Multifocal Sporadic Venous Malformations in a Child Caused by a Novel Somatic Double Mutation of the TEK Gene: A Case Report and Literature Review

ABSTRACT

Background

Multifocal sporadic venous malformations (MSVM) are a rare and severe subtype of venous malformations (VMs). Although somatic mutations in the TEK gene are frequently implicated in VMs, the clinical and molecular significance of harboring two distinct somatic mutations within a single lesion remains poorly understood.

Methods

DNA was extracted from both the lesion and peripheral blood, followed by whole‐exome sequencing (WES) and structural modeling of the TIE2 protein.

Results

The proband, a 6‐year‐old Chinese boy, presented at birth with soft‐tissue swelling of the hands and occipital region, which progressively involved the tongue, neck, trunk, extremities, and external genitalia. Imaging demonstrated extensive venous ectasia infiltrating subcutaneous and muscular layers. Despite 20 sessions of sclerotherapy, the lesions continued to progress and showed a strong tendency for recurrence. WES of the lesion revealed two pathogenic somatic mutations in the TEK gene: c.2740C>T (p.L914F) and c.2545C>T (p.R849W), which weren't detected in peripheral blood. Structural modeling suggested no major global conformational changes but showed possible local structural differences within the kinase domain. The double‐mutant variant exhibited predicted enhanced hydrogen bonding at the mutation sites, potentially contributing to local structural rearrangements and altered conformational stability.

Conclusion

This report describes a previously unreported case of dual somatic TEK mutations in MSVM. The findings expand the mutational spectrum of TEK‐related VMs and underscore the importance of lesion‐based genetic testing in complex cases.

This study reports a rare case of multifocal sporadic venous malformations (MSVM) harboring dual somatic TEK mutations (p.L914F and p.R849W). Unlike classical mosaic‐somatic “second‐hit” patterns, these variants were identified exclusively in lesional tissue. Structural modeling suggests that these dual mutations may alter the local structure of the TIE2 kinase domain.

1. Background

Venous malformations (VMs) are the most common type of low‐flow vascular anomaly, characterized by abnormally formed, dilated venous channels that are present at birth and progressively enlarge over time (Markovic and Shortell 2021). Clinically, VMs can present as localized or multifocal lesions involving the skin, subcutaneous tissue, muscle, or internal organs, and are often associated with pain, swelling, bleeding, and functional impairment (Cooke‐Barber et al. 2020). In more extensive cases, particularly those with multifocal involvement, VMs may lead to complications such as coagulopathy and localized intravascular coagulopathy (LIC) (Budge et al. 2021).

Recent advances in genetic research have revealed that mutations in the TEK gene, which encodes the endothelial‐specific receptor tyrosine kinase TIE2, play a central role in the pathogenesis of VMs (Kangas et al. 2018). In multifocal sporadic venous malformations (MSVM), which are non‐hereditary in nature, the most frequently reported mutational pattern involves the p.R915C variant in cis with p.Y897C (Revencu et al. 2025; Soblet et al. 2017). These cases are thought to follow a “second‐hit” mechanism, in which a postzygotic mosaic mutation (e.g., p.R915C), detectable in peripheral blood, is accompanied by a subsequent somatic mutation (e.g., p.Y897C) restricted to the lesional tissue. The coexistence of these mutations may synergistically perturb TIE2 signaling, contributing to lesion development.

In this report, we describe a pediatric case of MSVM associated with a previously unreported somatic double mutation (p.L914F and p.R849W) in the TEK gene. Unlike the well‐characterized mosaic‐somatic “second‐hit” combinations, both mutations in this case were identified exclusively in lesional tissue and were not detected in peripheral blood. Notably, p.R849W, a variant typically associated with familial mucocutaneous venous malformations (VMCM) as a germline mutation, was interpreted as a lesion‐associated somatic event in this case (Wouters et al. 2010). To explore their potential pathogenic role, we conducted structural modeling to evaluate the effects of these mutations on TIE2 conformation and stability.

