Vascular Ehlers Danlos Syndrome and Chromosome 2q32 Microdeletion Syndrome

Abstract

Interstitial deletions of 2q32 are typically identified after investigation for developmental delay. Two genes associated with Ehlers Danlos Syndrome (EDS); COL3A1 and COL5A2 associated with vascular EDS and classical EDS respectively, may be incorporated in the region. Although many reports of 2q32 microdeletion patients exist, there is little mention of these genes with only a few reports highlighting features potentially linked with EDS. This paper reviews the literature and presents eleven new patients with 2q32 deletions that encompass COL3A1 and COL5A2. We describe their clinical manifestations with a particular focus on the EDS phenotype. Most patients showed some minor features of vascular EDS and one patient had vessel rupture at a young age. Analysis of skin biopsy findings from two patients showed features consistent with vascular EDS but no features of classical EDS. The findings from this cohort provide additional evidence that haploinsufficiency is an important disease mechanism in COL3A1 but not COL5A2. We highlight the importance of pre-test counselling for incidental findings from broad genetic testing and appropriate post-test counselling to ensure follow up is provided to manage the implications of a vascular EDS diagnosis.

Introduction

Interstitial deletions of 2q32 were first identified by G-banding and described in 1989 by Glass, Swindlehurst, Aitken, McCrea and Boyd [1]. Since then, the phenotype of the 2q32 microdeletion syndrome has been further delineated. Features include learning difficulties, growth retardation, abnormal shape/size of teeth, high nasal bridge, thin and sparse hair, feeding difficulties, and cleft or high palate. Where included, SATB2 is considered the causative gene for most of this phenotype [2, 3]. However, multiple other genes with phenotypic implications are frequently present in the deleted region. This paper focuses on eleven new patients with 2q32 deletions encompassing COL3A1 and COL5A2, associated with vascular and classical Ehlers Danlos Syndrome, respectively. It also reviews the existing literature.

Ehlers Danlos Syndrome (EDS) is a clinically and genetically heterogeneous group of heritable connective tissue disorders. EDS is classified into fourteen types, each type is a separate and distinct condition but with overlapping clinical features [4, 5]. The 2017 diagnostic criteria clearly emphasise the importance of molecular confirmation in all EDS types where the genetic basis is known [4].

Vascular EDS (vEDS) is caused by heterozygous pathogenic variants (PVs) in the COL3A1 gene. COL3A1, located at 2q32.2, comprises 51 exons distributed over 38 kb. It encodes the procollagen peptide for type III collagen, a major structural component of skin, blood vessels, and hollow organs. The majority of COL3A1 PVs are due to missense substitutions in the Gly-X-Y repeat of the triple helical region of the type III procollagen molecule or splice site variants. The dominant negative mechanism of these PVs leads to a 7:1 ratio of abnormal to normal collagen molecules and thus only 10–15% of normal collagen is produced. The remainder, ~5%, of COL3A1 PVs result in haploinsufficiency, either by true haploinsufficiency in the form of partial and whole gene deletions or functional haploinsufficiency through the creation of a premature termination codon (PTC) [4]. The consequence of haploinsufficiency is synthesis of about half of the normal amount of type III procollagen and the phenotype of these patients is reported to be ‘less severe’ compared to other PVs [6, 7].

Vascular EDS is a rare life-threatening form of EDS, characterised by vascular complications and hollow organ fragility. The latter particularly manifests in perforation of the colon and pneumothorax. Thin translucent skin and easy bruising are frequently seen. Historically, the diagnosis was often made only after a catastrophic complication or at post-mortem examination. Establishing a correct diagnosis of vEDS can influence the individual’s prognosis and has implications for screening and medical/surgical management strategies. It has been strongly argued that molecular confirmation is particularly important in vEDS due to the challenges of making a clinical diagnosis, even for the most experienced clinicians: while some vEDS patients do demonstrate a clear phenotype of easy bruising, translucent skin, and a distinctive facial appearance, consisting of a delicate midface, thin lips, attached earlobes and prominent eyes, these findings are not present in all individuals with vEDS [4].

