Relationship of brainstem infarction to rupture of nonsaccular vertebrobasilar aneurysms

View More View Less
  • 1 Departments of Neurosurgery,
  • | 2 Neurology, and
  • | 3 Radiology, Mayo Clinic, Rochester, Minnesota
Free access

OBJECTIVE

Symptomatic nonsaccular vertebrobasilar aneurysms (NSVBAs) are associated with high rates of aneurysm-related death. Anecdotal evidence suggests that brainstem infarction may be a harbinger of aneurysm rupture. The authors aimed to investigate the association between brainstem infarction and subsequent NSVBA rupture.

METHODS

The clinical records and radiographic imaging studies of patients presenting to the authors’ institution between 1996 and 2019 for evaluation and management of an NSVBA were retrospectively reviewed to determine the effect of perforating artery infarction on the natural history of NSVBAs. Kaplan-Meier curves for patients with and patients without perforator infarction were constructed, and predictors of aneurysm rupture were identified using a multivariate Cox proportional hazards model.

RESULTS

There were 98 patients with 591.3 person-years of follow-up who met the inclusion criteria for analysis. There were 20 patients who experienced perforator infarction during follow-up. Ten patients (10.2%) experienced aneurysm rupture during follow-up and 26 patients (26.5%) died due to aneurysm-related complications, with annual rates of rupture and aneurysm-related death of 1.7% and 4.4%, respectively. Five patients with a perforator infarction later experienced aneurysm rupture, with a median time between infarction and rupture of 3 months (range 0–35 months). On multivariate analysis, the presence of intraaneurysmal thrombus (risk ratio [RR] 4.01, 95% confidence interval [CI] 1.12–14.44, p = 0.033) and perforator infarction (RR 6.37, 95% CI 1.07–37.95, p = 0.042) were independently associated with risk of aneurysm rupture.

CONCLUSIONS

NSVBAs continue to be extremely challenging clinical entities with a poor prognosis. These results suggest that brainstem infarction due to perforating artery occlusion may be a harbinger of near-term aneurysm rupture.

ABBREVIATIONS

CI = confidence interval; NSVBA = nonsaccular vertebrobasilar aneurysm; RR = risk ratio.

OBJECTIVE

Symptomatic nonsaccular vertebrobasilar aneurysms (NSVBAs) are associated with high rates of aneurysm-related death. Anecdotal evidence suggests that brainstem infarction may be a harbinger of aneurysm rupture. The authors aimed to investigate the association between brainstem infarction and subsequent NSVBA rupture.

METHODS

The clinical records and radiographic imaging studies of patients presenting to the authors’ institution between 1996 and 2019 for evaluation and management of an NSVBA were retrospectively reviewed to determine the effect of perforating artery infarction on the natural history of NSVBAs. Kaplan-Meier curves for patients with and patients without perforator infarction were constructed, and predictors of aneurysm rupture were identified using a multivariate Cox proportional hazards model.

RESULTS

There were 98 patients with 591.3 person-years of follow-up who met the inclusion criteria for analysis. There were 20 patients who experienced perforator infarction during follow-up. Ten patients (10.2%) experienced aneurysm rupture during follow-up and 26 patients (26.5%) died due to aneurysm-related complications, with annual rates of rupture and aneurysm-related death of 1.7% and 4.4%, respectively. Five patients with a perforator infarction later experienced aneurysm rupture, with a median time between infarction and rupture of 3 months (range 0–35 months). On multivariate analysis, the presence of intraaneurysmal thrombus (risk ratio [RR] 4.01, 95% confidence interval [CI] 1.12–14.44, p = 0.033) and perforator infarction (RR 6.37, 95% CI 1.07–37.95, p = 0.042) were independently associated with risk of aneurysm rupture.

CONCLUSIONS

NSVBAs continue to be extremely challenging clinical entities with a poor prognosis. These results suggest that brainstem infarction due to perforating artery occlusion may be a harbinger of near-term aneurysm rupture.

symptomatic nonsaccular vertebrobasilar aneurysms (NSVBAs) remain associated with a dismal prognosis. Not only do these aneurysms rupture at a fairly high annual rate,1–4 but as they enlarge, substantial morbidity and mortality can result due to mass effect on the brainstem.5 In addition, once a thrombus forms within the aneurysm sac, patients are at high risk for thromboembolic events, with a high incidence of recurrent infarction after an initial ischemic event.6

