Tissue vulnerability revealed by ultrastructural skin analysis as a surrogate for arterial integrity in two cases of vascular Ehlers-Danlos syndrome

Abstract

Clinical presentation and operative outcomes in vascular Ehlers-Danlos syndrome (VEDS) are heterogeneous. We present two patients with genetically confirmed VEDS who exhibited markedly different iliac pathology phenotypes. One ruptured a common iliac artery at a near-normal diameter and the other tolerated a large aortoiliac aneurysm without rupture. The Aortic and Arterial Vulnerability Spectrum is a conceptual framework that characterizes aortic and arterial failure (dissection/rupture) along a continuum of underlying tissue vulnerability, integrating extracellular matrix (ECM) ultrastructure and clinical phenotypes beyond traditional diameter-based models. Using ultrastructural analysis of skin biopsy as a window into arterial integrity, we demonstrate distinct differences in ECM organization that paralleled the vascular phenotypes. Viewed through the Aortic and Arterial Vulnerability Spectrum, these cases show how genotype-driven vulnerability shapes surgical phenotype in VEDS and highlight ultrastructural skin analysis as a tool for ECM assessment and possibly biologic risk stratification.

Individuals with vascular Ehlers-Danlos syndrome (VEDS) exhibit striking heterogeneity in arterial complications and survival with genotype-phenotype correlations offering partial predictive value.1, 2, 3, 4, 5, 6

The Aortic and Arterial Vulnerability Spectrum (AAVS) was recently proposed as a biologically informed framework to contextualize the risk of dissection or rupture beyond anatomy by integrating extracellular matrix (ECM) ultrastructural abnormalities and clinical vascular phenotypes of fragility.¹ In this report, we present two patients with genetically confirmed VEDS and iliac artery pathology who exhibited strikingly different vascular phenotypes. Case 1 exhibited profound vascular fragility, manifested by rupture at a near-normal diameter, rapid contralateral aneurysmal progression, and extreme fragility during open surgical repair. Case 2 demonstrated a clinical pattern consistent with a relatively preserved tensile integrity, allowing for iliac artery aneurysmal degeneration without rupture and uncomplicated open aortobi-iliac reconstruction. Using ultrastructural analysis of dermal biopsies, we illustrate how variations in ECM structural elements influence rupture risk and intraoperative tissue handling. This study was approved by the institutional review board. Informed consent and separate publication consent were obtained from all patients.

Case 1

A 25-year-old man presented with spontaneous rupture of a 12-mm dissected left common iliac artery (CIA) with a large retroperitoneal hematoma 1 year after a motor vehicle crash. He underwent emergent left CIA repair using 8 × 100-mm and 10 × 100-mm GORE VIABAHN (W. L. Gore & Associates) stent grafts from the CIA to the external iliac artery (EIA) with coverage of the left internal iliac artery. Femoral access was managed with an 8F Angio-Seal device (Terumo Medical) without complications. He was discharged on postoperative day 3.

Genetic evaluation, prompted by his young age and a family history of paternal death from aortic dissection at age 49, identified a pathogenic COL3A1 splice-donor variant (c.1347+1G>A), resulting in an exon skip, confirming the VEDS diagnosis. In retrospect, the patient reported easy bruising and sleeping with his eyes partially open, and physical examination demonstrated characteristic facial features (attached earlobes, thin lips and nose), and thin translucent skin of the chest (Fig 1) and back, consistent with the diagnosis.

Fig 1.

Case 1. Thin, translucent skin of the anterior chest with visible vasculature, consistent with vascular Ehlers-Danlos syndrome (VEDS).

