Article Text

Download PDFPDF

Original research article
Prevalence and predictors of intracranial aneurysms in patients with bicuspid aortic valve
  1. Alexander C Egbe1,
  2. Ratnasari Padang1,
  3. Robert D Brown2,
  4. Arooj R Khan1,
  5. Sushil A Luis1,
  6. John Huston III3,
  7. Emmanuel Akintoye4,
  8. Heidi M Connolly1
  1. 1 Department of Cardiovascular Diseases, Mayo Clinic, Rochester, Minnesota, USA
  2. 2 Department of Neurology, Mayo Clinic, Rochester, Minnesota, USA
  3. 3 Department of Radiology, Mayo Clinic, Rochester, Minnesota, USA
  4. 4 Department of Internal Medicine, Wayne State University/Detroit Medical Center, Detroit, Michigan, UK
  1. Correspondence to Dr Alexander C Egbe, Department of Cardiovascular Diseases, Mayo Clinic, 200 First Street SW, Rochester, MN 55905, USA; cegbe2001{at}yahoo.com

Abstract

Objective To determine the prevalence and outcomes of intracranial aneurysm (IA) in patients with bicuspid aortic valve (BAV).

Methods Retrospective review of patients with BAV who underwent brain MR angiography at the Mayo Clinic from 1994 to 2013.

Results There were 678 patients included in this study—mean age 57±13 years, men 480 (71%), mean follow-up 10±3 years (5913 patient-years). Coarctation of aorta (COA) was present in 154 (23%) patients.

There were 59 IAs identified in 52 of 678 patients (7.7%). IA was present in 20/154 patients (12.9%) with COA and 32/524 patients (5.7%) without COA (p<0.001). For the patients without COA, female gender and right–left cusp fusion were risks factors for IA in women after adjustment for all potential variables (HR 1.76, CI 1.31 to 2.68, p=0.03). There was no significant trend in the risk for IA across age tertiles: age ≤40 years versus 41–60 years (HR 1.19, p=0.34), and age 41–60 years versus 61–80 years (HR 1.06, p=0.56).

Among the 52 patients with IA, enlargement occurred in three patients (6%), rupture in two patients (4%) and four patients (8%) underwent coil embolisation. For the 626 patients without IA at baseline, no patient developed IA over 7±2 years of imaging follow-up.

Conclusions BAV is associated with a higher prevalence of IA compared to the general population, and this risk is higher in patients with COA, right–left cusp fusion and female gender.

  • Bicuspid aortic valve
  • Coarctation of aorta
  • Intracranial aneurysm
  • Magnetic resonance angiography

Statistics from Altmetric.com

Introduction

Bicuspid aortic valves (BAV) affect 1.5% of the general population and are the most common cause of aortic stenosis in patients younger than 60 years of age.1–3 The patients with BAV may have associated aortic dilation and coarctation of aorta (COA).4–6

Prior studies have shown that patients with COA with or without BAV have a fivefold risk of intracranial aneurysm (IA) compared with the general population.7 However, there are limited data about the occurrence of IA in patients with BAV.8 The purpose of this study was to determine the prevalence, risk factors for occurrence and outcomes of IAs in patients with BAV.

Methods

Patient selection

This is a retrospective study of adult patients (age >18 years) with a diagnosis of BAV who underwent brain MR angiography (MRA) at Mayo Clinic from January 1994 to December 2013. The Mayo Clinic Institutional Review Board approved the study protocol.

The inclusion criteria for the study were BAV diagnosis by echocardiography, at least one brain MRA and at least 2 years of clinical follow-up. The patients with concomitant diagnoses of connective tissue diseases and other genetic aortopathies were excluded. The inclusion criterion was restricted to only brain imaging with MRA because MRA was the imaging modality of choice for IA at our institution. The echocardiographic images were viewed in all patients to determine cusp morphology. Pathology reports were viewed for the patients that underwent aortic valve replacement at the Mayo Clinic.

Data collection and study design

Clinical, surgical, echocardiographic and radiological data were collected from the comprehensive medical record. The definitions for clinical variables in this study were similar to those used in prior studies.7 9

The primary objective was to determine the prevalence of and risk factors for IA in patients with BAV with and without COA. The secondary objective was to determine the outcomes of IA in this population.