Classical EDS (cEDS) is caused by heterozygous PVs in COL5A1 and COL5A2. The genes COL5A1 and COL5A2 code for type V collagen, a fibrillar collagen that is widely distributed in a variety of tissues primarily in skin, bone, and tendon. The COL5A2 gene, located at 2q32.2, comprises 54 exons distributed over 147 kb and encodes the α2 chain of type V collagen (α2(V)). Most cEDS diagnoses are attributable to PVs in COL5A1 (75–78%) with haploinsufficiency the most common disease mechanism. All reported PVs in COL5A2 are structural changes suggesting that disease mechanism is likely due to a ‘dominant negative’ effect of the altered α2(V)-chains that incorporate into collagen molecules and interfere with the formation of heterotrimers [8]. Haploinsufficiency of COL5A2 has not been reported as a disease mechanism in cEDS.

Classical EDS is mainly characterised by skin hyperextensibility, atrophic scarring, and generalised joint hypermobility (GJH). The simultaneous presence of skin hyperextensibility and atrophic scarring is required for a clinical diagnosis of cEDS, together with either generalised joint hypermobility or ≥3 minor criteria. Indeed, skin hyperextensibility and atrophic scarring has been shown to be a reliable and reproducible feature of cEDS [9].

There is limited clinical detail available for patients with 2q deletions encompassing COL3A1 and COL5A2. Given that haploinsufficiency is a recognised disease mechanism for COL3A1 and COL5A1 but has not been reported in COL5A2, it is valuable to examine this further. Accordingly, we report eleven new patients and review the phenotype of 26 patients reported in the literature, specifically with a focus on clinical manifestations of EDS. The aim is to explore whether an EDS phenotype exists in these patients and discuss the implications for genetic counselling and management.

Subjects and methods

The EDS National Diagnostic Service was established for the diagnosis of the rare types of EDS. We have diagnosed over 500 patients with vEDS and 300 with cEDS. The EDS service in Sheffield received referrals for five unrelated patients with chromosomal copy number variants (CNVs) which encompass two genes associated with EDS. These patients were clinically evaluated by EDS specialist consultants for features of vascular and classical EDS. Generalised Joint Hypermobility (GJH) was measured using the Beighton score [4]. Clinical data for six patients was provided to the author from other services: the London EDS service (n = 2), regional genetics services in the UK (n = 2), and from University of Brescia, Italy (n = 2). Patients 9 and 10 are father and son.

Affected individuals had a high-resolution analysis for copy number abnormalities using Microarray-based Comparative Genomic Hybridization or single nucleotide polymorphism array (SNP), see Table 1 for details. Patient 4 also had multiplex ligation-dependent amplification using the MRC Holland kit P155-B1 for COL3A1. Depending on presenting features, some individuals had additional genetic testing. Patients 2 and 6 had a karyotype. Patient 5 had trio whole exome sequencing (WES), Patient 11 had a full R101 EDS gene panel [10], the remainder did not have the second COL3A1 allele sequenced. Table 1 outlines the genetic test results for the eleven new cases. Figure 1 is a schematic diagram of the 2q deletions of all 37 patients including our eleven new cases and 26 patients described in the literature.

Table 1.

Genotypes of the eleven new 2q microdeletion cases.

