::MET (int6::int1) (n = 1) | N/A (n = 1) | ||
| BC | PTPRZ1::MET (int8::int1) (n = 1) | N/A (n = 1) | ||
| BC | PTPRZ1::MET (int13::int1) (n = 1) | Failed (n = 1) | ||
|
ST7::MET (n = 16) |
BC | ST7::MET (int1::int1) (n = 7) | ST7:MET (ex1::ex2) (n = 3) | 66.7% (4/6) |
| Failed (n = 1) | ||||
| N/A (n = 3) | ||||
| STS | ST7::MET (int1::int1) (n = 1) | CAPZA2-MET(ex1::ex2) | ||
| GC | ST7::MET (int1::int1) (n = 1) | N/A (n = 1) | ||
| BC | ST7::MET (int2::int1) (n = 4) | ST7::MET (ex2::ex2) (n = 1) | ||
| Neg (n = 1) | ||||
| N/A (n = 2) | ||||
| BC | ST7::MET (int3::int1) (n = 2) | N/A | ||
| LC | ST7::MET (int3::int1) (n = 1) | N/A | ||
|
CAPZA2::MET (n = 12) |
BC | CAPZA2::MET (int1::int1) (n = 8) | CAPZA2::MET (ex1::ex2) (n = 7) | 100% (8/8) |
| N/A (n = 1) | ||||
| BC | CAPZA2::MET (int2::int1) (n = 3) | CAPZA2::MET (ex2::ex2) (n = 1) | ||
| Failed (n = 1) | ||||
| N/A (n = 1) | ||||
| BDC | CAPZA2::MET (int9::int1) (n = 1) | N/A (n = 1) |
For uncommon MET fusions, transcriptional validation revealed an overall low consistency of 43.8% (14/32) (Fig. 4A, Table 2). The complete consistency rates were: Group A (5’−retained): 14.3% (1/7); Group B (intergenic/exonic): 18.2% (2/11); Group C (rare partners): 63.6% (7/11); Group D (dual fusions): 0% (0/3).
Fig. 4.
Consistency of uncommon MET fusions validated by DNA-based NGS and RNA-based NGS. (A) consistency of different subgroups of uncommon MET detected by DNA- based NGS and RNA-based NGS. (B-E) Structural and molecular characterization of four representative MET fusions (intergenic::MET fusion transferred into PTPRZ1::MET and ZKSCAN1::MET, respectively in B and C; E14 skipping in dual MET fusions in C, and loss expression of KANK1::MET in D)
Table 2.
Comparison of 64 uncommon MET fusions identified from by DNA and RNA-based NGS
| Classification | Cancer type | DNA NGS | RNA NGS | Consistency |
|---|---|---|---|---|
| Group A | IC | MET::CFTR (int12::int3) | MET::CFTR (ex12::ex4) | 14.3% (1/7) |
| BC | MET::EXOC4 (int1::int4) | Neg | ||
| BC | MET::PODXL (int2::downstream) | Neg | ||
| GC | MET::TES (ex21::int3) | Neg | ||
| HCC | MET::ASZ1 (ex1::int2) | Neg | ||
| OC | MET::CAV1 (int1::int2) | Neg | ||
| STS | MET::IMMP2L (int1::int4) | Neg | ||
| BC | MET::CNTNAP2 (int1::int13) | N/A | ||
| BC | MET::PTPRZ1 (int1::int13) | N/A | ||
| BC | MET::ST7 (int1::int1) | N/A | ||
| GC | MET::ST7 (int1::int1) | N/A | ||
| LC | MET::HLA-DRB1 (int14::int4) | N/A | ||
| Group B | BC | Intergenic::MET (-::int1) | PTPRZ1::MET (ex1::ex2) | 18.2% (2/11) |
| BC | Intergenic::MET (-::int13) | ZKSCAN1::MET (ex1::ex14) | ||
| BC | KANK1::MET (int7::ex14) | KANK1::MET (ex7::ex15) | ||
| BC | PTPRZ1::MET (ex12::int1) | PTPRZ1::MET (ex12::ex2) | ||
| BC | PTPRZ1::MET (int10::ex2) | PTPRZ1::MET (ex10::ex2) | ||
| BC | Intergenic::MET (-::int1) | Neg | ||
| BC | MCM7::MET (ex1::ex1) | Neg | ||
| BC | Intergenic::MET (-::int19) | Neg | ||
| IC | Intergenic::MET (-::int1) | Neg | ||
| IC | ZKSCAN1::MET (ex4::int14) | Neg | ||