2. Methods

2.1. Ethical Compliance

The study was approved by the Ethics Committee of the West China Second Hospital of Sichuan University (approval no. 2024105). All procedures were conducted in strict accordance with the Declaration of Helsinki. Written informed consent was obtained from the patient's parents prior to performing whole‐exome sequencing (WES) and for the inclusion of the patient's clinical and imaging details in publications.

2.2. DNA Extraction and WES Analysis

During surgical resection of the patient's chest wall lesion, lesional skin and subcutaneous tissue were collected, along with a peripheral blood sample. Genomic DNA was extracted from both samples using the Magen Universal DNA Extraction Kit. WES was performed by CIPHER Gene Co. Ltd. (Beijing, China). Exonic regions were enriched using the IDT xGen Exome Research Panel via a multiplex probe hybridization method. Sequencing was conducted on the DNBSEQ‐T7 platform (MGI Tech Co. Ltd., Shenzhen, China) with a total data volume of approximately 30 Gb. The mean sequencing depth was 483.99 $\times$ for the lesion sample and 148.62 $\times$ for the peripheral blood sample. For both samples, the base coverage was 99.00% at ≥ 20 $\times$ depth. Raw sequencing reads in FastQ format were aligned to the human reference genome (GRCh38/hg38) using BWA. Subsequent data processing was performed using Samtools, Picard, and the GATK suite. Variant calling for single nucleotide variants (SNVs) and small insertions/deletions (InDels) was executed, and variants were annotated using Annovar. Pathogenic variants were prioritized based on the following criteria: (i) non‐synonymous mutations in exonic or splicing regions; (ii) a minor allele frequency (MAF) < 5% in public databases, including ExAC (EAS and ALL) and the 1000 Genomes Project; and (iii) functional impact assessment using multiple computational tools such as SIFT, PolyPhen‐2, LRT, MutationTaster, and FATHMM. Final pathogenicity classification and clinical correlation were performed according to the American College of Medical Genetics and Genomics (ACMG) guidelines and public databases including gnomAD, ClinVar, and OMIM. Variant nomenclature followed the standards of the Human Genome Organisation (HUGO) and the Human Genome Variation Society (HGVS).

2.3. Mutation Analysis of TEK

To explore the possible structural impact of TEK mutations, we used the amino acid sequence of the reference transcript NM_000459.5 to generate wild‐type (WT) and mutant TIE2 protein sequences. Tertiary structures were predicted using AlphaFold3 (https://alphafoldserver.com/), and the top five models were selected based on per‐residue confidence scores (predicted Local Distance Difference Test, pLDDT). To evaluate the potential effects of the mutations on protein stability, DynaMut2 (http://biosig.unimelb.edu.au/dynamut2/) was used to calculate changes in Gibbs free energy (ΔΔG). Structural alignment of the kinase domains between WT and mutant models was performed using PyMOL (v2.6.0a0), and root mean square deviation (RMSD) was calculated to assess global conformational changes. In addition, hydrogen bonding patterns surrounding the mutation sites were analyzed to explore possible local structural alterations.

3. Results

3.1. Case Presentation

The proband, a 6‐year‐old Chinese male child, presented at birth with painless, multifocal soft‐tissue swelling. Diffuse enlargement of both palms was apparent immediately after birth (Figure 1A,B). In addition, a compressible, non‐pulsatile hemispheric mass measuring approximately 5*5 cm was palpable in the occipital region, with mildly increased overlying skin temperature. Over the subsequent 6 years, serpiginous venous ectasia and localized swellings progressively emerged on the tongue, neck, post‐auricular area, thoracic and dorsal trunk, buttocks, external genitalia and all four extremities, without erythema, ulceration or exudation (Figure 1C–F). Limb edema resulted in an unsteady gait and functional limitation. Intermittent pain in the left upper limb prompted Doppler ultrasonography, which demonstrated VMs complicated by partial thrombosis; the symptoms resolved following anticoagulation with low‐molecular‐weight heparin and thrombolysis with urokinase. The patient had no history of melena, hematochezia, hematemesis, or unexplained gastrointestinal bleeding. The patient's parents denied any family history of vascular anomalies.

