Neurovascular Findings in Children and Young Adults with Loeys-Dietz Syndromes: Informing Recommendations for Screening

Abstract

Introduction:

Loeys-Dietz Syndromes (LDS) are a group of connective tissue disorders associated with vascular abnormalities, including arterial tortuosity, aneurysms, and dissections. While neurovascular involvement is common, no pediatric or young adult recommendations for screening exist. We aimed to review our institution’s experience with special focus on neurovascular imaging to better understand the pathology and guide screening.

Methods:

A retrospective cohort study of patients with LDS was performed. Demographics, genetic subtype, clinical and radiographical data were analyzed. Primary outcome measures included pathology on neurovascular imaging, time to progression, and arterial tortuosity indexes for bilateral cervical internal carotid arteries (ICA) and vertebral arteries (VA).

Results:

Of 47 patients with LDS identified, 39 (83.0%) were found to have neuroimaging. Intracranial and cervical vascular tortuosity were seen in 79.5% and 64.1%, respectively. Twenty-one patients (44.7%) received follow-up screening, of which 3 were found to have progression. Time to progression was an average of 2.1 years. Average follow-up was 607 days (range 123–3070 days). Mean Arterial Tortuosity Index for the right ICA, left ICA, right VA, and left VA were 18, 20, 49, and 47, respectively. Comparison of interval percent change in Arterial Tortuosity Index over the course of follow-up demonstrated small changes in the right ICA (mean 5%), left ICA (mean 1%), right VA (mean 1%), and left VA (mean 2%).

Conclusions:

Arterial tortuosity was most prevalent, though it did not progress significantly over time. We suggest an algorithm for management and serial screening to guide management of pediatric and young adults with LDS.

Introduction

Loeys-Dietz syndromes (LDS) are a group of autosomal dominant connective tissue disorders with complex phenotypes including vascular manifestations and craniofacial abnormalities.^{1,2,9,10,16,19} First described by Loeys and colleagues¹⁰ in 2005, LDS is caused by genetic mutations altering transforming growth factor (TGF) beta signaling pathways. Mutations in TGFBR2, TGFBR1, TGFB3, SMAD3, and SMAD2 have been linked to various LDS phenotypes.^1,2,9,16,19

Despite heterogeneity, all LDS types have been associated with vascular pathology. Regarding cerebrovascular pathology, patients with LDS have an increased incidence of intracranial aneurysms and other neurovascular anomalies.¹⁰ Arterial tortuosity is a nearly universal finding and can be evident at as early as a few months of age.¹⁴ Tortuosity is typically most pronounced in the carotid and vertebrobasilar systems.³ Tortuosity scoring of these vessels has been correlated with risk of aortic dissection and may be an indicator of further neurovascular involvement or severity.^6,8,15 In non-LDS series, vascular tortuosity has been associated with increased risk of aneurysm development¹⁸, portending the question of whether or not aneurysm development is associated with syndromic arterial tortuosity in LDS and how best to monitor this cohort.

Adult guidelines supported by multiple societies state that “patients with Loeys-Dietz syndrome should have yearly magnetic resonance imaging from the cerebrovascular circulation to the pelvis.”⁵ However, the guidelines explicitly exclude patients younger than 18 years. There are other single-institution publications regarding LDS that recommend baseline head to pelvis magnetic resonance or computed tomography angiography (MRA/CTA) imaging at diagnosis and 1-year, and repeat imaging thereafter at least every 2–3 years, regardless of age.¹³ Currently, no formal evidence-based guidelines exist that apply to children. It is important to identify the best interval of screening to minimize unnecessary tests, including anesthetized or sedated imaging studies, or studies involving ionizing radiation, which may have implications in the developing brains and bodies of pediatric patients.⁴

The purpose of this study was to review guidelines for neurovascular screening in this population by evaluating serial cerebrovascular imaging performed in pediatric and young adult patients with LDS at our institution. We aimed to explore imaging findings at diagnosis and follow-up, to identify the best intervals of follow-up in these patients and understand how surveillance can best be applied to monitor for progression.

Methods

We performed a retrospective cohort study of patients carrying a diagnosis of any LDS subtype cared for at our tertiary pediatric center, with all sub-specialties, from 2004 through 2018. Patients were identified from the institutional Cardiovascular Genetics program database, which began in 2004. LDS diagnosis was defined by presence TGFBR1, TGFBR2, TGFB2, SMAD3, or TGFB3 mutation in conjunction with phenotypic features of LDS, including aortic dilation, tortuosity, and/or craniofacial features. Baylor College of Medicine Institutional Review Board approval was obtained with a waiver of consent.