| LC | Intergenic::MET (-::int1) | Neg | ||
| BC | Intergenic::MET (-::int1) | Failed | ||
| BC | Intergenic::MET (-::int1) | N/A | ||
| BC | Intergenic::MET (-::int1) | N/A | ||
| BC | Intergenic::MET (-::int1) | N/A | ||
| BC | PTPRZ1::MET (int10::ex5) | N/A | ||
| BC | ZKSCAN1::MET (int1::ex13) | N/A | ||
| GC | Intergenic::MET (-::int1) | N/A | ||
| BRC | ST7::MET (int1::ex17) | N/A | ||
| Group C | BC | BMT2::MET (int1::int1) | BMT2::MET (ex1::ex2) | 63.6% (7/11) |
| BC | CLIP2::MET (int1::int1) | CLIP2::MET (ex1::ex2) | ||
| BC | DNAJB6::MET (int1::int1) | DNAJB6::MET (ex1::ex2) | ||
| BC | HSF2::MET (int1::int1) | HSF2::MET (ex1::ex2) | ||
| BC | LSM8::MET (int3::int1) | LSM8::MET (ex3::ex2) | ||
| - | PTPRZ1::MET (ex2::ex2) | |||
| BC | NXPH1::MET (int2::int1) | NXPH1::MET (ex2::ex2) | ||
| BC | SEMA3A::MET (int1::int1) | SEMA3A::MET (ex1::ex2) | ||
| BC | VKORC1L1::MET (int1::int1) | VKORC1L1::MET (ex1::ex2) | ||
| - | CUX1::MET (ex3::ex2) | |||
| LC | PRKAR1A::MET (int5::int14) | PRKAR1A::MET (ex5::ex15) | ||
| BC | TNRC6B::MET (int1::int1) | Neg | ||
| LC | TFEC::MET (int1::int1) | Neg | ||
| BC | DPP6::MET (int16::int1) | N/A | ||
| BC | MAGI2::MET (int7::int1) | Failed | ||
| BC | MARK3::MET (int1::int1) | N/A | ||
| BC | PPP1R9A::MET (int1::int1) | Failed | ||
| BC | PTN::MET (int4::int1) | N/A | ||
| BC | WASL::MET (int1::int1) | Failed | ||
| BC | ZNF138::MET (int2::int1) | N/A | ||
| LC | LRFN2::MET (int1::int1) | Failed | ||
| LC | SEMA3D::MET (int2::int11) | N/A | ||
| STS | CAV2::MET (int1::int1) | N/A | ||
| BDC | PLGRKT::MET (int1::int1) | N/A | ||
| GC | TES::MET (int1::int1) | N/A | ||
| Group D | BC | CTTNBP2::MET (int4::int11) | Failed | 0% (0/3) |
| MET::Intergenic (int11::-) | ||||
| BC | PTPRZ1::MET (int1::int1) | PTPRZ1::MET (ex1::ex2) | ||
| PTPRZ1::MET (int1::int1) | Neg | |||
| BDC | KANK1::MET (int7::int14) | KANK1::MET (ex7::ex15) | ||
| MET::KIF3B (int1::int10) | Neg | |||
| LC | MET::DENND2A (ex14::int3) | MET::DENND2A (ex13::ex4) | ||
| CD47::MET (int7::ex14) | CD47::MET (ex7::ex15) | |||
| BC | DOCK4::MET (int20::int1) | N/A | ||
| DOCK4::MET (int1::int1) | ||||
| BC | MET::PTPRZ1 (int1::int10) | N/A | ||
| PTPRZ1::MET (int10::int1) | ||||
| BC | MET::ST7 (int1::int1) | N/A | ||
| FAM133B::MET (int1::int1) | ||||
| BC | Intergenic::MET (-::int1) | N/A | ||
| POM121C::MET (int1::int1) | ||||
| BC | PTPRZ1::MET (int2::int1) | N/A | ||
| PTPRZ1::MET (int1::int1) | ||||
| STS | Intergenic:MET (-::int1) | N/A | ||
| Intergenic::MET (-::int1) |
Key mechanistic insights emerged: 1) Most MET 5’-retained fusions (Group A, 6/7 tested) were transcriptionally silent; 2) Several intergenic::MET fusions (Group B) resolved into expressed canonical fusions (e.g., PTPRZ1::MET, ZKSCAN1::MET) at the RNA level (Fig. 4B, C); 3) In Group D (dual fusions), RNA sequencing often revealed simplification (e.g., expression of only one fusion) or complex processing, such as exon 14 skipping in both fusions of a lung adenocarcinoma case (Fig. 4D, E).