FIGURE 1.

Clinical and radiology manifestations in the current proband. (A–F) Clinical photographs showing extensive bluish soft‐tissue swellings involving multiple body regions, including the palm, upper limb, neck, chest wall, perineum, and lower limb. (G) MRI of the left hand shows multiple tortuous venous channels in the distal forearm, wrist, and hand, extending into the subcutaneous fat and intrinsic muscles, with a prominent nodular focus in the palmar subcutis. (H) Pelvic MRI demonstrates diffuse subcutaneous and intramuscular involvement with serpiginous vascular channels extending into the lower limbs.

Complete blood count revealed normocytic anemia, with a decreased red blood cells (RBCs) count of 3.5–3.8 × 10¹²/L and low hemoglobin (Hb) of 80–110 g/L. The platelet count was normal. The fecal occult blood test was negative. Coagulation function tests showed markedly elevated D‐dimer (32.97 mg/L) and FDPs (96.20 μg/mL), in association with hypofibrinogenemia (fibrinogen 52–139 mg/dL). Prothrombin time (PT), activated partial thromboplastin time (APTT) and thrombin time (TT) were all within normal limits.

Occipital ultrasound disclosed an ill‐defined, mildly hyperechoic subcutaneous mass with abundant intralesional flow signals. Cranio‐cervicofacial MRI disclosed patch‐like lesions with extensive serpiginous vascular structures involving the right lateral tongue, hemi‐epiglottis, parapharyngeal space, parotid and submandibular glands, auricular soft tissue, paraspinal tissue at C2 and the right occipital musculature. Left‐upper‐extremity MRI showed similar tortuous channels in the distal forearm, wrist and hand, extending into the subcutaneous fat and intrinsic muscles, with a prominent nodular focus in the palmar subcutis (Figure 1G). Pelvic MRI demonstrated numerous tortuous flow‐void vascular channels extending from the lower lumbar region to the buttocks, proximal thighs, inguinal areas and scrotum. The malformation infiltrated subcutaneous fat, underlying muscles and intermuscular spaces, with nodular venous clusters in both buttocks and the right proximal thigh (Figure 1H). On imaging, the lesions were iso‐ to hypointense on T1‐weighted images (T1WI) and markedly hyperintense on T2‐weighted images (T2WI) and fat‐suppressed T2‐weighted sequences.

Based on the clinical findings and ancillary investigations, a diagnosis of MSVM was established. The patient received fibrinogen supplementation as supportive therapy. Over a 6‐year period, the patient underwent a total of 20 sessions of endovascular sclerotherapy under general anesthesia, utilizing agents such as absolute ethanol, polidocanol, bleomycin, and lauromacrogol. Post‐sclerotherapy, antiplatelet therapy with clopidogrel was administered. While partial lesion reduction was achieved following sclerotherapy, new or progressive lesions continued to emerge.

To further elucidate the underlying etiology, the patient underwent surgical excision of truncal lesion under general anesthesia. Both lesion tissue and peripheral blood samples were submitted for WES. Histopathological examination of the resected specimen revealed diffusely dilated, thin‐walled vessels extending from the subcutaneous tissue into the muscle layer, with irregular vessel wall thickness, which was consistent with the pathological features of VMs (Figure 2A,B).

FIGURE 2.

Histopathological features of the resected lesion. (A, B) Hematoxylin and eosin (H&E) staining of the lesional tissue reveals diffusely dilated, thin‐walled venous channels extending from the subcutaneous tissue into the underlying muscle layer. The vessel walls demonstrate irregular thickness.

3.2. Sequencing With WES

WES of the lesion identified two mutations in the TEK gene: c.2740C>T (p.Leu914Phe, variant allele frequency [VAF] 14.3%) and c.2545C>T (p.Arg849Trp, VAF 1.42%), located within exons 17 and 15, respectively. Both variants map to the kinase domain of the TIE2 protein (Figure 3A,B). Given the low VAF of p.R849W, the WES alignment at this locus was further reviewed using IGV (Figure S1). Manual IGV review showed 12 alternate reads among 742 total reads at the p.R849W site, corresponding to an approximate VAF of 1.6%, basically consistent with the pipeline‐reported VAF of 1.42%. The alternate reads were evenly distributed between the forward and reverse strands (6+/6−), and no obvious strand bias or local alignment abnormalities were observed in the surrounding region. No corresponding mutations were detected in peripheral blood.