One year later, surveillance imaging showed a new 2.6-cm right CIA aneurysm with dissection into the femoral artery (Figs 2, A and B). Given the prior rupture at a near-normal diameter and current rapid aneurysmal degeneration, he was offered an elective transperitoneal open repair of the abdominal aorta and iliac arteries with a planned reimplantation of the right internal iliac artery to preserve pelvic blood flow. The planned operation was a bifurcated Dacron graft with felt reinforced anastomoses. Intraoperatively, severe tissue fragility of the aortic wall resulted in posterior wall tearing during the placement of the first suture of the proximal anastomosis. Thus, the technique was modified to an interrupted anastomosis, and the graft was downsized to a 12-mm Dacron graft sized 1:1 to the aortic diameter to minimize any tension on the anastomosis. The anastomosis was reinforced with circumferential felt. The right limb was anastomosed to the EIA with reimplantation of the internal iliac artery, and the left limb was anastomosed to the left CIA, incorporating the prior stent graft (Fig 2, C). A neoretroperitoneum was constructed using bovine pericardium patches (Fig 2, D). After removal of the abdominal retractor, a large proximal jejunum serosal hematoma was noted (Fig 2, E) and repaired primarily with Lembert sutures. The procedure was associated with an estimated blood loss of 2.2 L and an operative duration of 5 hours. The patient was discharged on postoperative day 6 and remains alive at 15 months without evidence of new arteriopathy (Fig 2, F). His recovery was complicated by retrograde ejaculation. At extended follow-up, he has developed two splenic artery aneurysms.

Fig 2.

Sequential iliac artery pathology and operative management in case 1, a 25-year-old patient with vascular Ehlers-Danlos syndrome (VEDS) (COL3A1 c.1347+1G>A, exon skip variant). (A) Computed tomography angiography demonstrating spontaneous rupture of the left common iliac artery (CIA) with associated retroperitoneal hematoma. (B) Surveillance imaging 1 year later revealing interval development of a right CIA aneurysm with associated dissection extending into the common femoral artery, prompting elective aortoiliac reconstruction. (C) Completed bifurcated Dacron aortobi-iliac graft incorporating the previously placed left CIA stent graft with reimplantation of the right internal iliac artery. (D) Neoretroperitoneum reconstruction using bovine pericardial patches to cover the aortobi-iliac graft. (E) Intraoperative proximal jejunal serosal hematoma related to traction. (F) Postoperative computed tomography angiography after right CIA repair demonstrating durable reconstruction without anastomotic pseudoaneurysm.

Case 2

A 44-year-old man presented with intact abdominal aortic and right CIA aneurysms measuring 5.5 × 5.2 and 7.9 × 6.8 cm, respectively, with an associated dissection extending into both CIAs (Fig 3). At the time of presentation, the patient did not exhibit overt phenotypic features suggestive of VEDS. The patient underwent an elective open transperitoneal repair using a 16 × 8-mm bifurcated Dacron graft. Distally, the right limb was anastomosed to the internal iliac artery with reimplantation of the EIA onto the graft limb. On the left, the limb was anastomosed to the EIA with reimplantation of the internal iliac artery onto the limb. The operation was completed without intraoperative complications. The patient had an estimated blood loss of 5 L (back-bleeding lumbar arteries) and an operative duration of 2 hours. The patient was discharged home uneventfully on postoperative day 8.

Fig 3.

Iliac and aortic aneurysmal disease with dissection in case 2, a 44-year-old patient with vascular Ehlers-Danlos syndrome (VEDS) (c.852+1 G>A, Ala280∗, null variant). (A) Computed tomography angiography demonstrating an intact right common iliac artery (CIA) aneurysm measuring 7.9 × 6.8 cm and an intact infrarenal abdominal aortic aneurysm measuring 5.5 × 5.2 cm, with an associated dissection extending into both common iliac arteries. (B) Postoperative computed tomography angiography 6 years post repair demonstrating durable reconstruction without anastomotic pseudoaneurysm.

Subsequent genetic testing identified a pathogenic COL3A1 splice-site variant (c.852+1G>A), predicted to result in functional haploinsufficiency (null variant) and consistent with a VEDS diagnosis. The patient was subsequently admitted 10 months later with spontaneous intraperitoneal hemorrhage without active extravasation managed nonoperatively with a 2-day hospitalization. The patient is alive approximately 6 years later without any vascular complications.