Patient	Patient 1	Patient 2	Patient 3	Patient 4	Patient 5	Patient 6	Patient 7	Patient 8	Patient 9	Patient 10	Patient 11
Deletion	arr[GRCh37] 2q32.1q33.1(186408968_201568314)x1	arr[GRCh37] 2q32.1q33.1(186534129_202840810)x1 46,XX,der(2)inv(2)(p14q32.1)del(2)(q32.1q33.1)	arr[GRCh37] 2q32.1q33.1(188405810_200275296)x1	arr[GRCh37] 2q32.1q34(185234253_214491747)x1	arr[GRCh37] 2q32.1q34(187976169_196400288)x1	arr[GRCh37] 2q32.1q32.3 (189300906_194170822) x1	arr[GRCh37] 2q31.3q32.3(182,078,611-193,402,654)x1	arr[GRCh37] 2q31.3q32.3(181,590,598-192,715,878)x1	arr[GRCh37] 2q32.1q32.3 (188,179,931_193,750,372)x1	arr[GRCh37] 2q32.1q32.3 (188,179,931_193,750,372)x1	arr[GRCh37] 2q32.1q32.2(189377509_190258812)x1
Other results		46,XX,der(2)inv(2)(p14q32.1)del(2)(q32.1q33.1)				46,XY,ins(9;2)(p24;q22q33.3) and arr[GRCh37] 17p13.3(24257_547720)x1		arr[GRCh37] Xp11.4(37,735,039-37,838,524)x3
Test method	GeneChip System 3000 platform and Affymetrix CytoScan 750K Array	Array CGH BlueGnome ISCA 60K Cytochip (v2)	Array CGH BlueGnome ISCA 60K Cytochip (v2)	Array CGH BlueGnome ISCA 60K Cytochip (v2)	Array CGH BlueGnome ISCA 60K Cytochip (v2)	Array CGH BlueGnome ISCA 60K Cytochip (v2)	Array CGH BlueGnome ISCA 60K Cytochip (v2) Normal Karyotype	Array CGH BlueGnome ISCA 60K Cytochip (v2)	Array CGH BlueGnome ISCA 60K Cytochip (v2)	Array CGH BlueGnome ISCA 60K Cytochip (v2)	SNP microarray Illumina GSA-Cyto (v1)

Fig. 1.

Schematic diagram of location of 2q deletions.

Variable diameter of collagen fibrils, a sign of disruption to the processing and organisation of collagen is a reported feature of vEDS [11, 12]. Updated nosology states that the finding of collagen flowers can support a clinical diagnosis of cEDS [13]. Therefore, transmission electron microscopy (TEM) was performed on skin biopsy in Patient 2 and Patient 6. Written consent for publication, including photographs, was obtained from all individuals.

Patient’s reported medical history pertaining to significant EDS related events was confirmed/refuted by accessing medical records.

Results

None of the patients from this cohort (n = 11) fulfilled major diagnostic criteria for cEDS. Easy bruising was present in most patients (7/11), which is also a minor criterion for vEDS. Only one patient had GJH (patient 2). All other minor criteria for cEDS were absent. See Table 2 for comparison of the clinical features against the 2017 diagnostic criteria.

Table 2.

Clinical features of the eleven new cases based on the 2017 EDS diagnostic criteria for vascular EDS and classical EDS.

One patient demonstrated a major diagnostic criterion for vEDS (aortic dissection at 15), but other patients showed several minor criteria particularly regarding characteristic facial appearance. For example, eight had thin/pinched noses, eight had lobeless/attached earlobes, eight had prominent eyes and five had thin vermillion. Other minor criteria were also prevalent, such as, thin translucent skin (7/11), acrogeria (4/11), and small joint hypermobility (7/11). Two of three patients over the age of 16, had early onset varicose veins. One patient had both talipes equinovarus and congenital hip dislocation. Figure 2 shows photographs of clinical features. A full clinical case description of each patient is included in the supplementary documents.

Fig. 2. Photographs of clinical features.

Images show some of the features seen in vascular EDS. These include facial features such as thin pinched nose, prominent deep-set eyes, thin upper vermillion. Images also show prominent vessels with thin, translucent skin and excessive bruising. Other features such as triangular-shaped face, high anterior hairline, and prominent upper front incisors are part of SATB2 facial phenotype.