It is possible that in certain cases, the most serious clinical manifestations of NSVBAs—specifically rupture, mass effect, and ischemia—are related to a single underlying disease process. Fusiform intracranial aneurysms are characterized histopathologically by disruption of the internal elastic lamina and neoangiogenesis within a thickened intimal layer. These nascent vessels are prone to hemorrhage, leading to deposition of intramural thrombus and associated aneurysm enlargement. As neoangiogenesis continues, this process repeats itself, resulting in a vicious cycle of aneurysm enlargement, intramural hemorrhage, and deposition of thrombus.7,8 Over time, mass effect on the brainstem increases, and ischemia results from occlusion of vital basilar perforating arteries to the brainstem, along with emboli to the distal posterior circulation. While the intraaneurysmal flow dynamics leading to growth and rupture are undoubtedly complex and defy oversimplification,9 it is possible that in some instances wall inflammation associated with a large thrombus may precipitate aneurysm rupture, particularly in the setting of a characteristically degenerated tunica media.10 In our experience, we have observed patients with NSVBAs initially presenting with brainstem infarction who subsequently died from aneurysm rupture within weeks or even days of the initial ischemic event (Fig. 1).11 We hypothesize that the infarction, which we presume results from perforating artery occlusion, denotes acceleration of the vicious cycle underlying aneurysm enlargement and may be a harbinger of rupture in certain cases. To address this question, we reviewed the clinical course of patients with NSVBA presenting to our institution over a 23-year period, with special attention to the effect of brainstem infarction on subsequent aneurysmal rupture.

FIG. 1.
FIG. 1.

Representative case. A: A middle-aged man with hypertension and a tobacco-use disorder presented with double vision, dysarthria, left-sided hemiparesis, and ataxia and was found to have a brainstem infarct at the pontomedullary junction (arrow), as well as multifocal infarcts in the right cerebellar hemisphere and middle cerebellar peduncle (arrowheads) on diffusion-weighted MRI sequences. B: MR angiography demonstrated a large, partially thrombosed transitional-type aneurysm spanning the vertebrobasilar junction. C and D: Several days after inpatient admission, the patient became acutely unresponsive and was found to have a large subarachnoid hemorrhage on noncontrast-enhanced CT of the head.

Methods

Patient Selection and Determination of Follow-Up Time

After approval from our IRB, we retrospectively reviewed the clinical course and radiological images of patients presenting to our institution for management of an NSVBA from 1996 to 2019. Patients were identified by searching radiology reports within our institutional database for the keywords “vertebrobasilar dolichoectasia” or “vertebral and/or basilar fusiform aneurysm.” Patients confirmed to have fusiform- or transitional-type aneurysms of the vertebrobasilar system were included for analysis. In accordance with established criteria, a fusiform aneurysm was defined as an aneurysmal dilation measuring at least 1.5 times the normal diameter of the affected vessel without a definable neck, while a transitional aneurysm was defined as a fusiform aneurysm superimposed on a dolichoectatic vertebral or basilar artery.6 Patients with isolated vertebrobasilar dolichoectasia without an aneurysmal component were excluded, given that these lesions are associated with a significantly lower risk of rupture.1

Both clinical and imaging follow-up durations were recorded. Clinical follow-up duration was defined as the time of first intracranial imaging study, specifically MRI of the brain with contrast, CT, MRI, or digital subtraction angiography, to the time of last clinical follow-up at our institution, or death due to the aneurysm or other cause. Imaging follow-up duration was defined as the time between the first and last imaging study. Follow-up duration was reported in years unless otherwise specified.

Variables of Interest

Patient demographic information was collected and included patient sex, age at initial imaging study demonstrating the NSVBA, and underlying comorbidities. Relevant comorbidities noted included hypertension, hyperlipidemia, coronary artery disease, current smoking at the time of presentation, type 2 diabetes mellitus, and presence of an abdominal aortic aneurysm. Collected information on aneurysm characteristics included aneurysm type, location, maximum diameter as determined on MRI or CT angiography, and presence of thrombus within the aneurysm sac on the initial imaging study. Aneurysm type was defined as fusiform or transitional morphology, and location was defined as basilar, vertebral, or vertebrobasilar. Maximal dimension was determined on MRI or CT angiography as opposed to formal angiography, in order to include noncontrast-filling portions of the aneurysms (i.e., thrombosed portions) in the measurement of diameter. Information on presenting symptomatology was also collected, and the reason for NSVBA evaluation was categorized as one of the following: incidental, posterior circulation infarction, symptoms of brainstem mass effect, neuralgia, or headache. Patients with anterior circulation infarction or cerebrovascular pathology other than the NSVBA were assigned to the incidental group. Patients categorized as presenting due to brainstem mass effect most frequently had symptoms such as cranial nerve dysfunction, motor weakness, or ataxia in the setting of significant radiographic brainstem compression and absence of brainstem or posterior circulation restricted diffusion on brain MRI that could otherwise explain their symptoms. Patients with restricted diffusion were characterized as presenting due to posterior circulation infarction regardless of the presence of brainstem mass effect. We also noted the presence of brainstem perforator infarction, either on presentation or during follow-up, which was defined as restricted diffusion within the brainstem parenchyma. Examples of perforator infarction can be seen in Fig. 1. Information on medical management was also collected, specifically the use of aspirin, clopidogrel, or oral anticoagulation during the follow-up period.