FIGURE 3.

The identified TEK mutations and their predicted structural impact on the TIE2 kinase domain. (A) Schematic diagram of the TEK gene (NM_000459.5), showing the positions of two somatic mutations: C.2545C>T and c.2740C>T, located in exons 15 and 17, respectively. (B) Domain structure of the TIE2 protein (UniProt: Q02763), with corresponding amino acid substitutions p.Arg849Trp and p.Leu914Phe mapped to the intracellular protein kinase domain. (C) Tertiary structure of wild‐type (WT) TIE2, showing the kinase domain and local hydrogen bonding interactions surrounding the Leu914 and Arg849 residues. (D) Structural comparison of the kinase domain between the p.L914F mutant (orange) and the WT (gray), showing hydrogen bond interactions around the Phe914 residue. (E) Structural comparison of the kinase domain between the p.R849W mutant (pink) and the WT (gray), showing hydrogen bond interactions around the Trp849 residue. (F) Structural comparison of the kinase domain between the double mutant (p.L914F + p.R849W, green) and the WT (gray), showing hydrogen bond interactions around both Phe914 and Trp849 residues. All structures were generated using AlphaFold3 and visualized in PyMOL (v2.6.0a0), with residues highlighted in red (Leu914; Phe914) and green (Arg849; Trp849), and hydrogen bonds represented by yellow dashed lines.

According to ACMG guidelines, both mutations were classified as pathogenic. The p.L914F variant met the criteria PS4 + PM1 + PS3_Supporting + PM2_Supporting + PP3 + PP4, and the p.R849W variant met the criteria PS4 + PP1_Strong + PS3_Supporting + PM2_Supporting + PP3 + PP4. Given that ACMG criteria were originally developed for germline variants, this classification was interpreted cautiously as supportive evidence in the context of the clinical and molecular findings.

3.3. Protein Structure Prediction

To explore the structural and stability‐related effects of TEK mutations, we used AlphaFold3 to generate tertiary structure models of the WT TIE2 protein and three variants: L914F, R849W, and the double mutant. Structural alignment of the kinase domains revealed RMSD values relative to the WT (L914F: 0.185 Å; R849W: 0.109 Å; double mutant: 0.143 Å), suggesting that none of the mutations induced significant global conformational changes (Figure 3C–F). Stability predictions from DynaMut2 indicated that both L914F and the double mutant showed slightly destabilizing effects, with ΔΔG values of −1.43 kcal/mol and −1.17 kcal/mol, respectively, whereas R849W showed minimal effects (ΔΔG = 0.27 kcal/mol). Hydrogen bond analysis showed that L914F increased bonding interactions around residue F914, while R849W similarly enhanced local hydrogen bonding near W849 (Figure 3D,E). In the double mutant, hydrogen bonding surrounding F914 was further intensified (Figure 3F). Collectively, these in silico findings suggest that although the overall fold of TIE2 is preserved, the mutations might promote subtle local rearrangements through altered hydrogen‐bond networks, thereby potentially modulating the conformation of the kinase active site. However, these observations should be interpreted as exploratory, and further functional validation is required.

