Background CT angiography (CTA) is increasingly used for the detection, characterization, and follow-up of intracranial aneurysms. A lower threshold to request a CT angiogram may render a patient population that differs from previous studies primarily evaluated with conventional angiography. Our objective was to broaden our knowledge of the factors associated with aneurysm rupture and patient mortality in this population.
Methods All CTA studies performed over a 10-year period at a large neurovascular referral center were reviewed for the presence of an intracranial aneurysm. Patient demographics, mortality, CTA indication, aneurysm location, size, and rupture status were recorded.
Results 2927 patients with aneurysms were identified among 29 003 CTAs. 17% of the aneurysms were ruptured at the time of imaging, 24% of aneurysms were incidentally identified, and multiple aneurysms were identified in 34% of patients. Aneurysms most commonly arose from the supraclinoid internal carotid artery (22%), the middle cerebral artery (18%), and the anterior communicating artery (13%). Male sex, age <50 years, aneurysms >6 mm, and aneurysms arising from the anterior communicating artery, posterior communicating artery, or the posterior circulation were independent predictors of aneurysm rupture. Independent mortality predictors included male sex, posterior circulation aneurysms, intraventricular hemorrhage, and intraparenchymal hemorrhage.
Conclusions These results indicate that aneurysms detected on CTA that arise from the anterior communicating artery, posterior communicating artery, or the posterior circulation, measure >6 mm in size, occur in men, and in patients aged <50 years are associated with rupture.
- CT Angiography
Statistics from Altmetric.com
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.
Intracranial aneurysms (IA) are found in 2–7% of people, and subarachnoid hemorrhage secondary to IA rupture affects 30 000 people annually in the USA.1–5 The significance of IA rupture has led to several large studies intended to determine the natural history of IA and characteristics associated with IA rupture.6–10 These studies identified factors associated with IA rupture and patient mortality, including IA size >7–10 mm, IA within the posterior circulation, IA growth, hypertension, and smoking. Other studies have found that aneurysms <5 mm in diameter also have a significant risk of rupture, which indicates that the natural history of IA remains incompletely understood.10 ,11
Digital subtraction angiography (DSA) is the gold standard imaging modality in the evaluation of IA, given the excellent sensitivity and specificity of this technique.12 However, CT angiography (CTA) is increasingly used for the detection and characterization of both ruptured and unruptured IA, given its non-invasive nature and similar sensitivity and specificity.7 ,9 ,13 ,14 CTA is increasingly used to evaluate stroke, trauma, intracranial hemorrhage, and headaches, which often leads to the incidental identification of IA.7 ,13 ,14 The increased screening and discovery of IA by CTA has resulted in a patient population with IA that differs from previous large trials in which patients with IA were evaluated by DSA.6 ,15 Consequently, further studies are needed to elucidate the natural history of IA discovered in the population evaluated by CTA.
We sought to determine the characteristics of IA identified by CTA and factors associated with IA rupture and patient mortality. We reviewed all CTA studies that identified IA performed at our institution over a 10-year period to explore this population.
Our study was approved by the hospital's Institutional Review Board and complied with HIPAA regulations. We searched the radiology database with the Render program16 and identified all patients who underwent CTA from January 2000 to December 2010 with evidence of an IA. Duplicated studies were removed and only one study per patient was included in the dataset. For duplicated studies, the first CTA was included in the study unless the patient developed rupture of an IA, in which case the CTA performed at the time of rupture was included. Aneurysms that were discovered at outside institutions by CTA, MRA, or DSA underwent follow-up imaging by CTA at our institution prior to inclusion in this study. Among patients with ruptured aneurysms, all patients underwent a non-contrast CT (NCCT) prior to CTA of the head and/or neck. DSA was performed at the discretion of the treating neurosurgeon. Approximately 10% of the aneurysms identified in this study were confirmed by DSA, and there was no significant discrepancy between CTA and DSA in these instances.
NCCT was performed to assess for aneurysm rupture on 16- or 64-section General Electric helical CT scanners (LightSpeed; GE Medical Systems, Waukesha, Wisconsin, USA). NCCTs were performed using the helical technique with 120–140 kVp, auto mA (10–500), and 5 mm slice thickness reconstruction. Multidetector CTAs (MDCTA) were performed by scanning from the base of the C1 vertebral body or from the aortic arch to the vertex using the axial technique, 0.5 pitch, 1.25 mm collimation, 235 mA, 120kVp, 22 cm field of view, and 65–85 mL of iodinated contrast administered by power injector at 4–5 mL/s with either a fixed 25 s scanning delay or Smart-Prep, a semi-automatic contrast bolus triggering technique. MDCTA and NCCT acquisitions were both performed on the same hardware platform (LightSpeed).