Demographics, genetic subtype, and initial and follow-up imaging findings were collected by electronic medical record review. Neurovascular pathology was diagnosed radiographically from the official imaging report performed by a board-certified neuroradiologist for each image reviewed. Imaging of the head and neck were performed with magnetic resonance angiography (MRA) or computed tomography angiography (CTA). MRA of the neck was performed in conjunction with cardiac MRI using Philips 1.5T MRI Scanners (Ingenia, Achieva, Intera; Philips Medical Systems, Best, Netherlands) with a dynamic contrast enhanced T1-weighted fast field gradient echo sequence with bolus tracking to optimize arterial enhancement with the following parameters: TE 1.1–1.7 ms, TR 3.6–5.5 ms, 2–3 mm slice thickness, and depending on body size a field of view 30–50 cm, and matrix 200–400 × 200–400). MRA of the brain was performed with 3D time of flight technique with the following parameters: TE 6.9 ms, TR review 23 ms, 1.2 mm slice thickness, field of view 15–18 cm, matrix 256–308 × 196–292. CTA of the head and neck was performed with an Aquillon One multidetector CT scanner (Philips Medical Systems, Best, Netherlands) with 100 kV, 300 mA, and 1 mm slice thickness.

Arterial Tortuosity Index was used to quantify the excess vessel length, measured by a board-certified neuroradiologist using a Vitrea workstation (Vital Images, Minnetonka, MN). Arterial Tortuosity Index of the cervical right internal carotid artery (RICA), left internal carotid artery (LICA), right vertebral artery (RVA), and left vertebral artery (RVA) were measured using a published algorithm of [(length/straight length)-1]*100, applied to volumetric images of the calculated distance of the actual length of the vessel along the curvatures and the straight length of the vessel measured linearly.¹⁵ An example measuring Arterial Tortuosity Index is demonstrated in Figure 1. The interval percent change in Arterial Tortuosity Index over time was calculated by subtracting the prior raw score Arterial Tortuosity Index from the more recent raw score Arterial Tortuosity Index, without conversion to excess calculation, multiplied by 100%, for each imaging interval for each patient.

Figure 1. Example of Arterial Tortuosity Index Measurements in Pediatric LDS.

Figure 1 Description: MRA neck of a pediatric patient with LDS showing severe generalized tortuosity of bilateral cervical vertebral arteries. The markings on this figure demonstrate the measurement of the tortuosity index, as the arterial length divided by the shortest craniocaudal distance. In this patient, the left vertebral artery was 2.51 and right vertebral artery was 2.00.

Data were analyzed with descriptive statistics for further assessment. Patients with missing or incomplete data were excluded from analysis where this data was incomplete.

Results

Forty-seven patients diagnosed with LDS were identified at our institution; 39 (83.0%) had neurovascular screening with CTA or MRA. Those that did not undergoing screening had not yet received imaging studies, were transferring care to our institution, or had imaging that was unavailable for review in our electronic medical records. Twenty-seven (57.4%) were male and 20 (46.8%) female, with a mean age of 8.9 years at LDS diagnosis (SD 8.2, median 6.8, range 1 month – 33.9 years). The cohort was 55.6% Non-Hispanic White and 37.8% Hispanic. Genetic testing was performed in all patients with the most common mutations being TGFBR2 (42.6%) and TGFB2 (29.8%). Table 1 depicts the demographics data for the study cohort as well as findings regarding neurovascular imaging and follow-up.

Table 1.