4. Discussion

In this study, we reported a rare pediatric case of MSVM, characterized by widespread VMs involving the skin and subcutaneous tissues throughout the body, accompanied by coagulation abnormalities and venous thrombosis. Notably, the patient harbored a previously unreported somatic double mutation in the TEK gene: p.L914F and p.R849W. VMs represent a heterogeneous group of congenital vascular anomalies characterized by a strong genotype–phenotype correlation. Increasing evidence indicates that the clinical presentation and severity of VMs are closely related to the type and location of the causative mutations (Revencu et al. 2025). Based on current genetic insights, VMs can be broadly classified into four major subtypes. First, inherited cutaneous–mucosal VMs (VMCM) are typically associated with germline mutations in the TEK gene, most commonly p.R849W, combined with a somatic second‐hit mutation (e.g., p.Y1180*). These cases follow an autosomal dominant inheritance pattern and frequently exhibit a strong familial clustering (Limaye et al. 2009). Second, sporadic unifocal VMs, the most common subtype, are predominantly associated with somatic mutations in TEK (p.L914F) (Limaye et al. 2009; Soblet et al. 2013). These patients generally present with isolated and localized venous lesions. Third, blue rubber bleb nevus syndrome (BRBNs) is another subtype linked to TEK somatic double mutations, such as p.T1105N and p.T1106P. This syndrome is distinguished by widespread involvement of the skin, mucosa, and gastrointestinal tract (Soblet et al. 2017). Lastly, MSVM typically involve cis double mutations in the TEK gene, most notably p.Y897C and p.R915C, where the former is confined to lesional tissue, while the latter represents a low‐level mosaic mutation that can be detected in peripheral blood (Soblet et al. 2017). In the present case, the absence of gastrointestinal symptoms, negative fecal occult blood testing, and the predominance of multifocal cutaneous, subcutaneous, and muscular involvement supported the diagnosis of MSVM rather than BRBNs. We reviewed previously reported cases definitively diagnosed as MSVM and summarized their associated genetic alterations (Table 1). Including the present case, a pattern involving multiple somatic mutations and lacking detectable mosaicism in peripheral blood was observed in 5 out of 21 patients (~24%). Although the rarity of MSVM limits available case data, this observation raises the possibility that, in addition to the classically proposed mosaic–somatic “second‐hit” model, the accumulation of multiple somatic events within the lesional tissue may underlie the pathogenesis of MSVM in certain patients.

TABLE 1.

Summary of gene mutations reported in multifocal sporadic venous malformations.

Case	Year	Nationality	Gender	Age	Lowest hemoglobin (g/dL)	D‐dimer (ng/mL)	Gene	Presence in lesion				Cis/Trans	Presence in blood	VAF %	Mutation type	Ref
								Mut 1	VAF %	Mut 2	VAF %
The proband	2025	China	M	6	8.0	3297	TEK	p.L914F	14.3	p.R849W	1.42	NA	Neg		Somatic+Somatic
1	2009	Belgium	NA	NA	NA	NA	TEK	p.Y897F	22.29	p.R915L	34.90	Cis	Neg		Somatic+Somatic	Limaye et al. (2009)
2	2009	Belgium	NA	NA	NA	NA	TEK	p.Y897C	4.55	p.R915C	10.79	Cis	Neg		Somatic+Somatic
3	2016	USA	M	54	NA	NA	TEK	p.L914F	10.16	/	/	/	NA		Somatic	Castel et al. (2016)
4	2017	Belgium	NA	37	13.8	9850	TEK	p.R915C	NA	p.Y897C	NA	NA	p.R915C	4.7	Mosaic+Somatic	Soblet et al. (2017)
5	2017	Belgium	NA	30	11.7	1000	TEK	p.R915C	10.79	p.Y897C	4.55	Cis	p.R915C	1.8	Mosaic+Somatic
6	2017	Belgium	NA	28	13.2	728	TEK	p.R915C	NA	p.Y897C	NA	Cis	p.R915C	1.1	Mosaic+Somatic
7	2017	Belgium	NA	41	10.9	340	TEK	p.R915C	NA	p.L914F	NA	NA	Neg		Somatic+Somatic
8	2017	Belgium	NA	76	13.4	962	ND						NA		NA
9	2017	Belgium	NA	20	13.2	8320	TEK	p.R1099*	NA	/	/	/	Neg		Somatic
10	2018	USA	F	15	NA	NA	TEK	p.L914F	NA	/	/	/	Neg		Somatic	Goines et al. (2018)
11	2018	USA	M	9	NA	NA	TEK	p.L914F	NA	/	/	/	NA		Somatic
12	2018	USA	F	12	NA	NA	TEK	p.L915C	NA	/	/	/	Neg		Somatic
13	2022	Japan	M	12	NA	NA	TEK	p.L914F	10.3	/	/	/	NA		Somatic	Nozawa et al. (2022)
14	2022	Japan	M	54	NA	NA	TEK	p.L914F	4.2	/	/	/	NA		Somatic
15	2022	Japan	F	43	NA	NA	TEK	p.R915C	12.6	/	/	/	NA		Somatic
16	2022	Japan	F	8	NA	NA	TEK	p.R1099*	3.7	/	/	/	NA		Somatic
17	2022	Japan	F	20	NA	NA	PIK3CA	p.E545K	11.8	/	/	/	NA		Somatic
18	2022	Japan	M	5	NA	NA	PIK3CA	p.E545K	13.0	/	/	/	NA		Somatic
19	2026	Thailand	M	50	NA	NA	PIK3CA	p.E542K	3.97	/	/	/	NA		Somatic	Pakhathirathien et al. (2026)
20	2026	Thailand	M	41	NA	NA	TEK	p.Y897C	2.75	TEK	p.R1099*	8.18	NA		Somatic+Somatic