Skin and vascular tissue biopsy ultrastructure and collagen fibril morphometry

Transmission electron microscopy (TEM) was performed on full-thickness ventral forearm skin biopsies from both patients, obtained intraoperatively for case 1 and at the 6-year follow-up in case 2. Collagen fibril diameter was quantified using a validated, custom ImageJ-based macro (National Institutes of Health).²

In case 1, ultrastructural analysis revealed marked ECM disorganization, specifically, pronounced heterogeneity in collagen fibril diameter (Fig 4, A-C). Quantitative analysis of the reticular dermis, based on 4857 measured fibrils, showed a mean fibril diameter of 0.08611 ± 0.01165 μm (range, 0.041-0.133 μm). Compared with an age- and sex-matched controls without known arteriopathy (mean, 0.07482 ± 0.00691 μm), case 1's biopsy showed broader and more variable fibril diameters, suggesting disease-specific ECM disruption rather than normal interindividual variation.

Fig 4.

Collagen fibrils in the reticular dermis from two patients with vascular Ehlers-Danlos syndrome (VEDS) compared with age- and sex-matched controls. (A) Representative transmission electron micrograph of reticular dermal collagen from case 1 (COL3A1 c.1347+1G>A, exon skip variant). (B) Representative transmission electron micrograph of reticular dermal collagen from an age- and sex-matched control corresponding to case 1. (C) Histogram of collagen fibril diameter measurements demonstrating a right-shifted and broadened distribution in case 1 compared with control. (D) Representative transmission electron micrograph of reticular dermal collagen from case 2 (COL3A1 c.852+1G>A, null variant). (E) Representative transmission electron micrograph of reticular dermal collagen from an age- and sex-matched control corresponding to case 2. (F) Histogram of collagen fibril diameter measurements demonstrating an altered and heterogeneous fibril diameter distribution in case 2 compared with control. Histograms display area-equivalent collagen fibril diameters measured from true cross-sections in the reticular dermis using a standardized semiautomated Fiji/ImageJ workflow. Patient and control distributions are overlaid to facilitate direct comparison within each age- and sex-matched pair. Scale bar, 500 nm.

In contrast, case 2 displayed a narrower fibril diameter distribution and, thus, less heterogeneity (Fig 4, D-F). Quantitative analysis of 865 fibrils in the reticular dermis yielded a mean area-equivalent fibril diameter of 0.1106 ± 0.0104 μm (range, 0.0616-0.1626 μm). Compared with the age- and sex-matched control (2536 fibrils; mean, 0.0808 ± 0.0075 μm; range, 0.0424-0.1018 μm), case 2 exhibited larger but more uniform fibrils, with less heterogeneity than observed in case 1 (Fig 4, B).

Ultrastructural analysis of the right CIA in case 1 demonstrated abnormalities involving both the intimal and medial layers (Fig 5, A-D). The intima exhibited an irregular, fenestrated appearance with focal discontinuity (Fig 5, B). The tunica media showed areas of continuous elastic lamellae at low magnification (Fig 5, A); however, at higher magnification, focal discontinuity of these lamellae was observed (Fig 5, B). Selected regions of the media lacked continuous elastic lamellae (Fig 5, C), within which smooth muscle cells appeared to be round and dysmorphic, lacking their characteristic elongated orientation (Fig 5, C). Within this disrupted medial ECM, collagen fibrils exhibited pronounced size variability, with a mean diameter of 0.05565 ± 0.02301 μm (range, 0.02700-0.1345 μm) (Fig 5, D).

Fig 5.