Electron microscopy analysis of skin biopsy findings in Patient 2 and Patient 6 were consistent with a diagnosis of vEDS. The biopsy for Patient 2 showed variation in collagen diameter with mildly loose collagen bundle packing. The biopsy on Patient 6 showed patchy distribution of collagen fibrils with an irregular diameter, fibroblasts showed a mildly dilated endoplasmic reticulum. Ultrastructural studies showed no collagen flowers characteristic of cEDS in either patient (Fig. 3).

Fig. 3. Skin biopsy images.

Electron micrographs from patient 2 and patient 6 showing features of vascular EDS compared with a control image for 30-year-old individual.

Discussion

This section summarises the clinical features of our eleven new cases and 26 patients identified through a literature search [14–21] (see supplementary table for detailed summary). We compare these 37 cases against the literature available for vEDS and cEDS. Although there are additional patients reported in the literature and on the DECIPHER database, they were not included if the clinical descriptions were sparse, or there was limited cytogenetic reporting.

Vascular Ehlers Danlos Syndrome

There was a total of eight aortic findings: six aortic dissections (four patients) and two aortic dilations (two patients). There were three deaths due to aortic dissection at ages 34, 51 and 62. The frequency of aortic dissections, in patients over 15, was 4/18 (22.2%). Data in the literature regarding the frequency of vascular events in haploinsufficiency patients is mixed which is probably due to the small numbers reported. For example; a natural history study of 68 vEDS patients, nine of whom had haploinsufficiency variants, reported a higher prevalence of aortic disease (50% of index cases) in comparison to patients with glycine substitutions (19%) or splice-site variants (13%) [22]. This aligns with data from Frank et al. [23], who reported that 13/14 patients with haploinsufficiency variants had ≥1 vascular complication (93%). Conversely, a Dutch study of 142 patients, including 39 haploinsufficiency patients, reported no difference in the frequency of aortic aneurysms between variant type but aortic dissections were more prevalent in patients with a glycine substitution (23.0%) vs a haploinsufficiency variant (10.3%) [24]. The age of first complication for haploinsufficiency variants is reported to be around 10-15 years later than other variant groups [6, 7]. The ages at first aortic dissection in this cohort were 15, 34, 43 and 62. Significantly, our report of an aortic dissection at age 15 highlights the variability of vascular EDS phenotypes, even for variants considered ‘less severe’.

In addition to major vascular complications, there were reports of bleeding tendencies and superficial venous problems. Six patients had early onset varicose veins and five patients reported bleeding problems, from gastrostomy, after minor injury and tooth extraction.

It has previously been suggested that major complications in haploinsufficiency patients may be limited to vascular events [6]. Indeed, there were no gastrointestinal (GI) manifestations reported in this 2q cohort, concordant with a French study of 131 vEDS patients [25]. However, non-vascular manifestations have subsequently been described [24, 26] and one bladder rupture was reported in the 2q literature [18]. This suggests that major complications in haploinsufficiency patients are not limited to vascular events. Of note, these non-vascular complications occurred at later ages than other variants.

Considering that the major criteria for vEDS focus on events that mostly occur in adulthood, it is not surprising that most of this cohort did not fulfil the criteria (only ten patients were over 30 and 17 were 10 or under). Major complications are uncommon in childhood, it is more likely that children will present with easy bruising and translucent skin with increased venous visibility [27]. Our study was consistent with this; nineteen patients had thin translucent skin and vessel visibility (19/27 70.4%) and thirteen report easy bruising (13/24 54.2%). A vEDS facial phenotype which includes prominent eyes (due to lack of subcutaneous adipose tissue around the eyes), a thin, pinched nose and thin lips, and lobeless ears [22, 23, 28, 29], was present in a substantial number of the 37 patients. This is concordant with findings from 180 vEDS patients in the Sheffield EDS service, see Table 3 [26]. Small joint hypermobility was specifically highlighted in 12 patients (12/24 50%). In vEDS, hypermobility is usually limited to the small joints with studies describing frequencies of 19–94% with most reporting that around half of their patients had small joint hypermobility [23, 26, 29, 30].