Abbreviations: Hb, hemoglobin; NA, not applicable; ND, Not Detected; VAF, variant allele frequency.

Furthermore, although the VAF of p.R849W in our patient was relatively low (1.42%), its detection was supported by high sequencing depth and additional manual review of the WES alignment. Re‐examination of this locus using IGV showed high local coverage, balanced forward and reverse strand support, and no obvious local alignment abnormality or clustered mismatches. Together with the patient's severe clinical phenotype, these findings support the plausibility of p.R849W as a low‐frequency, lesion‐associated somatic variant. This is consistent with previous reports of low‐abundance variants (1%–2% VAF) that co‐drive progression in MSVM. Nevertheless, because orthogonal validation could not be performed due to exhaustion of the lesional tissue, this finding should be interpreted with caution.

TEK gene encodes the TIE2, a receptor tyrosine kinase that is broadly expressed on the surface of endothelial cells and plays a critical role in vascular development and homeostasis (Joussen et al. 2021). Both p.L914F and p.R849W are frequently reported pathogenic mutations in VMs. Importantly, although p.R849W has mainly been described as a germline mutation associated with VMCM, in our case it was identified as a somatic mutation, thereby expanding the mutational spectrum of the TEK gene. Limaye et al. reported that overexpression of the L914F mutation in human umbilical vein endothelial cells (HUVECs) significantly enhanced TIE2 autophosphorylation and activated downstream AKT and ERK signaling pathways (Limaye et al. 2009, 2015). Building on this, Boscolo et al. transplanted HUVECs harboring the L914F mutation into the subcutaneous tissue of nude mice, successfully recapitulated the vascular ectasia phenotype characteristic of VMs (Boscolo et al. 2015), and further applied this model for pharmacological intervention studies (Li et al. 2019). In contrast, although Vikkula et al. also reported that the R849W mutation can promote TIE2 phosphorylation (Vikkula et al. 1996), subsequent research by Limaye et al. indicated that its activating effect is markedly weaker than that of L914F (Limaye et al. 2009). To date, no molecular or functional studies have investigated this dual mutation. Whether dual mutations confer a stronger activating effect on TIE2 compared to single mutations remains inconclusive. Several studies have reported that dual mutations such as Y897H‐R915C, Y897F‐R915L, Y897S‐S917I, and Y897C‐R915C result in significantly enhanced TIE2 phosphorylation relative to their corresponding single mutants (Limaye et al. 2009; Nätynki et al. 2015; Soblet et al. 2017). Nätynki et al. further demonstrated that transplantation of these dual‐mutant constructs into SCID mice led to vascular lesions with substantially larger vessel areas than those induced by single mutations (Nätynki et al. 2015). However, in other combinations such as Y897F‐R915L, T1105N‐T1106P, and T1105N‐G1115X, no additional increase in TIE2 autophosphorylation was observed compared to the single mutation (Nätynki et al. 2015; Soblet et al. 2017). In this study, structural modeling suggested that the L914F and R849W dual‐mutation configuration may be associated with increased hydrogen bonding at the mutation sites, potentially contributing to local structural rearrangements and altered conformational stability. These findings raise the possibility that the dual mutation may influence TIE2 receptor structure or activity. However, despite the structural insights provided by in silico modeling, further functional studies remain essential to confirm the effects of these dual mutations.