Magnetic resonance imaging

All MRA examinations were performed at our institution and the study protocol has been described in a prior study.7 Of all the MRAs included in this study, 391 (58%) were performed on a 3.0T MRI unit with a three-dimensional time-of-flight sequence while 287 (42%) MRAs were performed on 1.5T scanners. The parameters included 43/6.9, time to repetition/time to excitation; 25° flip angle, 256×224 matrix; 18 cm field of view with 0.9-phase field of view; one excitation; three slabs of 32 1.4 mm thick sections per slab acquired with six overlapped sections, inferior to superior ramped tilted optimised non-saturating excitation pulse, flow compensation, magnetisation transfer and zero filling in the frequency and phase directions.

Postprocessing maximum intensity projection images of the right and left internal carotid arteries and posterior circulation subvolumes were created to eliminate the problem of overlapping vessels. The three subvolumes were viewed in a cine loop on an independent monitor to enhance the perception of the three-dimensional relationship of the vessels and facilitate the identification of small IAs. Standard MRI of the head was also performed with sagittal T1-weighted and transaxial T2-weighted sequences.

Statistical analysis

All statistical analyses were performed using JMP V.10.0 software (SAS Institute, Cary, USA). Categorical variables were compared using the χ2 test; continuous variables were compared with a two-sided unpaired Student’s t-test or Wilcoxon rank-sum test, as appropriate. Cox proportional hazards models were used to identify risk factors for IA. Schoenfeld residual method was used for testing the proportional hazard assumption. Incremental model building was performed to assess the impact of cusp fusion and gender on the risk for IA. Model 1 was created to adjust for the pattern of cusp fusion. Model 2 included all variables in model 1 plus adjustments for female gender and family history of IA or intracranial haemorrhage. Model 3 included all variables in model 2 plus adjustment for current/prior smoking. Model 4 included all variables in model 3 plus adjustment for hypertension, aortic aneurysm, diabetes and use of oestrogen oral contraceptive pills. Exploratory models for right non-coronary cusp fusion and left non-coronary cusp fusion were also performed, but did not substantively alter the results. The risk associated with each variable was expressed as HR and 95% CI. All p values were two sided, and p values <0.05 were considered significant.

Results

Patient characteristics

A review of the medical records showed that there were 4861 patients with BAV who had an echocardiogram within the study period. Among these patients, 729 of them also underwent brain MRA during the study period; 13 were excluded because of concomitant diagnoses of connective tissue diseases or other genetic aortopathies; and 38 patients were excluded because of indeterminate cusp morphology. A total of 678 patients met the inclusion criteria for this study; mean age was 57±13 years and 480 (71%) were men. The age at the time of first brain MRA scan was 48±11 years, and the duration of clinical follow-up from the time of first brain MRA scan was 10±3 years (5913 patient-years).

There were 66 deaths (all-cause mortality) during follow-up, and the causes of death were perioperative after cardiac surgery (n=11), myocardial infarction (n=10), heart failure (n=7), endocarditis (n=3), sudden (n=4), mechanical aortic prosthetic thrombosis (n=1), malignancy (n=15), sepsis (n=3), renal failure (n=3) and unknown cause (n=9). None of these deaths were directly related to IA. There was no difference in all-cause mortality between the patients with IA versus patients without IA, 10% versus 9%, p=0.18.

BAV cusp morphology was right and non-coronary cusp fusion in 212 (31%), right–left cusp fusion in 428 (63%) patients and left and non-coronary cusp fusion in 38 (6%). COA was present in 154 patients (23%; table 1).

Table 1

Baseline characteristics of entire cohort

The indication for brain MRA was primarily because of a history of BAV in 478 patients (70%) and neurological symptoms such as headache, weakness, vision loss, gait disturbance, paresthesia and seizure in 175 patients (26%) (table 1).

Prevalence of intracranial aneurysm

IA was presents in 52 of 678 patients (7.7%) in this study. Table 1 shows a comparison of the baseline clinical and echocardiographic characteristics of the patients with and without IAs. The patients with IA were more likely to have COA and right–left cusp fusion, and more likely to have an aortic dimension >40 mm.