Comparative ultrastructural features of iliac artery aneurysm and nonaneurysmal abdominal aorta in case 1, a 25-year-old patient with vascular Ehlers-Danlos syndrome (VEDS) (COL3A1 c.1347+1G>A, exon skip variant). (A-D) Iliac artery aneurysm. (A) Low-magnification view of the aneurysmal right common iliac artery (CIA) wall demonstrating severe disruption of normal vascular architecture. In a healthy artery, the tunica media consists of concentrically arranged elastic lamellae interspersed with smooth muscle cells. In the aneurysmal segment shown here, this lamellar organization is markedly disordered. The boxed region on the far left corresponds with the intimal layer, shown at higher magnification in (B), and the central boxed region corresponds with a portion of the tunica media, shown at higher magnification in (C). Widespread extracellular matrix (ECM) disorganization is present throughout the medial layer (scale bar, 10 μm). (B) Higher-magnification view of the intima, which appears fragmented (arrow) and unusually fenestrated (∗) with areas of intimal discontinuity (arrowhead). Elastic lamella displays areas of abnormal morphology (∗) (scale bar, 2 μm). (C) Selected area from the central portion of the media. Smooth muscle cells, which normally align parallel to elastic lamellae (∗), are round and dysmorphic (arrow). Fragmented and disorganized elastic fibers are present throughout this region, indicating loss of normal elastic fiber continuity (scale bar, 2 μm). (D) High-magnification view of collagen fibrils within the aneurysmal iliac artery media. This illustrates marked heterogeneity in collagen fibril diameter and packing density, rather than the relatively uniform fibril size and spacing expected in healthy arterial tissue (scale bar, 500 nm). (E-H) Abdominal aorta from the same patient. (E) Low-magnification view of the nonaneurysmal abdominal aortic wall indicating medial lamellar architecture. Concentric elastic lamellae and intervening smooth muscle cells are identifiable (scale bar, 10 μm). (F) Higher-magnification view of the tunica media showing a smooth muscle cell with an elongated orientation (∗) aligned with adjacent elastic lamella (arrow) (scale bar, 2 μm). (G) Further magnification of the boxed region in (F), demonstrating elastic fibers with focal fragmentation (arrow). Although the elastic lamellae remain recognizable and continuous, the fiber morphology is disrupted with areas of abnormal blebs and thinning (arrowheads). The basement membrane appears to be thickened (∗) (scale bar, 2 μm). (H) High-magnification image of collagen fibrils within the abdominal aorta. This highlights persistent heterogeneity in collagen fibril diameter and organization, similar to that seen in the iliac artery aneurysm, suggesting a systemic collagen ultrastructural abnormality (scale bar, 500 nm).

Similarly, ultrastructural analysis of the nonaneurysmal infrarenal abdominal aorta revealed abnormalities involving both the intimal and medial layers (Fig 5, E-H). Despite relatively preserved medial lamellar architecture at low magnification (Fig 5, E), higher magnification demonstrated focal elastic fiber fragmentation, thinning, and abnormal morphology (Fig 5, F and G). In contrast with the round, dysmorphic smooth muscle cells observed in the CIA (Fig 5, C), aortic smooth muscle cells remained elongated and aligned along morphologically abnormal elastic lamellae (Fig 5, F), but exhibited thickened basement membranes (Fig 5, G). Collagen fibrils also demonstrated substantial heterogeneity (mean, 0.09098 ± 0.02310 μm; range, 0.04063-0.1789 μm) (Fig 5, H). These findings across skin and vascular beds support a systemic collagen ultrastructural abnormality rather than a lesion-restricted process (Figs 4 and 5).

Discussion

These two cases highlight the marked biological heterogeneity in VEDS and demonstrate how differences in ECM ultrastructure translate into distinct clinical phenotypes ranging from rupture at a near-normal diameter to tolerance of large aneurysmal degeneration. They also exhibited heterogeneity in intraoperative tissue handling, ranging from extremely fragile to seemingly normal. This variability underscores the limitations of anatomy-based risk assessment alone and supports the need for biologically informed frameworks to better contextualize vascular risk and guide surgical decision-making.

Importantly, these cases reflect real-world clinical scenarios in which diagnosis is often established after an index vascular event, and operative decisions must be made in the setting of incomplete biological information. In this context, surgical planning frequently relies on clinical suspicion informed by patient age, family history, and presenting vascular pathology rather than confirmed genotype.