Demographic characteristics of 47 Patients with Loeys-Dietz Syndromes

Characteristics	Total Cohort (N=47)
Age of LDS Diagnosis (yr)
Mean (SD, Median)	8.9 (8.2, 6.8)
Range	0.1–33.9
Female Sex (%)	46.8
Race/Ethnicity1 (%)
Non-Hispanic White	55.6
Hispanic	37.8
Asian	2.2
Non-Hispanic Black	4.4
Genetically confirmed LDS (%)	100
Genetic Variant of LDS (%)
TGFBR1	19.1
TGFBR2	42.6
TGFB2	29.8
TGFB3	2.1
SMAD3	6.4
Age of Neurovascular Tortuosity Diagnosis (yr)
Mean (SD, Median)	10.1 (8.5, 9)
Range	0.1–34.4
Time to Progression of Neurovascular Tortuosity (yr)*
Mean (SD, Median)	2.1 (1.6, 1.4)
Range	1–4
Received Follow-up Neurovascular Imaging (%)	44.7%
Follow-up screening time (days)
Mean (SD, Median)	607 (523.8, 385)
Range	123–3070
Neurovascular Imaging Findings
Intracranial Vascular Tortuosity (%)	79.5
Anterior Circulation Tortuosity	12
Posterior Circulation Tortuosity	16
Both Anterior and Posterior Circulation Tortuosity	72
Cervical Vascular Tortuosity (%)	64.1
Anterior Circulation Tortuosity	24
Posterior Circulation Tortuosity	8
Both Anterior and Posterior Circulation Tortuosity	68
Intracranial Aneurysms (%)	5.1
Other Vascular Lesions (%)	0
Intracranial hemorrhage (%)	0
PRES/RCVS (%)	2.6

Abbreviations: LDS= Loeys-Dietz Syndromes, TGFBR1= Transforming Growth Factor Beta Receptor 1, TGFBR2=Transforming Growth Factor Beta Receptor 2, TGFB2= Transforming Growth Factor Beta 2, TGFB3= Transforming Growth Factor Beta 3, SMAD3= gene pathway related protein, Yr= years, d= days, PRES=Posterior reversible encephalopathy syndrome, RCVS=reversible cerebral vasoconstriction syndrome.

¹Race/ethnicity was self-reported

^*Time to Progression of Neurovascular Tortuosity was measured from 3 patients who had new diagnosis of neurovascular tortuosity or progression of tortuosity based on at tending neuroradiologist report of follow-up imaging.

Numbers presented in this chart were rounded, so percentages may not total 100%.

As no standardized serial screening protocol was applied, twenty-one (44.7%) patients underwent follow-up neurovascular imaging; the number of interval screens ranged from 1–10, with a mean of 3.1 follow-up scans (SD 8.5, median 9). Mean interval scan follow-up was 607 days (SD 523.8, median 385, range 123–3070 days). Comparing follow-up with initial neurovascular imaging, 3 patients were subjectively found to have new or progressive anomalies. Of those with progression of disease, all 3 (14.3%) were found to have one or more new findings of arterial tortuosity. All 3 patients with progression of disease had a TGFBR2 mutation. Mean time to progression was 2.1 years (SD 1.6, median 1.4, range of 1–4 years).

Overall, 79.5% of patients had intracranial neurovascular tortuosity on imaging. Of those, 72% had both anterior and posterior circulation involvement, 12% had only anterior circulation involvement, and 16% had only posterior circulation involvement. In total, 64.1% of patients had cervical neurovascular tortuosity on imaging. Of those, 68% had both anterior and posterior circulation involvement, 24% had only anterior circulation involvement, and 8% only posterior circulation involvement. Of those initially screened, one patient was found to have multiple intracranial aneurysms including an 8mm right vertebral artery fusiform aneurysm, a 4mm left vertebral artery aneurysm, and a 6mm left internal carotid artery aneurysm; another patient had a small 1mm right posterior cerebral artery aneurysm, and one other patient had a 1–2mm outpouching of the left supraclinoid ICA, which represented a small aneurysm or infundibulum. The 3 patients found to have at least one intracranial aneurysm had varying genetic subtypes, including variants in TGFBR2, TGFBR1, and TGFB3. No aneurysmal progression was seen during length of follow-up. No other vascular anomalies or lesions were diagnosed in this cohort. No intracranial hemorrhage or vascular rupture was reported. No patients underwent intervention or treatment for identified neurovascular pathology. Descriptive statistics are presented in Table 1.

Of the 39 patients who underwent neurovascular imaging, 31 had adequate image quality to measure the Arterial Tortuosity Index. Those 31 patients had a total of 99 assessments, with either CTA or MRA. Each patient had a mean of 3 scans (SD 2.4, median 2, range1–10). Mean Arterial Tortuosity Index scores were 18, 20, 49, and 47 for the RICA, LICA, RVA, and LVA, respectively. During the study period, the mean percent interval change in Arterial Tortuosity Index was 5% for the RICA, 1% for the LICA, 1% for the RVA, and 2% for the LVA. (Table 2) While objective measurements demonstrate general stability over time, a subjective improvement in vascular tortuosity over time was noted by the board-certified attending neuroradiologist as the patient’s neck elongated with growth and age.