Clinical outcomes of targeted therapy of MET inhibitor
This study analyzed four cases of advanced malignancies harboring uncommon MET fusions treated with MET inhibitors, including two pediatric diffuse midline gliomas (DMG) and two adult lung adenocarcinomas (Table 3). Three patients #63, #65, and #84 received first-line treatment of MET inhibitor. One patient #61 was treated with a MET inhibitor as second-line therapy, with first-line temozolomide (TMZ) stopped after only one cycle due to disease progression. All patients derived clinical response from MET inhibitors, with partial response (PR) observed in 4/4 cases (100%).
Table 3.
Clinical response to MET inhibitor
| Case No. | Sex | Age | Cancer type | Classification | DNA NGS | RNA NGS | Co-occuring driver gene mutations | First-line therapy/ DOT (months) |
PFS1 (months) | Second-line therapy/DOT (months) | PFS2 (months) | Best response to MET inhibitor | OS (months) |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 61 | F | 4 | diffuse midline glioma | Uncommon fusion | DOCK4::MET (int20::int1) | N/A |
H3F3A p.K28M TP53 p.R273C |
TMZ+radiotherapy/ < 1 |
< 1 | crizotinib/2.4 | 2.4 | PR | 8.3 |
| DOCK4::MET (int1::int1) | |||||||||||||
| 63 | F | 68 | lung adenocarcinoma | Uncommon fusion | PRKAR1A::MET (int5::int14) | PRKAR1A::MET (ex5::ex15) | none |
crizotinib/ 1 |
3.5 | crizotinib/9 | 9.1 | nCR | 29 |
| 65 | F | 31 | lung adenocarcinoma with brain METastasis | Uncommon fusion | MET::DENND2A (ex14::int3) | MET::DENND2A (ex13::ex4) | none | crizotinib/23 | 23 | PLB-1001 | > 16 | PR | > 39 |
| CD47::MET (int7::ex14) | CD47::MET (ex7::ex15) | ||||||||||||
| 84 | M | 9 | diffuse midline glioma | Uncommon fusion | Intergenic::MET (-::int1) | N/A |
H3F3A p.K28M TP53 c0.267deletion |
crizotinib+TMZ/ 8.4 |
8.4 | best supportive care | - | PR | 14 |
DOT: duration of therapy; nCR: nearly complete response; TMZ: temozolomide; PR: partial response
Two pediatric DMG cases 61# and 84# exhibited aggressive clinical courses before targeted therapy. Case 61# (4-year-old female) carried DOCK4::MET. Despite rapid progression on first-line temozolomide (TMZ) plus radiotherapy (PFS1 < 1 month), crizotinib as second-line therapy induced a partial response (PR) with PFS2 of 2.4 months better than PFS1, though overall survival (OS) remained limited to 8.3 months. Case 84# (9-year-old male) harbored an intergenic::MET fusion. Combined crizotinib and TMZ achieved a PR with prolonged PFS1 of 8.4 months, yet subsequent progression led to an OS of 14 months. In contrast, two adult lung adenocarcinoma cases demonstrated durable responses. Case 63# (68-year-old female) with a PRKAR1A::MET fusion received crizotinib as first-line therapy (duration of therapy [DOT]: 1 month) achieving a near-complete response (nCR). Upon disease progression, crizotinib rechallenge achieved a PR with PFS2 of 9.1 months and OS of 29 months, was reported in the previous study [21]. Case 65# (31-year-old female, brain METastasis) carried double MET fusions of MET::DENND2A and CD47::MET. First-line crizotinib yielded a remarkable PFS1 of 23 months, followed by PLB-1001 (a next-generation MET inhibitor) with ongoing PR ( > 16 months PFS2) and survival.
Discussion
In clinical practice, it lacks a standard METhod to detect MET fusions, and it is a significant challenge for clinicians and pathologists to interpret whether the MET fusion is functional or unreliable genomic structure variation; should MET inhibitors be recommended for patients with advanced cancer carrying MET fusion? To address these issues, we performed a large-scale study to identify MET fusions and assessed the clinical outcomes of targeted therapy of MET inhibitor.