Among the 52 patients with IA, 20 patients (38%) had COA. Table 2 shows a comparison of the baseline characteristics of the patients with IA who had COA compared with those without COA. The patients with COA were younger at the time of IA diagnosis, more likely to be males, and more likely to have hypertension.

Table 2

Baseline characteristics of patients with IA (n=52) 

The prevalence of IA in the entire cohort (n=678) was 7.7% (6.4–8.5); the prevalence of IA in the subset of patients with COA (n=154) was 12.9% (11.5–13.8); and the prevalence of IA in the subset of patients without COA (n=524) was 5.7% (5.1–6.9). IA was more common in patients with COA compared with those without COA, p<0.001.

Location and outcomes of intracranial aneurysms

There were 59 IAs identified in 52 patients, and 7 of these 52 patients (12%) had two IAs. figure 1 shows the location and size of IAs. The median size of IA was 4 mm (3-6), and 48 (81%) IAs were located in the anterior circulation.

Figure 1

Size of intracranial aneurysm by location. Each black dot represents aneurysm size per location, and the white-coloured number within each dot represents the number of patients with the corresponding size of aneurysm at that particular arterial location.

Aneurysm rupture occurred in 2 of 52 patients (4%) during 11±4 years of clinical follow-up. One patient was a 41-year-old man with COA who had a 6 mm right posterior cerebral artery IA detected on MRA performed 2 years prior to IA rupture. The second patient was a woman without COA who had a 7 mm left anterior choroidal artery IA detected on MRA at the age 46 years. She was lost to follow-up and presented with subarachnoid haemorrhage at the age of 53. The frequency of IA rupture was 4% within 7 years.

Four patients (8%) underwent coil embolisation of IA. One patient was a 48-year-old man with COA who underwent coil embolisation of an 8 mm right middle cerebral artery IA. The second patient was a 55-year-old man with COA who underwent coil embolisation of a 10 mm left anterior cerebral artery IA. The third patient was a 69-year-old woman without COA who underwent coil embolisation of a 7 mm right ophthalmic artery IA. The fourth patient was a 64-year-old man without COA who underwent coil embolisation of a 10 mm right anterior choroidal artery IA.

A 78-year-old with a 9 mm left anterior cerebral artery IA underwent neurosurgical evaluation but conservative management was recommended because of patient’s age and comorbidities. He died 2 years later from a malignancy.

Risk factors for intracranial aneurysm

Schoenfeld residual method was used for testing the proportional hazard assumption. COA was the only multivariable risk factor for IA in the cohort of 678 patients (HR 3.21, 95% CI 2.12 to 5.13, p<0.001; figure 2). However, in the subset of patients without COA (n=524), the presence of right–left cusp fusion was a risk factor for IA (table 3, model 1). Although this risk was attenuated after adjusting for female gender, right–left cusp fusion remained a significant risk factor for IA, HR 1.76, CI 1.33 to 2.69 (table 3, model 2).The inclusion of smoking history (table 3, model 3) and comorbidities (table 3, model 4) did not alter these findings.

Figure 2

Predictors of intracranial aneurysm (IA). DBP, diastolic blood pressure; IIH, intracranial haemorrhage; OCP, contraceptive pill/oral contraceptives; SBP, systolic blood pressure.

Table 3

Incrementally adjusted risk for IA in patients without coarctation of aorta (n=524) 

We then divided the patients without COA into three age tertiles: ≤40 years, 41–60 years and 61–80 years. The presence of right–left cusp fusion and female gender was associated with increased risk of IA for patients <40 years of age (HR 1.68, CI 1.31 to 2.62), 40–60 years of age (HR 1.76, CI 1.33 to 2.69) and 61–80 years of age (HR 1.77, CI 1.34 to 2.69). There was no significant trend in the risk for IA across age tertiles (HR 1.19, p=0.34 and HR 1.06, p=0.56, respectively). After adjusting for demographics, cardiovascular risk factors and comorbidities, there was no significant difference by age in the association of female gender and cusp fusion with development of IA (p interaction=0.12).