Table 2:

Bilateral Internal Carotid and Vertebral Artery Tortuosity Indexes

Characteristics	Total Cohort (N=31)
Arterial Tortuosity Index^	Mean (Range)
Right Internal Carotid Artery	18 (2–112)
Left Internal Carotid Artery	20 (3–119)
Right Vertebral Artery	49 (0–175)
Left Vertebral Artery	47 (2–218)
Interval Percent Change in Arterial Tortuosity Index*	Mean Percent (Range)
Right Internal Carotid Artery	5% (−23% – 107%)
Left Internal Carotid Artery	1% (−2% – 74%)
Right Vertebral Artery	1% (−75% – 99%)
Left Vertebral Artery	2% (−75% – 149%)

^{^}Arterial Tortuosity Index was measured using a published algorithm applied to volumetric images of the calculated distance of the actual length of the vessel along the curvatures and the straight length of the vessel measured linearly, this measures only the excess as “1” was subtracted from the ratio and multiplied by 100.¹⁵

^*The interval percent change in Arterial Tortuosity Index over time was calculated by subtracting the raw score, before being calculated as only excess, of prior Vascular Tortuosity Index from the more recent raw score Vascular Tortuosity Index for each imaging interval for each patient, multiplied by 100%.

Discussion

We report a series of 47 pediatric and young adult patients with LDS. We identify arterial tortuosity as the most common cerebrovascular abnormality observed, marking it as an important early radiographic indicator of disease, and we quantify the Arterial Tortuosity Index in this population, noting change in Arterial Tortuosity Index over follow-up.

While in non-LDS series, arterial tortuosity is often associated with progression to aneurysmal disease, our findings suggest overall stability, if not subjective improvement in radiographic appearance of vascular tortuosity over time, and no findings of progression to aneurysmal disease or dissection in this population. Our findings of grossly stable arterial tortuosity on follow-up imaging raise the consideration of what time interval is most appropriate for screening in this cohort, as frequent imaging in pediatric patients carries associated risks.

This study highlights several key considerations when approaching the care of these patients. First, due to the complexity of pathology and systemic involvement, a multidisciplinary approach is instrumental. Coordination of care by a primary cardiologist is common, and referral to consulting teams, including cardiothoracic surgery, neuroradiology, and neurosurgery are often warranted. Secondly, in pediatric and young adult patients with LDS, lifelong medical care and follow-up is required; therefore, radiological concerns such as cumulative radiation doses from serial CTs and contrast implications from vascular imaging must be considered. Lastly, due to the inability of many pediatric patients to tolerate prolonged testing, such as MRI/MRAs, anesthetics are employed to perform sedated exams, and consideration of the implications of this in development in the growing pediatric brain is important.

Management of LDS is largely derived from past experiences with other connective tissue disorders, such as Marfan syndrome and vascular Ehlers-Danlos syndrome, and is continually evolving as the natural history further declares itself. Recommendations for screening following initial diagnosis includes MRA or CTA with 3D reconstruction from head to pelvis to identify arterial aneurysms and tortuosity throughout the arterial tree.¹¹ To assess for cerebrovascular pathology, screening is indicated at the time of diagnosis. However, there is opportunity to inform guidelines for surveillance screening. The current recommendation is to tailor the frequency of MRA or CTA evaluations to clinical findings.¹¹ This ambiguity may be due to insufficient knowledge regarding the incidence or rate of progression. Due to the low number of those who progressed in our cohort, it is important to identify patients that may have a higher risk for neurovascular pathology and progression. This will help to select patients to follow more closely, and reduce unnecessary screening and the associated exposure to radiation, contrast, and anesthesia in other patients.⁴

Our findings suggest that frequent serial neurovascular screening may be limited in pediatric patients with LDS and low risk features for neurovascular involvement. Intracranial aneurysms provide an indication for follow-up screening, as they present an increased risk of morbidity and mortality due to rupture risk,^12,14,17 though no patients in our cohort experienced aneurysmal rupture or required intervention. Our study demonstrates that arterial tortuosity, though common, remained stable over time, if not subjectively improved radiographically as the patient’s neck elongated with maturity. Follow-up screening for arterial tortuosity should be considered due to the theoretically associated risk of intracranial aneurysm development.^6–8 However, we found this rarely occurred, with no patients developing new aneurysmal disease during the study period. While few patients in this study experienced progression, all had a TGFBR2 mutation, perhaps signifying a higher risk of cerebrovascular pathology in that group. Based on the aforementioned findings, we believe that LDS patients with no high-risk features, aneurysms, or arterial tortuosity on initial or subsequent follow-up screening, may not need screening as frequently as they currently receive; additionally, the intensity of surveillance may be relaxed with less frequent neuroimaging assessments.