Outcomes of Interest

Outcomes of interest included aneurysm growth during follow-up, the incidence of new cerebral infarction related to the NSVBA, aneurysm rupture, and death due to the aneurysm. Growth was defined as enlargement of at least 2 mm on follow-up imaging. All new infarcts were confirmed by MRI, and aneurysm rupture was confirmed on noncontrast head CT demonstrating subarachnoid hemorrhage in the basilar cisterns. The clinical records of all deceased patients were reviewed, and the cause of death was ascertained when possible. The incidence of death secondary to the aneurysm was determined and was categorized as due to aneurysm rupture, complications of recurrent ischemia, or complications of brainstem mass effect. The cause of death could not be ascertained with a reasonable degree of certainty for 3 of the 43 patients who died during follow-up. In the remaining cases, based on available clinical information it was ascertained that these patients died from causes unrelated to the aneurysm.

Information on treatment of NSVBA was also collected. Although treatment strategies for NSVBA evolved over the 20-year study period, patients were considered for treatment if they presented with symptoms primarily related to brainstem mass effect in the setting of an enlarging aneurysm. Patients presenting with ischemia were typically not considered for treatment due to the potential to induce further thrombosis and recurrent ischemia with endovascular therapy. All patients were treated endovascularly, and specific information on the treatment modality was noted. The incidence of ischemic complications related to treatment was recorded, as was the time to death after treatment.

Statistical Analyses

Descriptive statistics for continuous and categorical variables were presented as mean and standard deviation and frequency and percentage, respectively. Comparative statistics for continuous and categorical variables were performed using the Student t-test and Pearson chi-square test, where appropriate. Kaplan-Meier curves were constructed to perform survival analyses evaluating time to aneurysm rupture and death related to the aneurysm. Results were reported as median time to event, defined as the time in years at which point 50% of patients had experienced the event of interest. Separate Kaplan-Meier curves comparing the survival distribution of patients according to presentation symptomatology, aneurysm type, and presence or absence of perforator infarction were also constructed. Clinical follow-up duration, defined as time from initial imaging study to last clinical follow-up or death, was used as time to event or censorship, except for curves comparing patients with and without perforator infarction. For these latter graphs, the time from the last perforator infarction to the event of interest or censorship was used. Curves comparing the distribution of patients according to presentation and aneurysm type were compared using the log-rank test, while curves for patients with and without perforator infarction were compared using the generalized Wilcoxon rank-sum test. The Wilcoxon rank-sum test was selected over the log-rank test to more heavily weight early events after brainstem infarction. Predictors of aneurysm rupture were determined using a multivariate Cox proportional hazards model. Univariate analysis was performed, and variables significantly associated with aneurysm rupture were included in a multivariate model. The alpha level for statistical significance was set at 0.05. Statistical analyses were performed using commercially available software (JMP version 10.0.0, SAS Institute Inc.).

Results

Patient and Aneurysm Characteristics

There were 98 patients who met the inclusion criteria for analysis. A majority of patients were male (80.6%), and the mean age at presentation was 60.1 years. The NSVBA was an incidental discovery in 55.1% of patients. Posterior circulation infarction (19.4%) and symptoms related to brainstem compression (17.3%) were the two most common reasons for presentation in symptomatic cases. A majority of aneurysms had fusiform (57.1%) as opposed to transitional morphology (42.9%), and aneurysms most commonly spanned the vertebrobasilar junction (46.9%) as opposed to being isolated to the basilar (43.9%) or vertebral (9.2%) arteries. The mean maximal aneurysm diameter on presentation was 11.3 ± 5.5 mm, and 52.0% of aneurysms had radiographic evidence of thrombus within the aneurysm sac on initial imaging studies. A complete summary of patient and aneurysm characteristics is presented in Table 1.

TABLE 1.