Screening and surveillance

In addition to a baseline brain MRA, which was an inclusion criterion for this study, 428 patients (63%) had additional MRAs. All 52 patients with IA had a second brain scan and 29 (56%) of them had a third brain scan. The mean interval between the baseline MRA and first repeat MRA was 3±1 years and the mean imaging follow-up was 9±4 years. IA enlargement ≥1 mm occurred in three patients (6%), and one patient (2%) developed a second IA. The cases of IA enlargement occurred in the following patients: a 49-year-old man with 6 mm IA that enlarged to 8 mm 3 years after the initial MRA; a 38-year-old man with 4 mm IA that enlarged to 7 mm after 4 years; and a 53-year-old woman with 5 mm that enlarged to 6 mm after 2 years.

For the 626 patients without IA at baseline, 274 patients (44%) underwent a second MRA and 73 (12%) had a third scan. The mean interval between the baseline MRA and first repeat MRA was 4±1 years and the mean imaging follow-up was 7±2 years. No patient with an initial negative scan developed IA on subsequent scan.

Discussion

This is a review of 678 patients with BAV who underwent brain MRA at the Mayo Clinic, and the major findings are: (1) IA was present in 7.7% of all patients with BAV, in 5.7% of the subgroup of patients without COA and in 12.9% of those with COA; (2) In the absence of COA, the risk factors for IA were female gender and the presence of right–left cusp fusion.

Prevalence of IA

There were 524 patients is this study who had BAV without COA and the prevalence of IA was 5.7% in this subset. A recent study identified IA in 10% of 61 patients with BAV based on brain MRA and CT angiography.8 The demographic and clinical characteristics of that cohort were comparable to the current study.

The presence of COA is associated with the significant increase in the risk for IA.7 This association is based on a prospective study that showed a 10% prevalence of IA in 100 patients who underwent brain MRA. There were 154 patients with COA in the current study, and IA was identified in 20 of them yielding a prevalence of 12.9%. This result is concordant with the 10% prevalence of IA in patients with COA reported in the prior prospective study from our institution.7 The message from the current study is that BAV is associated with an increased risk of IA compared with the general population,10–15 and this risk is higher if there is concomitant COA.

About 70% of the cohort underwent brain MRA primarily because of a history of BAV. Although there is no institutional policy for IA screening at the Mayo Clinic, most of the physicians have a low threshold for IA screening in this population because of our previous data that showed a high prevalence of IA in patients with COA.

Risk factors for IA

Among the 524 patients without COA, the multivariable risk factors for IA were female gender and the presence of right–left cusp fusion. There were 53 patients with both risk factors (female and right–left cusp fusion) and IA was present in six of them (11.3%), a prevalence that was comparable to 12.9% in the patients with COA.

The pathophysiological mechanism underlying the higher prevalence of IA in women is not apparent from this study. However, we suspect that this is similar to the observation in the general population where the female gender is also a known risk factor for IA10 11 13 and in patients with BAV.8 This is the first study correlating the occurrence of IA with the pattern of cusp fusion. A combination of female gender and right–left cusp fusion identifies a high  group, whose risk of IA was comparable to that of patients with COA. It is important to note that right-left cusp fusion was more common in patients with COA compared with the rest of the cohort. Perhaps the higher prevalence of IAs in patients with COA may, to some extent, be related to their cusp morphology.

Aneurysm enlargement

Aneurysm enlargement occurred in 3 of 52 patients (6%), and one patient (2%) developed a second IA. A prospective study of 468 unruptured IAs showed that 57 patients (12%) had aneurysm enlargement within 2 years of follow-up.16 The risk factors for aneurysm enlargement identified in that study were larger aneurysm size at baseline, dome:neck ratio, multilobarity and aneurysm location in the posterior circulation. A retrospective study of 191 IAs in 165 patients showed that 10% of the IAs grew in size during a mean follow-up of 4 years.17 Aneurysm size at baseline was a predictor of subsequent enlargement.

Even with a somewhat more liberal definition of aneurysm growth in the current cohort, the frequency of aneurysm enlargement of 6% over 9 years in our cohort may be somewhat lower than those reported in the cohorts noted above.16 17 Possible explanations for the difference in enlargement would include overall differences in cohort characteristics in terms of the risk factors for aneurysm growth. However, the natural history of IA in BAV may be different from that of the general population. We were unable to assess for predictors of aneurysm growth in the current cohort because of small number of patients with growth.