The high frequency of uncommon MET fusions (55.2% of all MET rearrangements) contrasts sharply with the predominance of canonical partners in other actionable kinase fusions (e.g., EML4::ALK), explaining why traditional FISH or RT-PCR assays are not suited for MET fusion detection [2, 22]. While DNA-based NGS is a powerful discovery tool, our validation data align with concerns raised for ALK, ROS1, and RET, where DNA-level calls can be misleading [23–25]. We propose a refined diagnostic framework: DNA-based NGS for initial screening, followed by mandatory RNA confirmation for any uncommon or complex MET fusion call. This two-step approach is crucial to distinguish passive genomic rearrangements from bona fide, transcriptionally active oncogenic drivers.
Mechanistically, our subgroup analysis offers insights. The near-total lack of RNA expression from Group A (5’−retained) fusions suggests they rarely produce stable fusion proteins. The phenomenon of intergenic::MET fusions resolving into expressed canonical fusions (e.g., PTPRZ1::MET) in Group B has been observed with other kinases and may result from transcriptional processing that removes intervening non-genic sequences [17, 20, 26]. Moreover, exon 14 skipping of MET often results from mutations in the poly-pyrimidine tract, splice donor, or acceptor sites [3, 27, 28], as well as the whole exons 1 to 14 deleted by genomic structure rearrangement events [21, 29]. In our study, breakpoints in exon 14 of MET also resulted in the exon 14 skipping at the transcriptional level in two cases (one with brain cancer and one with lung cancer), probably due to an alternative spicing event. These findings underscore that the final oncogenic product is defined at the RNA level, not solely by DNA breakpoints.
Clinically, MET exon 14 skipping mutations and high-level amplifications (fold-change ≥10) represent established, guideline-recommended biomarkers for MET-targeted therapy in non-small cell lung cancer (NSCLC), with specific inhibitors like capmatinib and tepotinib approved for these indications [30–32]. However, to date, the standard TKI treatment has not been determined for patients with cancer carrying MET fusion. Despite the overall low functional consistency, our preliminary data indicate that the subset of uncommon MET fusions validated at the RNA level can be sensitive to MET inhibition. Responses were observed in aggressive cancers like H3F3A-mutant pediatric glioma and lung adenocarcinoma, with notable durability in the latter. This suggests that biological context (e.g., co-occurring drivers, blood-brain barrier) significantly influences therapeutic efficacy. The ongoing FUGEN trial (NCT06105619) of bozitinib in PTPRZ1::MET glioma further supports the clinical potential of targeting MET fusions [12]. Our work thus provides a stratification strategy to identify which patients with uncommon DNA-level findings are most likely to benefit from targeted therapy.
This study has limitations. The therapeutic cohort is small (n = 4) and retrospective, precluding definitive efficacy conclusions. RNA material was unavailable for all DNA-positive cases, which may affect the comprehensive assessment of transcriptional consistency. Future prospective studies in defined patient cohorts are needed to validate the clinical utility of this stratification framework. Future prospective studies in defined cohorts are needed to validate the clinical utility of this stratification framework.
In conclusion, uncommon MET fusions represent a heterogeneous collection of genomic alterations. We advocate for a two-stage interpretation framework comprising: (i) DNA-based NGS screening categorizes the structural variant (Groups A-D); and (ii) subsequent RNA validation to determine the transcriptional outcome, which is critical for defining clinical actionability, especially in Groups A, B, and D. In addition, our preliminary clinical observations suggest potential benefit from MET inhibitor in a subset of patients, highlighting the need for prospective studies to validate the therapeutic potential of these rare alterations.
Electronic supplementary material
Below is the link to the electronic supplementary material.
Acknowledgements
This study was supported by the Science and Technology Cooperation and Exchange Program of Shanxi Provincial Science and Technology Department (No. 202204041101042), the Innovation Talent Team of Shanxi Province (No. 202204051001031) and the central government guides local funds for science and technology development of Shanxi Provincial Science and Technology Department (No. YDZJSX20231A067).
Funding
This work was supported by the Liaoning Provincial Science and Technology Program Joint Program (No. 2025JH2/101800096).
Data availability
The data that support the findings of this study are available from the corresponding author upon request.
Declarations
Ethics statement
The protocol was approved by the Ethics Committee of medical research, Shanxi Bethune Hospital Cancer Center (No. SBQKL-2022–113). All the patients were fully informed and signed the informed consent form adhered to the ethical principles of the Declaration of Helsinki and Good Clinical Practice Guidelines.
Conflict of interest
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Footnotes
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Wenhui Yang, Yanxiang Zhang, Tonghui Ma and Haiyang Liang contributed equally to this study.