Patient and aneurysm characteristics

VariableValue (%)
No. of patients98
Sex
 Male79 (80.6)
 Female19 (19.4)
Mean age ± SD, yrs60.1 ± 13.8
 Median61
 Range15–88
Comorbidities
 Hypertension72 (73.5)
 Hyperlipidemia55 (56.1)
 Coronary artery disease28 (28.6)
 Current smoker26 (26.5)
 Diabetes mellitus type 216 (16.3)
 Abdominal aortic aneurysm15 (15.3)
Presenting symptom
 Incidental54 (55.1)
 Posterior circulation infarction19 (19.4)
 Brainstem compression17 (17.3)
 Neuralgia5 (5.1)
 Headache3 (3.1)
Aneurysm type
 Fusiform56 (57.1)
 Transitional42 (42.9)
Aneurysm location
 Vertebrobasilar46 (46.9)
 Basilar artery43 (43.9)
 Vertebral artery9 (9.2)
Mean max diameter ± SD, mm11.3 ± 5.5
 Median9.1
 Range4.8–30.0
Thrombus formation
 No51 (52.0)
 Yes47 (48.0)
Antiplatelet therapy
 Aspirin81 (82.7)
 Clopidogrel25 (25.5)
 Dual antiplatelet therapy23 (23.5)
Vitamin K antagonist/direct oral anticoagulant27 (27.6)

Patient Outcomes

For the 98 patients in our study, there were 591.3 and 361.9 clinical and imaging person-years of follow-up, respectively. There were 40 aneurysms that showed radiographic evidence of growth on follow-up imaging, with an annual growth rate of 11.1%. Thirteen patients suffered a new posterior circulation infarction during follow-up, with an annual rate of 3.6%. Aneurysm rupture and death due to the aneurysm occurred in 10 and 26 patients, with respective annual rates of 1.7% and 4.4%. Among the aneurysm-related deaths, 10 (38.5%) were due to rupture, 10 (38.5%) were due to complications of brainstem compression, and 6 (23.1%) were due to complications of recurrent ischemia. Patient outcomes are summarized in Table 2.

TABLE 2.

Patient and treatment outcomes

VariableValue (%)
Aneurysm growth40/98 (40.8)
 Annual rate (%)11.1
New stroke attributable to aneurysm13/98 (13.3)
 Annual rate (%)3.6
Aneurysm rupture10/98 (10.2)
 Annual rate (%)1.7
Death due to aneurysm26/98 (26.5)
 Annual rate (%)4.4
Type of aneurysm-related death
 Rupture10 (38.5)
 Mass effect10 (38.5)
 Ischemia6 (23.1)
Aneurysm treatment14 (14.3)
 Flow diversion6 (42.9)
 Stenting5 (35.7)*
 Coiling2 (14.3)*
 Vertebral artery sacrifice2 (14.3)
Treatment-related ischemic complications6 (42.9)
Death after aneurysm treatment8/14 (57.1)
 Mean mos until death ± SD12.4 ± 15.8
 Median mos until death6

One patient was treated with stent-assisted coiling.

On Kaplan-Meier analysis, the median time to death related to the aneurysm was 14.3 years (Fig. 2A). The median time to rupture could not be calculated due to the insufficient number of events (Fig. 2B). Patients suffering a perforator infarction had significantly shorter median survival times compared to patients without perforator infarction (median 3.5 vs 14.1 years; Fig. 2C). The time to aneurysm rupture was also significantly different for patients with and without perforator infarction (p < 0.001; Fig. 2D). Kaplan-Meier curves stratifying patients according to presentation symptomatology and aneurysm morphology were also constructed. Survival rates were lowest for patients presenting with brainstem compression and with transitional aneurysm morphology (Supplementary Fig. 1).

FIG. 2.
FIG. 2.

A and B: Kaplan-Meier curves demonstrating time to aneurysm-related death (A) and aneurysm rupture for all patients (B). C and D: Kaplan-Meier curves stratified by incidence of perforator infarction demonstrating time to death related to the aneurysm (C) and aneurysm rupture (D). The p values were obtained using the Wilcoxon rank-sum test. Figure is available in color online only.

Treatment Outcomes

There were 14 patients who underwent treatment for their NSVBA. Seven of these patients (50.0%) initially presented with symptoms related to brainstem compression, while 3 patients (21.4%) presented with posterior circulation ischemia. Of the remaining patients, 2 aneurysms (14.3%) were initially incidental, 1 (7.1%) was causing trigeminal neuralgia, and 1 was deemed to be the cause of new occipital headaches. All patients experienced clinical worsening in the setting of radiographic aneurysm enlargement despite maximal medical therapy prior to treatment. Most aneurysms were treated with flow diversion (6/14, 42.9%), with basilar stenting as the next most common treatment modality (5/14, 35.7%). All patients undergoing treatment after 2011 received treatment with flow diversion. Treatment-related ischemic complications occurred in 6 patients (42.9%), with brainstem infarction followed by aneurysm rupture and death occurring in 1 patient. Eight patients (57.1%) eventually died after treatment, with a median time from treatment to death of 6 months. A summary of treatment information is presented in Table 2.