Table 3.

Comparison of phenotype in 2q cases against patients with vascular EDS reported by Bowen et al. [26].

Featuresa	2q cohort in the literature (n = 26)b	New patients (n = 11)	(Bowen et al. [26])
Prominent/deep set eyes	6/26 23.1%	8/11 72.7%	94.6%
Thin translucent skin	12/26 46.2%	7/11 63.6%	74%
Lobeless/attached ear lobes	5/26 19.2%	8/11 72.7%	85%
Easy Bruising	6/26 23.1%	7/11 63.6%	83%
Hair (sparse and thin)	4/26 15.4%	4/11 36.4%	29.7%
Small joint hypermobility	5/26 19.2%	7/11 63.6%	64.1%
Acrogeria	4/26 15.4%	4/11 36.4%	41.8%
Early onset varicose veins	4/26 15.4%	2/11 18.2%	15%
No patients who have experienced vascular Events	5/26 19.2%	1/11 9%	40%
Generalised joint hypermobility	2/26 7.7%	1/11 9%	32.8%
Contractures	2/26 7.7%	2/11 18.2%	3.8%
Keratoconus	1/26 3.8%	1/11 9%	1.11%

^aTable ordered by frequency of features identified in the total cohort of 37 patients.

^bEDS features were often not reported on in the previous 2q literature. This suggests that EDS features may have been underreported because the focus of these papers was not the EDS phenotype.

Congenital talipes have a frequency of ~10% in vEDS patients [7, 23, 26] and affect three patients here (7.8%). However, in one case there was a paternal family history of talipes whereas the vEDS history was maternal [21]. Two patients had keratoconus, which has also been reported in vEDS patients [31]. Four patients had contractures, which although not currently considered a key feature of vEDS, were found in 3.8% of the Sheffield cohort [26]. Acrogeria was described in eight patients (8/22 36.4%). Thin/sparse hair was reported in eight patients (8/17 47.1%). To our knowledge, there are two mentions of alopecia prevalence in vEDS retrospective studies which cite figures of 18% [28] and 41% [32]. However, it is our experience that although common in the Sheffield vEDS cohort (29.7%.), thin, sparse hair is usually not observed until adulthood in vEDS patients [26]. It is therefore unclear if ‘thin hair’ is due to haploinsufficiency of COL3A1. Of note, thick coarse hair was reported in four patients. Papers which report on smaller deletions often do not specify hair features. Some evidence that features such as thin upper lip, deep set eyes, talipes, thin and sparse hair, translucent skin and joint hypermobility may be due COL3A1 comes from Zarate and Fish (2017) [3], who reviewed 41 patients with SATB2 alterations. They compared large cytogenic deletions (encompassing COL3A1 and COL5A2) with intragenic SATB2 variants. They found that these features were often or exclusively found in patients with large cytogenic deletions.

Classical Ehlers Danlos Syndrome

The diagnostic criteria for cEDS were not fulfilled by any of the patients reviewed here. Hyperextensible skin was reported in four patients, but significant atrophic scarring was not seen. There were also no descriptions of skin fragility. GJH was present in three patients (3/24 12.5%), whereas Ritelli et al. [9], report GJH in 58.7% of cEDS patients.

Except for soft skin, other features consistent with cEDS also overlap with other conditions with genes deleted in the region, e.g., vEDS and SATB2-associated syndrome (SAS). Strabismus, which was present in six patients here (five of whom have SATB2 included in their deletion) has been reported in cEDS [9] and is also seen in patients with SATB2 point mutations, gene disruptions and intragenic deletions [3]. Similarly, hypotonia was found in seven patients; six of whom had deletions encompassing SATB2. Hypotonia is reported in 59% of patients with intragenic SATB2 variants. It is therefore possible that strabismus and hypotonia are due to SAS phenotype [3, 33].