An additional consideration in pediatric patients with LDS is the consideration among primary providers in choosing the most appropriate neuroimaging for evaluation and screening. Though ultrasound may assess extracranial vasculature, it lacks the ability to assess complete intracranial vascular and produce 3D reconstructed imaging for further evaluation. While CTA avoids unnecessary sedation, ionizing radiation can be harmful to the developing brain, reproductive system, and growing body of the pediatric patient, increasing the lifetime risk of cancer over time. However, while MRA does not contain ionizing radiation, the test is lengthy, often requiring anesthesia, holding neurocognitive implications over years of repeat imaging. These risks must be weighed on an individual basis for each patient and perhaps may change over time. Similarly, the timing of when to refer these patients for neurosurgical evaluation is another consideration primary providers may struggle with.

We thus recommend an algorithm to inform neuroimaging screenings and neurosurgical referral in LDS patients under 25 years of age, based on our experience. (Figure 2) We recommend initial screening for cerebrovascular involvement with head and neck MRA or CTA within the year of diagnosis. If intracranial aneurysms are present, we recommend follow-up screening in 6 months and referral to neurosurgery for further evaluation. If head and/or neck arterial tortuosity is present, defined as Arterial Tortuosity Index of greater than 10 to be abnormal, we recommend follow-up neurovascular screening in 2 years. If no cerebrovascular pathology is present on initial screening, we recommend follow-up neurovascular screening in 5 years. Additionally, baseline characteristics of patients should be factored when considering the frequency of follow-up screening for neurovascular pathology, tailoring the process, taking into account any high-risk features that may suggest a more severe phenotype of LDS, such as higher scores on the craniofacial index⁵ or certain genetic subtypes, such as TGFBR2, as well as cardiovascular risk factors.

Figure 2.

Cerebrovascular pathology screening recommendations for LDS patients ages 25 and under

Limitations

Limitations in this study include its small sample size and retrospective design. Due to the low number of patients who had progression of disease over time, no statistically significant conclusions can be draw from the data. Similarly, the small sample size precluded genetic subtype analysis to further stratify risk factors for progression of disease or severity of disease. The observational and retrospective nature of the study precluded comparison of best practice techniques for screening as the screening process was not standardized and therefore individual provider practice patterns not controlled for. Similarly, due to the retrospective nature of this study, no standardized serial screening protocols were applied, introducing possible selection bias as not all patients had follow-up imaging. Additionally, the recommendations in this study are based on findings in children and young adults and may not reflect the natural history of LDS in adult patients.

Conclusions

LDS is a rare genetic disorder with multisystem involvement including neurovascular abnormalities. Our findings suggest that the most common cerebrovascular pathology in patients with LDS is arterial tortuosity. However, follow-up screening after initial imaging with MRA or CTA is unlikely to reveal new or progressive pathology without adequate time, and most neurovascular pathology remains stable over several years. We suggest less frequent neurovascular imaging in the pediatric LDS population as a reasonable practice due to rarity of progression or clinically actionable findings. Further study and follow-up are warranted as well as correlation with cardiovascular disease severity to further stratify risk in this cohort.

Highlights.

Thank you for the feedback, we have found the peer review process informative and constructive and we believe your feedback has enhanced our manuscript. Revision highlights are included below:

1. We have edited the manuscript to remove redundancies and shorten where able.

2. We have created a new Figure to include the markings of how the arterial tortuosity index measurements were calculated to provide a visual representation for further clarity.

3. We have critically revised our manuscript to provide clarity over several points mentioned in the reviewer’s comments and overall believe that our intention, work, and findings are clearer.

For complete revision responses, please see the detailed response to reviewers. Additionally, our revisions are in the manuscript marked by tracked changes as well as can be viewed in whole in the unmarked version.