Relationship of Perforator Infarction to Aneurysm Rupture

There were 20 patients who experienced perforator infarction, 13 of whom initially presented with brainstem infarction and 7 who experienced such infarction during follow-up. There were 3 patients who experienced perforator infarction after aneurysm treatment, 1 of whom experienced aneurysm rupture shortly after said infarction. Overall, 5 of the 10 patients who ultimately suffered aneurysm rupture experienced a perforator infarction either at presentation or during follow-up. The mean and median times from perforator infarction to aneurysm rupture were 10.6 ± 14.8 months and 3 months, with a range of 0–35 months (Table 3). Baseline demographic and aneurysm characteristics for patients with and without perforator infarction at presentation or during follow-up were compared, and no significant differences were found (Supplementary Table 1).

TABLE 3.

Temporal relationship of perforator infarction to aneurysm rupture

VariableValue (%)
Patients with perforator infarction/all patients20/98 (20.4)
 Perforator infarction during follow-up/all patients with perforator infarction7/20 (35.0)
 Perforator infarction after treatment/all treated patients3/14 (21.4)
Patients with perforator infarction prior to aneurysm rupture/all patients with rupture5/10 (50.0)*
 Aneurysm ruptured after treatment/all treated patients1/14 (7.1)
Mean mos from last infarct to rupture ± SD10.6 ± 14.8
 Median3
 Range0–35

Denotes percentage of patients with aneurysm rupture who also had a perforator infarction.

Predictors of aneurysm rupture were identified using a Cox proportional hazards model. On univariate analysis, increasing aneurysm maximal diameter (unit risk ratio [RR] 1.85, 95% confidence interval [CI] 1.03–3.14, p = 0.025), presence of intraaneurysmal thrombus (RR 7.95, 95% CI 1.61–39.18, p = 0.011), aneurysm growth during follow-up (RR 3.92, 95% CI 1.01–15.26, p = 0.049), new posterior circulation infarction during follow-up (RR 4.44, 95% CI 1.23–16.09, p = 0.023), and perforator infarction (RR 4.50, 95% CI 1.30–15.63, p = 0.018) were associated with increased risk of subsequent aneurysm rupture. On multivariate analysis, the presence of intraaneurysmal thrombus (RR 4.01, 95% CI 1.12–14.44, p = 0.033) and perforator infarction (RR 6.37, 95% CI 1.07–37.95, p = 0.042) were independently associated with risk of aneurysm rupture. The complete results of the proportional hazards analysis are presented in Table 4.

TABLE 4.

Cox proportional hazards model identifying predictors of aneurysm rupture during follow-up

PredictorUnivariateMultivariate 
OR (95% CI)p ValueOR (95% CI)p Value 
Male sex1.12 (0.24–5.30)0.88 
Age, yrs0.87 (0.57–1.43)0.56 
Hypertension2.26 (0.28–18.18)0.44 
Hyperlipidemia1.33 (0.33–5.35)0.69 
Coronary artery disease2.81 (0.78–10.07)0.11 
Current smoker3.73 (0.99–13.96)0.051 
Diabetes mellitus type 20.46 (0.06–3.63)0.46 
Abdominal aortic aneurysm3.11 (0.77–12.49)0.11 
Aneurysm type 
 FusiformRef 
 Transitional3.85 (0.99–14.98)0.052 
Max diameter1.85 (1.03–3.14)*0.0251.14 (0.50–2.25)*0.73 
Thrombus formation7.95 (1.61–39.18)0.0114.01 (1.12–14.44)0.033 
Presenting symptom 
 IncidentalRef 
 Posterior circulation infarction4.29 (0.96–19.28)0.057 
 Brainstem compression2.06 (0.21–20.36)0.54 
 Neuralgia3.40 (0.51–22.80)0.21 
 HeadacheNA0.99 
Aneurysm growth3.92 (1.01–15.26)0.0493.07 (0.64–14.82)0.16 
New infarction4.44 (1.23–16.09)0.023 
Perforator infarction4.50 (1.30–15.63)0.0186.37 (1.07–37.95)0.042 

NA = not available; Ref = reference.

Boldface type indicates statistical significance.

Unit ORs denoting effect on risk for every decade of age and 5-mm increase in size.

New posterior circulation infarction was not included in the multivariate model due to colinearity with perforator infarction.

Discussion

In this study, we reviewed the clinical course of patients presenting to our institution with a fusiform- or transitional-type NSVBA over an approximately 23-year period. Our results largely confirm the poor natural history of these lesions, particularly for patients presenting with brainstem compression and/or infarction related to the aneurysm. We also demonstrate a potential association between brainstem perforator infarction and aneurysm rupture shortly thereafter. These findings may help clarify the pathophysiology of NSVBA rupture and help guide prognostic discussions in patients presenting for evaluation and management.