All CTA studies at our institution are reviewed by dedicated neuroradiologists with at least 3 years of experience, and all queried CTA reports were interpreted by these neuroradiologists. Aneurysms were defined as rounded areas of contrast outpouching measuring ≥2 mm in diameter. Infundibula were defined as an outpouching with a vessel clearly arising from the apex of the outpouching. Fusiform aneurysms were excluded from the analysis. Three-dimensional reformations of all CTA studies were created in three planes (axial, sagittal, and coronal), and maximum intensity pixel images of the intracranial internal carotid arteries and intracranial vertebral and basilar arteries were also used during imaging interpretation. CTA images were reviewed to verify the reported presence of an IA by two neuroradiologists with 1 (JJH) and 10 (JMR) years of experience. IA were classified by location, size, and ruptured status based on the radiologist report and subsequent imaging review by two neuroradiologists. IA arising from the basilar artery, superior cerebellar artery, anterior inferior cerebellar artery, posterior inferior cerebellar artery, or the vertebral artery were considered to arise from the posterior circulation. Aneurysms were considered to be incidental if the CTA was performed for an indication other than: (1) follow-up of a known aneurysm; (2) assessment for an IA; (3) evaluation of intracranial hemorrhage; or (4) if multiple IA were identified in a patient with a ruptured IA.
Patient electronic medical records (EMRs) were reviewed for demographic information at the time of imaging and for in-hospital mortality secondary to IA rupture. The CTA requisitions were reviewed to determine the imaging indications. Mortality not determined in the EMR was assessed through the National Death Index (reported by the US Social Security Administration). Mortality among patients with ruptured aneurysms was secondary to aneurysm rupture. Mortality of patients with unruptured aneurysms was secondary to other causes in all patients for whom follow-up data were available in the EMR.
Statistical analysis was performed using SAS V.9.1 software package (SAS Institute) and XLSTAT (Addinsoft, New York, USA). Receiver operating characteristic analysis was used to determine the area under the curve for the size of IA in the prediction of IA rupture; aneurysm size analysis was then dichotomized at 6 mm. All variables including age, sex, aneurysm location, aneurysm size, and the presence of multiple IA were recorded and compared using multivariate logistic regression analysis to find possible significant predictors of mortality, and the level of significance was set at p<0.05 for all statistical analyses. Linear regression analysis was performed for continuous variables and for the evaluation of aneurysm location to determine independent predictors of mortality.
A total of 29 003 patients who underwent CTA were screened, and one or more IA was identified in 2927 patients (10%). Common indications for performing a CTA in patients found to have an IA were for evaluation or follow-up of an IA (43%), headache (26%), intracranial hemorrhage (14%), or stroke (10%). Some patients had multiple indications for obtaining a study (see online supplementary table S1). IA were discovered incidentally during CTA studies performed for other indications in 24% of studies, and an incidental IA was more likely to be identified in men than in women (36% vs 26%, p<0.0001; see online supplementary table S2).
Characteristics of ruptured and unruptured aneurysms
Most IA were identified in women (72%, p<0.001) and most were unruptured (83%), although men were more likely than women to present with a ruptured IA (20% vs 15%; p=0.003). There was no significant difference between men and women with respect to the frequency of aneurysm rupture when the indication for the CTA examination was for aneurysm follow-up or when an aneurysm was incidentally discovered (see online supplementary table S3). Multiple IA were identified in 532 patients (15%), and these IA were found in 4% of men compared with 37% of women (p<0.0001). These data are summarized in online supplementary tables S2 and S3.
There was no significant difference in the age between women and men with IA. Female patients who died were slightly older than the male patients who died (mean age 72 vs 69 years, p=0.005; see online supplementary table S4). Men were more likely than women to die from other causes while harboring an unruptured IA (18% vs 14% mortality, p=0.004), but there was no significant difference in mortality between women and men with ruptured IA (see online supplementary table S4).
Intracranial aneurysm distribution
The majority of IA in this series arose from the supraclinoid internal carotid artery (ICA) (22%) followed by the middle cerebral artery (MCA) (18%) and the anterior communicating artery (Acom) (13%). IA arising from the posterior circulation were found in 18% of patients. IA arising from the intradural segments of the ICA were more common in women (p=0.002 or less for each of these segments; table 1). By contrast, IA arising from the anterior cerebral artery, Acom, or vertebral artery were more common in men (p=0.006, p<0.0001, and p=0.05, respectively). The distribution of IA is listed in table 1.
A linear regression analysis was performed to determine whether IA location correlated with IA rupture and patient mortality. IA arising from the Acom (p<0.0001), basilar artery (p=0.001), posterior communicating artery (Pcom) (p<0.0001), or the posterior inferior cerebellar artery (p<0.0001) were significantly associated with IA rupture (table 2). Similarly, patient mortality correlated with IA arising from the Acom (p<0.001), basilar artery (p=0.01), Pcom (p=0.0004), the posterior inferior cerebellar artery (p<0.0001), or the vertebral artery (p<0.0001; table 3).