References
- 1.Pal K, Bandyopadhyay A, Zhou XE, Xu Q, Marciano DP, Brunzelle JS, et al. Structural basis of TPR-Mediated oligomerization and Activation of oncogenic fusion kinases. Structure. 2017;25(6):867–77 e863. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Guo R, Luo J, Chang J, Rekhtman N, Arcila M, Drilon A. MET-dependent solid tumours - molecular diagnosis and targeted therapy. Nat Rev Clin Oncol. 2020;17(9):569–87. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Salgia R, Sattler M, Scheele J, Stroh C, Felip E. The promise of selective MET inhibitors in non-small cell lung cancer with MET exon 14 skipping. Cancer Treat Rev. 2020;87:102022. [DOI] [PubMed] [Google Scholar]
- 4.Tong JH, Yeung SF, Chan AW, Chung LY, Chau SL, Lung RW, et al. MET amplification and Exon 14 splice site mutation define unique molecular subgroups of non-small cell lung carcinoma with poor prognosis. Clin Cancer Res. 2016;22(12):3048–56. [DOI] [PubMed]
- 5.Riedel R, Fassunke J, Tumbrink HL, Scheel AH, Heydt C, Hieggelke L, et al. Resistance to MET inhibition in MET-dependent NSCLC and therapeutic activity after switching from type I to type II MET inhibitors. Eur J Cancer. 2023;179:124–35. [DOI] [PubMed] [Google Scholar]
- 6.Stransky N, Cerami E, Schalm S, Kim JL, Lengauer C. The landscape of kinase fusions in cancer. Nat Commun. 2014;5:4846. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Yang W, Zhao X, Zheng A, Liu Z, Ma J, Zhang X, et al. Identification of MET fusions in solid tumors: a multicenter, large scale study in China. Int J Cancer. 2023;152(6):1259–68. [DOI] [PubMed] [Google Scholar]
- 8.Kang J, Deng QM, Feng W, Chen ZH, Su JW, Chen HJ, et al. Response and acquired resistance to MET inhibitors in de novo MET fusion-positive advanced non-small cell lung cancer. Lung Cancer. 2023;178:66–74. [DOI] [PubMed] [Google Scholar]
- 9.Riedel R, Fassunke J, Scheel AH, Scheffler M, Heydt C, Nogova L, et al. MET fusions in NSCLC: clinicopathologic features and response to MET inhibition. J Thorac Oncol. 2024;19(1):160–65. [DOI] [PubMed]
- 10.International Cancer Genome Consortium PedBrain Tumor P: Recurrent MET fusion genes represent a drug target in pediatric glioblastoma. Nat Med. 2016;22(11):1314–20. [DOI] [PubMed]
- 11.Blanc-Durand F, Alameddine R, Iafrate AJ, Tran-Thanh D, Lo YC, Blais N, et al. Tepotinib efficacy in a patient with non-small cell lung cancer with brain METastasis harboring an HLA-DRB1-MET gene fusion. Oncologist. 2020;25(11):916–20. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Hu H, Mu Q, Bao Z, Chen Y, Liu Y, Chen J, et al. Mutational landscape of secondary glioblastoma guides MET-Targeted trial in brain tumor. Cell. 2018;175(6):1665–78 e 1618. [DOI] [PubMed] [Google Scholar]
- 13.Turpin A, Descarpentries C, Gregoire V, Farchi O, Cortot AB, Jamme P. Response to Capmatinib in a MET fusion-positive cholangiocarcinoma. Oncologist. 2023;28(1):80–83. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Zhu YC, Wang WX, Song ZB, Zhang QX, Xu CW, Chen G, et al. MET-UBE2H fusion as a novel mechanism of acquired EGFR Resistance in lung adenocarcinoma. J Thorac Oncol. 2018;13(10):e202–04. [DOI] [PubMed]
- 15.Harada G, Yang SR, Cocco E, Drilon A. Rare molecular subtypes of lung cancer. Nat Rev Clin Oncol. 2023;20(4):229–49. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Liang H, Zhou D, Dai L, Zhang M, Gao Z, Mu X. A novel c-mesenchymal-epithelial transition Factor Intergenic fusion response to crizotinib in a Chinese patient with lung adenocarcinoma: a case report. Front Oncol. 2021;11:727662. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Li W, Liu Y, Li W, Chen L, Ying J. Intergenic breakpoints identified by DNA sequencing confound targetable kinase fusion detection in NSCLC. J Thorac Oncol. 2020;15(7):1223–31. [DOI] [PubMed] [Google Scholar]