The observed annual stroke, rupture, and mortality rates due to NSVBA in our study are comparable to those observed in prior studies.3,5 Regarding risk factors for rupture, a prior study from our institution describing the clinical course of patients treated between 1989 and 2001 found fusiform- or transitional-type morphology to be predictive of aneurysm rupture, with a comparable annual rate of rupture among these aneurysm types to that observed in our study (1.7% vs 2.3%).1 Aneurysm growth was also predictive of rupture in this study, which in our study was associated with hemorrhage on univariate but not multivariate analysis. While prior studies have focused on morphology and size as predictors of rupture, the potential association between brainstem infarction and aneurysm rupture has not been previously investigated. We found perforator infarction, either as a presenting symptom or occurring during follow-up, to be the only factor in addition to intraaneurysmal thrombus to be independently associated with a risk of rupture. Of interest, in a few patients, rupture occurred within days or weeks following the brainstem infarct, suggesting a potential link between factors triggering the infarction and rupture. The sample size of our study is nevertheless small, as is the number of patients experiencing perforator infarction prior to aneurysm rupture (n = 5), and thus these findings would need to be confirmed in larger studies.

The potential mechanism underlying the possible association between brainstem infarction and rupture is not immediately clear. We suspect the infarction may represent an acceleration of the disease process responsible for aneurysm enlargement, specifically intramural hemorrhage from friable, newly formed blood vessels and subsequent thrombus deposition leading to aneurysm enlargement.7,8 While in some cases this process likely proceeds at a relatively slower pace, causing morbidity and mortality from complications of brainstem mass effect or recurrent ischemia, in others rapid aneurysm expansion due to fulminant thrombosis may precipitate rupture, potentially by extension of thrombus beyond the tunica media in a manner similar to that of subarachnoid hemorrhage from dissecting pseudoaneurysms.12 This process may also be initiated iatrogenically, which potentially occurred in 1 patient in our series who experienced acute aneurysm thrombosis and associated infarction after endovascular treatment, followed shortly thereafter by fatal subarachnoid hemorrhage. Alternatively, the inflammation associated with a large burden of thrombus may precipitate rupture of an already compromised aneurysmal wall with degenerated intimal and medial layers, much like the hypothesized mechanism underlying delayed rupture of giant aneurysms after endovascular flow diversion.13

Overall, there were 14 patients who underwent treatment of their NSVBA during follow-up, most of whom were treated with endovascular flow diversion. Similar to prior studies reporting outcomes of flow diversion for treatment of NSVBA,14–16 we observed considerable morbidity and mortality after treatment. Whether or not the potential benefits of treatment can ever outweigh the seemingly prohibitive risks of therapy remains an open question. In general, we consider flow diversion for patients presenting with symptoms of mass effect, with the aim of arresting further aneurysmal growth. Even with optimal device placement, however, acute aneurysm thrombosis can lead to debilitating infarction and potentially rupture. To improve treatment outcomes, there has been considerable interest in identifying the optimal treatment timing.5 For example, one could speculate that intervention earlier in the disease course could potentially alter disease natural history, while intervention on “end-stage” disease may be futile as the cycle of aneurysm thrombosis and enlargement is too advanced to disrupt and may, in fact, be accelerated by treatment. Such an approach would nevertheless be fraught, given the uncertain natural history of these lesions, particularly when they are incidentally discovered. A more in-depth understanding of this disease process, and the factors predicting malignant transformation, is needed to better manage these difficult lesions.

Study Limitations

Our study is limited by its single-center, retrospective methodology. Specifically, patients did not undergo standardized imaging follow-up, and there was a wide range in the number of imaging studies undergone by patients within our cohort, potentially influencing our results. In addition, our results were dependent on accurate reporting of clinical events, particularly surrounding patient deaths. It is possible that certain deaths related to the NSVBA were missed due to incomplete documentation. Similarly, the diagnosis of aneurysm rupture was based on CT imaging obtained after the ictus; it is possible that hemorrhages in patients who did not undergo imaging were missed, potentially affecting our results. Despite these limitations, our study represents one of the largest series to date reporting on the clinical history of patients with specifically fusiform- or transitional-type vertebrobasilar aneurysms.

Conclusions

We report a close temporal association between brainstem infarction and subsequent aneurysm rupture in patients with NSVBA. These findings may be of utility in the prognostication and management of these patients but require confirmation in larger studies.

Disclosures

Dr. Lanzino reports being a consultant for Superior Medical Editing and Nested Knowledge.

