Aortic comparison of monozygotic twins and their nontwin siblings

Abstract

Twin registries provide data on concordance of aneurysmal disease between twin sets, but provide no granular data. Four sets of monozygotic (MZ) twin families with aortic pathology and aortic imaging were identified within a single health care system to create a Dice score for aortic comparison. MZ twin Dice scores in families A through D were 32.3%, 34%, 46.5%, and 42.4% respectively. Nontwin siblings Dice scores were 31% and 33.2% for families B and C, respectively. Our case series illustrates that MZ twin aortas have higher rates of concordance compared with their nontwin siblings.

Aortic pathology is a multifactorial process that results from modifiable risk factors, familial inheritance, and developmental defects.^1,2 Twin registries provide data on concordance of aneurysmal disease between twin sets but provide no granular data. Current twin registries do not have data regarding imaging, size of the abdominal aortic aneurysm (AAA), medical or family history, or comparison with nontwin siblings (NTSs).^3,4

We present the first known case series investigating diverse aortic pathology among monozygotic (MZ) twins and their NTSs with aortic imaging. Our aortic twin case series seeks to investigate the aortic anatomical similarities between twin sets and their NTSs using computed tomography angiography (CTA) and concordance to investigate inheritance patterns of familial aortic disease.

Methods

Study design

Our case series was a retrospective review of a prospectively maintained database from 2000 to 2024 within a single health care system. Inclusion criteria were to be a MZ twin with documented aortic pathology, the sibling of a MZ twin with known aortic pathology, or a NTS of a MZ twin with aortic pathology. If the patient was a MZ twin with aortic pathology, all twin and NTSs were contacted for a one-time CTA CAP, medical history, and genetic screening. Informed consent was obtained from all participants included in the study.

Dice confusion matrix

CTA images were used to create an aortic vascular tree that was segmented and registered using previously validated algorithms (see the Supplementary Methods). From an image analysis perspective, the Dice score is a well-known metric for the comparison of two binary images or segmentations.^5,6 The confusion matrix illustrates the Dice similarity coefficients (DSCs) with higher values indicating greater morphological similarity.

Primary end points

Our primary end point was to analyze CTA CAP with three-dimensional reconstruction among MZ twin siblings with known aortic pathology to create Dice scores and confusion matrixes to compare the anatomical characteristics among MZ twins and their NTs.

Results

Our twin case series consisted of four sets of MZ twins for a total of eight MZ twins. Two of the twin sibling sets had 1 NTS for a total of 10 patients (Table). Aortic imaging was compared from three-dimensional reconstruction (Fig 1) and Dice score calculated for anatomical comparison (Figs 2 and 3).

Table.

Families A through D aortic pathology and history

Group	Pathology	History	Family history
A	Twin 1—AAA, BL CIA aneurysms, LIIA aneurysm	CAD, HTN, HLD, 30 PY	None
	Twin 2—AAA, L CIA aneurysm	HLD
B	Twin 1—AAA	CAD, HLD, HTN, colon cancer	Both parents—sudden cardiopulmonary arrest
	Twin 2—AAA	Unknown
	NTS—None	HTN, HLD, 20 PY
C	Twin 1—AAA	HTN, HLD, 40 PY	None
	Twin 2—None	HTN, 45 PY
	NTS—None	CAD
D	Twin 1—aberrant right subclavian artery and bovine arch	HTN, HLD, 50 PY	Father—AAA
	Twin 2—Bovine Arch	HLD, 34 PY

AAA, Abdominal aortic aneurysm; BL, bilateral; CAD, coronary artery disease; CIA, common iliac artery; HLD, hyperlipidemia; HTN, hypertension; L, left; LIIA, left internal iliac artery; PY, pack-year; NTS, nontwin sibling.

Fig 1.

Three-dimensional reconstruction of aortic imaging of families A through D. NTS, nontwin sibling.

Fig 2.

All aortas were segmented from the aortic root to the superficial femoral artery (SFA)-profunda bifurcation and include major visceral branches (celiac, SMA, IMA, and, when available, renals). After registrations, Dice score metric comparisons was used to quantify the aortic similarity between the two aortas (red, sibling 1 vs green, sibling 2) in each of the sibling pairs. Guide: family A (sibling pair #1), family B (sibling pairs #2,3), family C (sibling pairs #4,5), and family D (sibling pair 6).

Fig 3.

Confusion matrix illustrating the Dice similarity coefficients (DSCs) between the 10 aortas analyzed in this study.

Family A

Family A consisted of one set of MZ twins with no relevant family history. Twin 1 was a 67-year-old black man with a past medical history (PMH) of coronary artery disease (CAD), hypertension (HTN), hyperlipidemia (HLD), and a 30 pack-year (PY) smoking history with a 48-mm infrarenal aortic aneurysm and bilateral common iliac artery aneurysms. Twin 2 was a 67-year-old black male nonsmoker with a PMH of HLD with an infrarenal aortic aneurysm measuring 33 mm and a 34-mm left common iliac artery aneurysm. Dice score similarity between these paired aortas was 32.3%.