The authors were unable to find reports of cEDS patients with haploinsufficiency for COL5A2. The absence of major clinical features in this cohort suggest that this kind of variant does not result in a clinical phenotype. There is some animal evidence that COL5A2 haploinsufficiency could result in a mild phenotype. Postnatal COL5A2^+/− heterozygotes mice had relatively mild abnormalities in collagen fibril morphology with adults showing abnormally contoured dermal collagen fibrils on skin biopsy but without collagen flowers, congruous with biopsy findings from Patients 2 and 6 in this cohort. COL5A2^+/− skin had increased elasticity and decreased tensile strength. The paper’s authors suggest that human individuals with a COL5A2^+/− genotype are likely to be more susceptible to skin tearing and wounding than normal, even though they might not be diagnosed as having frank cEDS [34]. This did not appear to be the case in our cohort, but mild skin features may not have been reported in the wider 2q literature.

Implications for practice

Clinicians should be aware that the COL3A1 gene may be included in the deleted region of patients with 2q deletions, and care should be taken when explaining the implications of this to patients. Importantly, plans should be made for ongoing follow up which includes lifestyle advice, regular surveillance of the arterial tree, and antihypertensive medication from the teenage years, as discussed in Bowen et al. [26]. Clinician awareness of tissue fragility and bleeding tendency is particularly important in this cohort of patients who may have increased medical needs due to other genes in their deleted region. The children seen in our service were regularly admitted for medical treatment and investigation and there were reports of problems from minor interventions (e.g. cannulation resulted in an extravasation injury in patient 2). It is recommended that additional care is taken in consideration of their tissue fragility [23].

Future research

This is a small patient cohort and studies with additional numbers would provide stronger insights. It is likely that patients already reported in the literature fulfil some additional minor vEDS diagnostic criteria, but authors have not focused on these features or considered them notable findings. In addition, patients who have been reported are mostly younger than the age of significant events in vEDS, it would therefore be beneficial to review these patients over time. Patients with 2q deletions often require additional and complex health care support, and studies observing how families, patients, and clinicians managed these challenges may help inform management advice and psychosocial support for this group.

To summarise, pathogenic variants causing haploinsufficiency of the COL3A1 gene are well documented in the literature and our findings are concordant with reports of other patients with haploinsufficiency. We therefore conclude that diagnoses of vEDS should be made. There is not sufficient evidence from phenotypic features and EM results in this cohort that haploinsufficiency is a disease mechanism in COL5A2. Follow-up with greater numbers may help to further characterise natural history for these patients. Although findings may be complicated by other genes involved in their deleted region, longitudinal analysis to observe events related to vascular EDS may help to further clarify the phenotype of vEDS due to haploinsufficiency.

Author contributions

CG conceived and designed the work, acquired, and reviewed medical records and drafted the manuscript. DD and AM were involved in the conception and design of the work and contributed to the drafting of the first manuscripts. SA produced Fig. 1, DB, and SM provided genotype results for their patients and ensured correct nomenclature. GS, DD, AM, NG, FVD, MC, MR, EH, JR, EK, AD, and FS provided patient data. FVD, AM, and TK proofread and edited the manuscript, and GS, DSJ, JB, MC and MR provided considerable and detailed feedback throughout the project.

Funding

No financial assistance was received in support of this study.

Data availability

Data generated during this study can be found in the published article and supplementary files. It is not available in a repository.

Competing interests

The authors declare no competing interests.

Ethical approval

No formal research ethics approval or research and development approval was required as stipulated by the UK Policy Framework for Health and Social Care Research and the Health Research Authority decision tool. Written consent for publication, including photographs, was obtained from all individuals.

Supplementary information

The online version contains supplementary material available at 10.1038/s41431-025-01849-2.