Author Contributions

Conception and design: all authors. Acquisition of data: Rinaldo, Nasr, Flemming. Analysis and interpretation of data: Rinaldo, Flemming, Lanzino, Brinjikji. Drafting the article: Rinaldo, Nasr. Critically revising the article: all authors. Reviewed submitted version of manuscript: all authors. Approved the final version of the manuscript on behalf of all authors: Rinaldo. Statistical analysis: Rinaldo. Study supervision: Brinjikji.

Supplemental Information

Online-Only Content

Supplemental material is available with the online version of the article.

References

  • 1

    Flemming KD, Wiebers DO, Brown RD Jr, et al. Prospective risk of hemorrhage in patients with vertebrobasilar nonsaccular intracranial aneurysm. J Neurosurg. 2004;101(1):8287.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 2

    Mangrum WI, Huston J III, Link MJ, et al. Enlarging vertebrobasilar nonsaccular intracranial aneurysms: frequency, predictors, and clinical outcome of growth. J Neurosurg. 2005;102(1):7279.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 3

    Nasr DM, Brinjikji W, Rouchaud A, et al. Imaging characteristics of growing and ruptured vertebrobasilar non-saccular and dolichoectatic aneurysms. Stroke. 2016;47(1):106112.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 4

    Nasr DM, Flemming KD, Lanzino G, et al. Natural history of vertebrobasilar dolichoectatic and fusiform aneurysms: a systematic review and meta-analysis. Cerebrovasc Dis. 2018;45(1-2):6877.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 5

    Xu DS, Levitt MR, Kalani MYS, et al. Dolichoectatic aneurysms of the vertebrobasilar system: clinical and radiographic factors that predict poor outcomes. J Neurosurg. 2018;128(2):560566.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 6

    Flemming KD, Wiebers DO, Brown RD Jr, et al. The natural history of radiographically defined vertebrobasilar nonsaccular intracranial aneurysms. Cerebrovasc Dis. 2005;20(4):270279.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 7

    Nakayama Y, Tanaka A, Kumate S, et al. Giant fusiform aneurysm of the basilar artery: consideration of its pathogenesis. Surg Neurol. 1999;51(2):140145.

  • 8

    Nakatomi H, Segawa H, Kurata A, et al. Clinicopathological study of intracranial fusiform and dolichoectatic aneurysms: insight on the mechanism of growth. Stroke. 2000;31(4):896900.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 9

    Brinjikji W, Chung B, Yong-Hong D, et al. Hemodynamic characteristics of stable and unstable vertebrobasilar dolichoectatic and fusiform aneurysms. J Neurointerv Surg. 2018;10(11):11021107.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 10

    Schievink WI, Parisi JE, Piepgras DG, Michels VV. Intracranial aneurysms in Marfan’s syndrome: an autopsy study. Neurosurgery. 1997;41(4):866871.

  • 11

    Flemming KD, Josephs K, Wijdicks EF. Enlarging vertebrobasilar dolichoectasia with subarachnoid hemorrhage heralded by recurrent ischemia. Case illustration. J Neurosurg. 2000;92(3):504.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 12

    Barletta EA, Ricci RL, Silva RDG, et al. Fusiform aneurysms: a review from its pathogenesis to treatment options. Surg Neurol Int. 2018;9:189.

  • 13

    Rouchaud A, Brinjikji W, Lanzino G, et al. Delayed hemorrhagic complications after flow diversion for intracranial aneurysms: a literature overview. Neuroradiology. 2016;58(2):171177.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 14

    Ertl L, Holtmannspötter M, Patzig M, et al. Use of flow-diverting devices in fusiform vertebrobasilar giant aneurysms: a report on periprocedural course and long-term follow-up. AJNR Am J Neuroradiol. 2014;35(7):13461352.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 15

    Siddiqui AH, Abla AA, Kan P, et al. Panacea or problem: flow diverters in the treatment of symptomatic large or giant fusiform vertebrobasilar aneurysms. J Neurosurg. 2012;116(6):12581266.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 16

    Meckel S, McAuliffe W, Fiorella D, et al. Endovascular treatment of complex aneurysms at the vertebrobasilar junction with flow-diverting stents: initial experience. Neurosurgery. 2013;73(3):386394.

    • Crossref
    • Search Google Scholar
    • Export Citation

Selected panels from a figure in Satzer et al. (pp 1742–1751).