Intracranial aneurysm size
The size of IA ranged from 1 to 57 mm with a mean of 5.8 mm (median 4.0 mm). The majority of IA (n=2598) measured ≤6 mm in size, and 341 (13%) of these small IA ruptured (see online supplementary figure S1A). The average size of ruptured IA was 6.8 mm compared with 5.7 mm for unruptured IA (see online supplementary figure S1B; p<0.0001). When aneurysms measuring ≥25 mm (n=56) were excluded from the size analysis, the average size of ruptured and unruptured aneurysms was not significantly changed, which is consistent with the relatively small number of these outliers. When the distribution of ruptured and unruptured IA were compared, a significant increase in IA size was identified for ruptured IA arising from the Acom, supraclinoid ICA, MCA, and the Pcom (p<0.01 for all locations; table 2). The difference in size between ruptured and unruptured IA at other locations was not significant (table 2).
Factors associated with intracranial aneurysm rupture and patient mortality
A multivariate logistic regression analysis was performed to determine factors associated with IA rupture and patient mortality. Male sex (OR 1.3; 95% CI 1.1 to 1.6; p=0.01), IA size >6 mm (OR 2.5; 95% CI 2.1 to 2.9; p<0.0001), IA arising from the Acom (OR 4.2; 95% CI 3.3 to 5.3; p<0.0001), Pcom (OR 2.0; 95% CI 1.4 to 2.9; p<0.0001), or the posterior circulation (OR 2.5; 95% CI 2.1 to 3.1; p<0.0001), and patient age <50 years (OR 1.7; 95% CI 1.4 to 2.1; p<0.0001) were independent predictors of aneurysm rupture. Interestingly, the presence of multiple IA (OR 0.6; 95% CI 0.5 to 0.7; p<0.0001) was not associated with IA rupture, but rather it correlated with IA non-rupture (table 4).
Male sex (OR 1.5; 95% CI 1.2 to 1.8; p<0.0001) and IA located in the posterior circulation (OR 1.6; 95% CI 1.3 to 2.0; p<0.0001) were independently associated with patient mortality, regardless of whether an IA had ruptured. Among patients with ruptured IA, the presence of intraventricular hemorrhage (OR 1.7; 95% CI 1.3 to 2.2; p<0.0001) or intraparenchymal hemorrhage (OR 4.6; 95% CI 3.4 to 6.2; p<0.0001) was also associated with patient mortality. Patient age <50 years (OR 0.5; 95% CI 0.3 to 0.7; p<0.0001) and the presence of multiple IA (OR 0.7; 95% CI 0.5 to 0.8; p<0.0001) correlated with patient survival. IA size and IA arising from the Acom were not associated with patient mortality in this multivariate logistic regression analysis (table 4).
We screened for IA among 29 003 patients who underwent CTA at our hospital over a 10-year period and identified 3647 IA in a total of 2927 patients which, to our knowledge, represents the largest dataset of IA evaluated by CTA examinations. The 10% prevalence of IA in this study is slightly higher than the IA prevalence reported in other series.1–4
Nearly half of CTA studies that identified an aneurysm (48%) were performed for indications other than the assessment of an IA, and incidental IA were discovered in 24% of patients. Other studies have described a lower finding of incidental IA in 1–6% of patients undergoing CTA, whereas another recent study evaluating patients with IA found incidental IA in 91% of patients.9 ,17 This wide variation in the prevalence of incidental aneurysms probably reflects selection bias at our center and differences in patient populations and institutional thresholds to perform a CTA compared with DSA. The International Study of Unruptured Intracranial Aneurysms (ISUIA) identified a higher prevalence of IA in patients with headache (36% compared with 26% in this study) or stroke (18% compared with 10% in this study).6 These differences also likely reflect inherent differences in patients with IA identified by CTA compared with DSA.
Similar to previous studies, IA were much more commonly identified in women (72%),3 ,4 ,6 ,9 ,15 ,18 ,19 and multiple IA were identified in 34% of patients.6 ,8 ,20 Multiple IA were far more likely to be identified in women, which supports the notion that estrogen or other female sex-specific factors predispose patients to the development of IA.21
The presence of multiple IA has been associated with an increased risk of rupture.11 By contrast, we found that patients with multiple IA were less likely to develop rupture of any aneurysm. This unexpected result may be explained by more aggressive treatment of IA in patients with multiple IA or differences in the natural history of multiple IA within our patient population. Further studies are required to identify additional characteristics of patients with multiple IA that might influence the risk of IA rupture.