From a therapeutic perspective, lesion‐based genetic testing may also provide useful information for the subsequent management of patients with complex or refractory VMs. In the present case, the child continued to show lesion progression despite repeated sclerotherapy. TEK mutations are associated with abnormal activation of the PI3K–AKT–mTOR pathway, and mTOR inhibition, such as rapamycin, has been reported to improve clinical symptoms in some patients with refractory VMs (Boscolo et al. 2015; Ji et al. 2021). However, targeted therapy should be individualized based on the patient's clinical condition, risk–benefit assessment, and available evidence. In addition, the long‐term management of complex VMs usually requires multidisciplinary support from pediatricians, dermatologists, interventional radiologists, surgeons, and other care teams.

A major limitation of this study is the inability to determine the allelic phase of the two mutations, which are approximately 6 kb apart. Due to the short read lengths inherent to next‐generation sequencing (NGS) technologies, it was not possible to ascertain whether the mutations occur in cis or in trans (Hu et al. 2021). This leaves several mechanistic possibilities: (i) if the mutations occur in cis, they may act synergistically to constitutively activate the TIE2 receptor, providing a potential molecular rationale for the phenotype; (ii) if located in trans, they might interfere with receptor function through distinct structural disruptions; or (iii) if they exist in distinct subclonal populations, it would reflect the significant genetic heterogeneity of the lesion. To overcome this challenge, third‐generation sequencing platforms such as long‐read single‐molecule real‐time sequencing could be employed in future studies (Scarano et al. 2024). In addition, although the p.R849W variant was not detected in peripheral blood, the sequencing depth of the blood sample was limited. Therefore, very low‐level mosaicism below the detection threshold cannot be completely excluded.

5. Conclusion

In conclusion, we report a Chinese patient with MSVM harboring a novel somatic double mutation in the TEK gene (p.L914F and p.R849W). Unlike the previously described mosaic‐somatic “second‐hit” model, both mutations in this case were detected only in lesional tissue and were not detected in peripheral blood, supporting a lesion‐associated somatic origin. This finding raises the possibility that the accumulation of multiple somatic mutations may contribute to the development of MSVM in some cases. Structural modeling provided hypothesis‐generating observations suggesting that the co‐occurrence of these mutations may be associated with local structural differences and altered hydrogen‐bonding patterns within the TIE2 kinase domain. These findings highlight the value of lesion‐based genetic testing in patients with complex or treatment‐resistant MSVM. Further molecular and functional studies are warranted to elucidate the pathogenic mechanisms underlying the double somatic mutations in the TEK gene.

Author Contributions

L.L. and Y.C. contributed equally to this work. L.L. was responsible for writing the original draft, data curation, and visualization. Y.C. contributed to the original draft and methodology. Y.Z. participated in the preparation of the original draft. F.H. contributed to the conceptualization of the study and to the review and editing of the manuscript. All authors read and approved the final version of the manuscript for submission.

Funding

This research was supported by the China Scholarship Council (Grant No. 202406240246).

Ethics Statement

The study was approved by the Ethics Committee of the West China Second Hospital of Sichuan University (approval no. 2024105). All procedures were conducted in strict accordance with the Declaration of Helsinki.

Consent

Written informed consent to participate in this study was obtained from the patient's parents. In addition, written informed consent was provided by the patient's parents for the publication of any potentially identifiable images or data included in this article.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Figure S1: IGV screenshot of the low‐frequency TEK p.R849W variant. IGV screenshot showing WES read alignment at the TEK p.R849W locus. The variant site showed high local coverage, with 12 alternate reads among 742 total reads. The alternate reads were evenly distributed between the forward and reverse strands (6+/6−), with no obvious strand bias.

Data Availability Statement

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.