Family B

Family B consisted of one set of MZ twins and one NTS with a family history of both parents with sudden cardiopulmonary arrest. Twin 1 was an 82-year-old White man with a PMH of CAD, HLD, HTN, colon cancer, and nonsmoker with a 54-mm infrarenal aortic aneurysm. Twin 2 was a White 82-year-old man with an unknown PMH with a 45-mm infrarenal aortic aneurysm. The Dice score metric between the two twins within this family was 34.0%. The NTS was a White 72-year-old man with a PMH of HTN, HLD, and a 20 PY smoking history with no aortic pathology. Here, the Dice score metric between twin 1 and the NTS within this family was 31.0%.

Family C

Family C consisted of one set of MZ twins and one NTS with no significant family history. Twin 1 was a 64-year-old White man with a PMH of HTN, HLD, and a 40 PY smoking history with a 46-mm fusiform infrarenal aortic aneurysm. Twin 2 was a 64-year-old White man with a PMH of HTN and a 45 PY smoking history who had a heavily calcified aorta but no aneurysmal dilation. The Dice score metric between the two twins within this family was 46.5%. The NTS was a 52-year-old White female nonsmoker with a PMH of CAD and no aortic pathology. The Dice score metric between twin 1 and the NTS within this family was 33.2%.

Family D

Family D consisted of one set of MZ twins with a family history of their father with a known infrarenal aortic aneurysm. Twin 1 was a 52-year-old White woman with a PMH of HTN, HLD, and 50 PY smoking history with an aberrant right subclavian artery and a bovine arch. Twin 2 was a 52-year-old White woman with a PMH of HLD and a 34 PY smoking history with a bovine arch. The Dice score metric between the MZ twin siblings was 42.2%.

Discussion

In this case series, we investigated the role of medical, family, and smoking history on imaging concordance among four separate sets of MZ twins and their NTSs. Currently, there are only two large studies that investigate the role of genetics in aneurysmal disease among twins: the Swedish and Danish twin registry studies.^7,8 In the Swedish registry twin study, MZ twin comparison correlated with a 71 times higher risk of a AAA compared with a MZ sibling without AAA. The Danish twin study analyzed 95 MZ twins and found more than a 2.5 times higher proband-wise concordance in MZ twins (30%) compared with dizygotic twins (12%) and an overall heritability of 77%. Although these studies help to understand the heritability of aneurysmal disease, the datasets do not contain imaging or family history and only provide a binary reference to the presence or absence of aneurysm among MZ twin siblings. The registries also do not analyze the NTs or other family members. The dataset we present is unique in that all MZ twin sets and their NTS underwent cross-sectional imaging with accompanying detailed family and medical histories, allowing direct three-dimensional comparison of aortic morphology.

To quantify morphological similarity, we used the DSC, a clinically validated metric widely used in aortic imaging to compare three-dimensional geometries across individuals, disease states, and longitudinal time points. Although the DSC does not localize focal or directional deformation, it provides a reproducible summary measure of global aortic shape concordance and is therefore well suited for interindividual and familial comparisons.

Across all subjects, Fig 3 illustrates a confusion matrix comparing DSC values for within- and between-family aortic comparisons. DSC values for unrelated individuals and related siblings were 21.2% (interquartile range [IQR], 17.7%-24.5%) and 34.0% (IQR, 33.1%-39.2%), respectively (P < .001), highlighting the strong influence of shared genetics on aortic geometry. Within the sibling cohort, MZ twins demonstrated numerically higher DSC values (34.0%; IQR, 34.0%-38.1%) compared with NTSs (32.6%; IQR, 32.2%-40.5%); however, this difference did not attain statistical significance and should be interpreted as descriptive rather than inferential. Accordingly, this observation is hypothesis generating and warrants validation in larger, statistically powered cohorts.

The morphological discordance observed among some MZ twin pairs may reflect factors beyond shared genetics; however, this study did not directly assess underlying mechanisms such as de novo mutations or unshared environmental exposures.

This study has limitations, including a small sample size, lack of a control group, and limited longitudinal follow-up to fully age-match family members. Additionally, although the DSC captures global morphological similarity, it may underestimate clinically meaningful localized differences, particularly in regions such as the infrarenal aorta or iliac bifurcation.

Conclusions

Our case series illustrates that MZ twin aortas have higher rates of concordance compared with their NTSs which persists despite varying acquired behaviors such as smoking.

Funding

None.

Disclosures

None.

From the Southern Association for Vascular Surgery