Four practice-relevant lessons emerge from these two cases. First, emergent endovascular repair is a viable and life-saving temporizing strategy in VEDS-related rupture. Second, genetic diagnosis provides critical biologic context that informs perioperative planning in patients with VEDS. Because syndromic features may be subtle or absent in some patients,³ genetic testing is often required to establish the diagnosis and underlying genotype. Knowledge of the specific genotype allows the surgical team to anticipate the degree of tissue fragility and likelihood of intraoperative arterial and intestinal complications. Third, successful aortic and arterial interventions in VEDS require real-time adaptation to the tissue phenotype, rather than standardized approaches alone. Fourth, ultrastructural analysis of the skin may provide a biologically meaningful window into vascular vulnerability.7, 8, 9, 10

Several technical considerations warrant emphasis. Open arterial repair in VEDS warrants meticulous tissue handling. Vessel loops have been reported to tear through the arterial wall; thus, broad-based vessel control using an 0.25-inch Penrose drains minimizes this risk. Additionally, padded clamps, tension-free anastomoses, circumferential felt reinforcement, and the selective use of interrupted sutures can decrease the risk of anastomotic failure.² These considerations reflect the need to tailor operative strategy to the underlying tissue phenotype and anticipated fragility. Visceral manipulation should be minimized; even gentle traction may result in intestinal injury. Endovascular intervention is an excellent strategy for rupture of medium-sized arteries,¹¹ with attention to minimizing iatrogenic injury to iliac arteries and access-site complications through the smallest feasible sheath sizes, avoidance of balloon dilation, and open femoral exposure with pledgeted repair when large-bore access is used.^3,12,13 Finally, the genotype provides valuable biologic context for tailoring the operative strategy, balancing open vs endovascular approaches, and anticipating intraoperative risk in this highly heterogeneous population.^2,3,5

A novel aspect of this report is the preliminary investigation of dermal ECM as a surrogate for arterial ECM integrity and as a complement to imaging and clinical phenotyping.^7,8,10,14 The contrast between cases 1 and 2 reflects a central AAVS principle: rupture of an artery at a small diameter signals greater vulnerability than an artery that can form a large-diameter aneurysm, resolving the apparent size paradox in diameter-based models. This contrast also underscores the importance of biologic context in surgical decision-making. Moreover, in the case where both dermal and arterial tissues were available, both tissues had similar ultrastructural abnormalities. These findings support the concept that ultrastructural analysis may enable comparative assessment of collagen fibril abnormalities across patients and phenotypes. Future work should evaluate correlations between skin-based quantitative ultrastructural metrics, variant class, arterial tissue phenotype on imaging and intraoperatively, and clinical outcomes across VEDS cohorts and other arteriopathies.

From an implementation perspective, skin biopsy represents a minimally invasive and accessible method for sampling ECM architecture and is in the early phases of clinical evaluation. Although TEM requires specialized facilities and technical expertise, dedicated pathology-based TEM platforms do exist and may be leveraged with appropriate training. In this context, standardized acquisition and analysis workflows, including targeted measurement of collagen fibril diameter using reproducible tools such as ImageJ-based macros, may facilitate broader application. However, current processing and analysis timelines remain incompatible with real-time decision-making in emergent clinical scenarios. Accordingly, this approach is best viewed, at present, as a proof-of-concept and research tool for characterizing tissue vulnerability and comparing ultrastructural abnormalities that may parallel arterial behavior, rather than for immediate perioperative guidance or routine clinical use. Future work will focus on standardizing TEM image acquisition and analysis protocols and validating these findings across larger, multicenter cohorts.

Conclusions

Interpreted through the AAVS framework, these cases illustrate how genotype-driven differences in ECM ultrastructure manifest as distinct arterial phenotypes and operative tissue behavior in VEDS. The observation that rupture may occur at small diameters, while large aneurysms may remain stable, highlights the limitations of anatomy-based risk models alone and underscores the importance of biologic context in surgical decision-making. Dermal ultrastructural analysis offers a potential surrogate for arterial ultrastructure and provides a framework for integrating tissue-level insights into arterial risk assessment. Future studies will focus on validating these findings across larger cohorts and integrating quantitative tissue metrics with genetic and clinical data to advance precision and personalized vascular care.

Funding

Funding was provided to Dr. Sherene Shalhub (2023-2024 Gerlinger Research Award), Dr. Apoorva Bhandari (2025-2026 McGriff Scholar Program). The funding sources had no role in study design, data analysis, manuscript preparation, or the decision to submit for publication.

Disclosures

None.