- 18.Li W, Wan R, Guo L, Chang G, Jiang D, Meng L, et al. Reliability analysis of exonic-breakpoint fusions identified by DNA sequencing for predicting the efficacy of targeted therapy in non-small cell lung cancer. BMC Med. 2022;20(1):160. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Jang JS, Holicky E, Lau J, McDonough S, Mutawe M, Koster MJ, et al. Application of the 3’ mRNA-seq using unique molecular identifiers in highly degraded RNA derived from formalin-fixed, paraffin-embedded tissue. BMC Genomics. 2021;22(1):759. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Zeng Z, Fu J, Cibulskis C, Jhaveri A, Gumbs C, Das B, et al. Cross-site concordance evaluation of tumor DNA and RNA sequencing platforms for the CIMAC-CIDC network. Clin Cancer Res. 2021;27(18):5049–61. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Yang Y, Zhang Y, Zhao D, Li X, Ma T. A Novel PRKAR1A: MET fusion dramatic response to crizotinib in a patient with unresectable lung cancer. Clin Lung Cancer. 2023;24(1):e50–54. [DOI] [PubMed] [Google Scholar]
- 22.Heyer EE, Deveson IW, Wooi D, Selinger CI, Lyons RJ, Hayes VM, et al. Diagnosis of fusion genes using targeted RNA sequencing. Nat Commun. 2019;10(1):1388. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Xiang C, Guo L, Zhao R, Teng H, Wang Y, Xiong L, et al. Identification and validation of noncanonical RET fusions in non-small-cell lung cancer through DNA and RNA sequencing. J Mol Diagn. 2022;24(4):374–85. [DOI] [PubMed] [Google Scholar]
- 24.Li W, Guo L, Liu Y, Dong L, Yang L, Chen L, et al. Potential unreliability of uncommon ALK, ROS1, and RET genomic breakpoints in predicting the efficacy of targeted therapy in NSCLC. J Thorac Oncol. 2021;16(3):404–18. [DOI] [PubMed] [Google Scholar]
- 25.Yang SR, Aypar U, Rosen EY, Mata DA, Benayed R, Mullaney K, et al. A performance comparison of commonly used assays to detect RET fusions. Clin Cancer Res. 2021;27(5):1316–28. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Zhou S, Zhang F, Xu M, Zhang L, Liu Z, Yang Q, et al. Novel insights into molecular patterns of ROS1 fusions in a large Chinese NSCLC cohort: a multicenter study. Mol Oncol. 2023;17(10):2200–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Drilon A, Cappuzzo F, Ou SI, Camidge DR. Targeting MET in lung cancer: Will expectations finally Be MET? J Thorac Oncol. 2017;12(1):15–26. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Frampton GM, Ali SM, Rosenzweig M, Chmielecki J, Lu X, Bauer TM, et al. Activation of MET via diverse exon 14 splicing alterations occurs in multiple tumor types and confers clinical sensitivity to MET inhibitors. Cancer Discov. 2015;5(8):850–59. [DOI] [PubMed] [Google Scholar]
- 29.Plenker D, Bertrand M, de Langen AJ, Riedel R, Lorenz C, Scheel AH, et al. Structural alterations of MET trigger response to MET kinase inhibition in lung adenocarcinoma patients. Clin Cancer Res. 2018;24(6):1337–43. [DOI] [PubMed] [Google Scholar]
- 30.Paik PK, Felip E, Veillon R, Sakai H, Cortot AB, Garassino MC, et al. Tepotinib in non-small-cell lung cancer with MET exon 14 skipping mutations. N Engl J Med. 2020;383(10):931–43. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Wolf J, Seto T, Han JY, Reguart N, Garon EB, Groen HJM, et al. Capmatinib in MET exon 14-mutated or MET-amplified non-small-cell lung cancer. N Engl J Med. 2020;383(10):944–57. [DOI] [PubMed] [Google Scholar]
- 32.Wu YL, Guarneri V, Voon PJ, Lim BK, Yang JJ, Wislez M, et al. Tepotinib plus osimertinib in patients with EGFR-mutated non-small-cell lung cancer with MET amplification following progression on first-line osimertinib (INSIGHT 2): a multicentre, open-label, phase 2 trial. Lancet Oncol. 2024;25(8):989–1002. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data that support the findings of this study are available from the corresponding author upon request.