  • View in gallery

    Representative case. A: A middle-aged man with hypertension and a tobacco-use disorder presented with double vision, dysarthria, left-sided hemiparesis, and ataxia and was found to have a brainstem infarct at the pontomedullary junction (arrow), as well as multifocal infarcts in the right cerebellar hemisphere and middle cerebellar peduncle (arrowheads) on diffusion-weighted MRI sequences. B: MR angiography demonstrated a large, partially thrombosed transitional-type aneurysm spanning the vertebrobasilar junction. C and D: Several days after inpatient admission, the patient became acutely unresponsive and was found to have a large subarachnoid hemorrhage on noncontrast-enhanced CT of the head.

  • View in gallery

    A and B: Kaplan-Meier curves demonstrating time to aneurysm-related death (A) and aneurysm rupture for all patients (B). C and D: Kaplan-Meier curves stratified by incidence of perforator infarction demonstrating time to death related to the aneurysm (C) and aneurysm rupture (D). The p values were obtained using the Wilcoxon rank-sum test. Figure is available in color online only.

  • 1

    Flemming KD, Wiebers DO, Brown RD Jr, et al. Prospective risk of hemorrhage in patients with vertebrobasilar nonsaccular intracranial aneurysm. J Neurosurg. 2004;101(1):8287.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 2

    Mangrum WI, Huston J III, Link MJ, et al. Enlarging vertebrobasilar nonsaccular intracranial aneurysms: frequency, predictors, and clinical outcome of growth. J Neurosurg. 2005;102(1):7279.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 3

    Nasr DM, Brinjikji W, Rouchaud A, et al. Imaging characteristics of growing and ruptured vertebrobasilar non-saccular and dolichoectatic aneurysms. Stroke. 2016;47(1):106112.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 4

    Nasr DM, Flemming KD, Lanzino G, et al. Natural history of vertebrobasilar dolichoectatic and fusiform aneurysms: a systematic review and meta-analysis. Cerebrovasc Dis. 2018;45(1-2):6877.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 5

    Xu DS, Levitt MR, Kalani MYS, et al. Dolichoectatic aneurysms of the vertebrobasilar system: clinical and radiographic factors that predict poor outcomes. J Neurosurg. 2018;128(2):560566.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 6

    Flemming KD, Wiebers DO, Brown RD Jr, et al. The natural history of radiographically defined vertebrobasilar nonsaccular intracranial aneurysms. Cerebrovasc Dis. 2005;20(4):270279.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 7

    Nakayama Y, Tanaka A, Kumate S, et al. Giant fusiform aneurysm of the basilar artery: consideration of its pathogenesis. Surg Neurol. 1999;51(2):140145.

  • 8

    Nakatomi H, Segawa H, Kurata A, et al. Clinicopathological study of intracranial fusiform and dolichoectatic aneurysms: insight on the mechanism of growth. Stroke. 2000;31(4):896900.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 9

    Brinjikji W, Chung B, Yong-Hong D, et al. Hemodynamic characteristics of stable and unstable vertebrobasilar dolichoectatic and fusiform aneurysms. J Neurointerv Surg. 2018;10(11):11021107.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 10

    Schievink WI, Parisi JE, Piepgras DG, Michels VV. Intracranial aneurysms in Marfan’s syndrome: an autopsy study. Neurosurgery. 1997;41(4):866871.

  • 11

    Flemming KD, Josephs K, Wijdicks EF. Enlarging vertebrobasilar dolichoectasia with subarachnoid hemorrhage heralded by recurrent ischemia. Case illustration. J Neurosurg. 2000;92(3):504.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 12

    Barletta EA, Ricci RL, Silva RDG, et al. Fusiform aneurysms: a review from its pathogenesis to treatment options. Surg Neurol Int. 2018;9:189.

  • 13

    Rouchaud A, Brinjikji W, Lanzino G, et al. Delayed hemorrhagic complications after flow diversion for intracranial aneurysms: a literature overview. Neuroradiology. 2016;58(2):171177.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 14

    Ertl L, Holtmannspötter M, Patzig M, et al. Use of flow-diverting devices in fusiform vertebrobasilar giant aneurysms: a report on periprocedural course and long-term follow-up. AJNR Am J Neuroradiol. 2014;35(7):13461352.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 15

    Siddiqui AH, Abla AA, Kan P, et al. Panacea or problem: flow diverters in the treatment of symptomatic large or giant fusiform vertebrobasilar aneurysms. J Neurosurg. 2012;116(6):12581266.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 16

    Meckel S, McAuliffe W, Fiorella D, et al. Endovascular treatment of complex aneurysms at the vertebrobasilar junction with flow-diverting stents: initial experience. Neurosurgery. 2013;73(3):386394.

    • Crossref
    • Search Google Scholar
    • Export Citation

Metrics

All Time Past Year Past 30 Days
Abstract Views 1275 1275 0
Full Text Views 629 629 153
PDF Downloads 465 465 16
EPUB Downloads 0 0 0