IA most commonly arose from the supraclinoid ICA (22%), the MCA (18%) and the Acom (13%). The distribution of IA varies between different studies, but most series identify the ICA,2 ,6 ,17 ,22 MCA,9 ,10 or Acom23 as the most common locations for IA to develop. Thus, the distribution of IA in our study is not dissimilar to previous reports. We note that the predominance of supraclinoid IA in this study may also reflect a selection bias, given that the study was performed at a large neurovascular referral center.
The average IA size of 5.8 mm is similar to other studies.6 ,8–10 ,15 We found that IA measuring >6 mm were associated with rupture, which is smaller than the 7 mm size cited in prior studies.6 ,8 ,9 ,15 IA rupture was identified in 13% of IA measuring ≤6 mm in size, which contrasts with the low risk of rupture among small aneurysms in the ISUIA study.6 However, other studies have identified similar rates of rupture of 3–13% among IA measuring ≤6 mm.9 ,11 ,24
In a multivariate logistic regression we identified characteristics associated with IA rupture and patient mortality. Male sex, age <50 years, IA arising from the Acom, Pcom, or within the posterior circulation, and IA size >6 mm were associated with IA rupture. Other studies have found IA size >7–10 mm, IA arising from the Acom or the posterior circulation, and patient age <50 years to be associated with IA rupture.6 ,8 ,11 ,15 Previous studies have also found female sex, the presence of multiple IA, and patient age >60 years (which contrasts with the study described above) as risk factors for IA rupture.6 ,8 ,11 ,18 While these differences may reflect differences in the patient populations between these studies, it is also possible that men in this study had additional risk factors not identified in our image-based assessment that predispose them to IA rupture. It is not intuitive that patients with multiple IA would have a significantly reduced risk of IA rupture, but perhaps patients with multiple IA are more likely to undergo treatment of their aneurysms prior to rupture. Alternatively, patients with multiple IA in our study may have other clinical characteristics that are protective against IA rupture. It will be of interest to determine whether the male patients and patients with multiple IA in this study have differences in the prevalence of hypertension, smoking, or other risk factors that influence the possibility of IA rupture.
We found patient age >50 years, IA location within the posterior circulation, the presence of intraventricular hemorrhage, and the presence of intraparenchymal hemorrhage to be associated with patient mortality in the setting of aneurysm rupture, which is similar to previous studies.6 ,8 ,25 ,26 We also found male sex to be an independent predictor of patient mortality, which differs from other studies that have found a higher mortality rate among women.9 ,11 Other studies have found no significant effect of patient sex upon mortality.19 As noted above, it will be of interest to determine whether the men in this study had additional risk factors or if they presented in a worse clinical state than the women. We did not find IA size to be a predictor of patient mortality, which indicates that IA size may affect the probability of rupture but, once IA rupture has occurred, patient mortality is influenced by other factors.
The majority of patients in this series did not undergo serial CTA examinations. We have previously published the results of 152 patients in this series who underwent serial CTA studies for conservative aneurysm follow-up.27 Among these patients, six developed subarachnoid hemorrhage due to aneurysm rupture over the follow-up period, and aneurysm size, aneurysm growth, and aneurysm multilobulation were significant predictors of rupture. These results are similar to those reported by Villablanca et al,7 who also found aneurysm size and growth on serial CTA to be predictors of aneurysm rupture.
The limitations of this study include its retrospective design. Furthermore, this is a single-center study from a large neurovascular referral center, and the patient population included in this study may not be generalizable to a broader patient population. Despite our diligent search, it is possible that individual cases might have been missed. Moreover, the search relies on key words in the radiology report and thus discretionary cases (very small aneurysms) might be missed because the radiologist opted not to term the finding an aneurysm. Additionally, CTA studies that failed to identify an aneurysm would be excluded, although these false negative studies are probably few, given the excellent sensitivity of CTA in the detection of IA.7
Cerebral aneurysms detected with MDCT angiography that arise from the Acom, Pcom, or the posterior circulation, measure >6 mm in size, occur in men and in patients <50 years of age are associated with rupture.
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
- Data supplement 1 - Online supplement
Twitter Follow Edgar Ordonez-Rubiano at @edgargo88
Contributors JJH, JMR, JAH, and RGG conceived and designed the research. JJH, DS, and EGOR acquired the data. JJH, JMR, JAH, RGG, and PWS analyzed and interpreted the data. JJH and JMR drafted the manuscript. JJH, JMR, JAH, RGG, and HBB made critical revisions to the manuscript. JJH, JMR, JAH, RGG, DS, EGOR, HBB, and PWS approved the final version of the manuscript.
Competing interests None declared.
Ethics approval Ethics approval was obtained from the Massachusetts General Hospital IRB.
Provenance and peer review Not commissioned; externally peer reviewed.
Data sharing statement Data available upon request and collaboration agreement.