Aneurysm rupture and need for intervention

Aneurysm rupture occurred in two patients (4%) while another four patients (8%) underwent coil embolisation of IA. Data from a pooled analysis of six prospective cohort studies showed a 3.4% risk of aneurysm rupture in 5 years.18 In a different study, the investigators from the International Study of Unruptured Aneurysms reported a 3% risk of aneurysm rupture in 5 years based on follow-up of 2686 IAs in 1692 patients.19 The comparison of IA rupture risks between cohorts can be a challenge because of differences in IA characteristics, including those which may predict an increased risk of rupture. However, the rupture risk in the current cohort, 4% IA rupture over a mean of 7 years of follow-up, appears to be similar to those reported in large IA patient cohorts. The two patients that had IA rupture in our cohort had moderate-sized lesions (6 and 7 mm) at baseline. Perhaps, the patients with moderate-sized lesions (>5 mm) should have more frequent monitoring and a low threshold for intervention in order to prevent IA rupture.

Screening and surveillance

This study shows that BAV in the absence of COA was associated with an approximately threefold increase in risk of IA detection compared with the general population with IA prevalence of 1%–2%.10 14 20 Furthermore, women with right–left cusp fusion had an approximately sixfold risk of IA, a risk that was comparable to that of patients with COA.

Currently, there are no guideline recommendations for IA screening in patients with BAV, or for monitoring aneurysm enlargement after a diagnosis of IA is made in this population. While additional data, optimally from a large prospective cohort, is necessary to validate the current findings, the following could be considered based on the data from the current cohort: (1) All patients with BAV should consider having a baseline brain MRA or CT angiography to screen for IA. If a diagnosis of IA is made, monitoring for aneurysm enlargement should follow the general recommendation and expert consensus opinion for monitoring IA in the general population.21 22 (2) Serial imaging may not be necessary in patients with normal baseline MRA except for those patients with BAV who may have an overall higher risk of IA occurrence. Higher-risk patients include those with COA or female patients with right–left cusp fusion.

Limitations

This is a retrospective review of patients followed at a single tertiary centre. The authors were unable to accurately determine the indication for MRA in some of the cases because of the retrospective nature of the study. Although there is no institutional policy for IA screening at the Mayo Clinic, most of the physicians have a low threshold for IA screening in this population because of our previous data that showed a high prevalence of IA in patients with COA. The prevalence of IA may have been overestimated in the study due to ascertainment bias. Another limitation of the study is the absence of control groups for estimating IA prevalence in the general population. There were only 31 patients (4.6% of the entire cohort) with one or more missing variables of which 16 (2.4%) had two missing variables. We considered this to be insignificant and we did not take any special precaution to account for this. Finally, there were few IA ruptures or cases of aneurysm enlargement, which precluded analysis of predictors for such outcomes.

Conclusion

This is the largest study of IA in patients with BAV, and the study showed that the prevalence of IA in patients with BAV was 7.7%. IA was more common in certain high-risk subgroups such as those with COA and in women with right–left cusp fusion.

The frequency of aneurysm rupture was similar to data from large registries of patients with unruptured IAs. A baseline brain MRA may be reasonable in all patients with BAV. However, a prospective study with consistent follow-up imaging is required to validate our data and to determine which patients with negative baseline imaging should have repeat study during follow-up.

Key messages

What is already known about this subject?

There is an increased risk of intracranial aneurysm (IA) in patients with coarctation of aorta (COA). Most patients with COA have bicuspid aortic valve (BAV). However, the prevalence of IA in patients with BAV is unknown.

What does this study add?

There is an increased prevalence of IA in patients with BAV (12.9% in patients with BAV and COA, and 5.7% in patients without COA). Among the patients without COA, the risk factors for IA are female gender and presence of right–left cusp fusion.

How might this impact clinical practice?

A population-based study is required to validate these findings and determine the strength of association between BAV and IA.

References

Footnotes

  • Contributors ACE, RP, ARK, EA: study design, data collection, data analysis, initial drafting, critical review, final revision. HMC, RDB, SAL, JH: data analysis, initial drafting, critical review, final revision. All authors reviewed and approved the final draft of this manuscript.

  • Competing interests None declared.

  • Ethics approval Mayo Institutional Review Board.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Linked Articles