Aneurysms associated with brain arteriovenous malformations (AVMs)1,2 and moyamoya disease3 represent a distinct pathology compared with saccular aneurysms, including higher rupture risk and spontaneous resolution after AVM treatment.4 Multiple case reports and small series have collectively reported an elevated risk of associated aneurysm in patients with intracranial dural arteriovenous fistula (dAVF) ranging from 13% to 21%.5,6 A limitation in these studies is the lack of a control group of dAVF patients without aneurysms. Furthermore, little is known regarding their natural history, rate of spontaneous regression or dAVF treatment-related regression, and ideal management for this population.
The Consortium for Dural Arteriovenous Fistula Outcomes Research (CONDOR) has collected information for more than 1000 patients with intracranial dAVF. Using this database, we aimed to define 1) the prevalence of aneurysms in dAVF patients; 2) subtypes of aneurysms in dAVF patients; 3) differences in baseline characteristics between dAVF patients with and without aneurysms; and 4) the treatment trends of intracranial aneurysms.
Methods
The infrastructure for CONDOR is described elsewhere in detail.39 Briefly, collaborating institutions obtained individual institutional review board approval to conduct a retrospective analysis of all dAVFs treated at each institution. Details were collected from the clinical record and imaging databases, de-identified, and shared via a third-party host institution. Discrepancies in coded patient information were addressed directly with each contributing center. Twelve centers contributed a total of 1077 patients with dAVF presenting between 1990 and 2017. Patients were excluded if the status of a coexisting aneurysm was missing or unknown.
Patients were dichotomized into two cohorts based on the presence of aneurysms: 1) dAVF patients with coexisting aneurysms (dAVF+ cohort); and 2) dAVF patients without coexisting aneurysms (dAVF-only cohort). Statistical analysis was performed by an independent statistician, using the Mann-Whitney U-test for continuous variables and Fisher’s exact test for categorical variables. Continuous variables are presented as the mean ± SD. No adjustments were made for multiple comparisons. Aneurysms were then further categorized as either intra- or extradural, and as either a dAVF flow-related aneurysm (FRA), defined as an aneurysm located on an arterial pedicle to the dAVF, or a completely dAVF non–flow-related aneurysm (NFRA). Aneurysms were also categorized (separately) based on their location in relation to the dAVF (i.e., ipsilateral or contralateral). Patients with bilateral aneurysms, midline aneurysm(s), or midline dAVFs were assigned to the “other” group. Predictors of associated aneurysms were identified using a multivariable logistic regression model with forward selection algorithm (p < 0.05 to enter), on a per-patient basis (i.e., each patient was enrolled in the model once, irrespective of the number of aneurysms). Any potential bias associated with missing covariate information was accounted for using multiple imputation by chained equations.
Results
Of the 1077 patients maintained in the CONDOR data set, 1043 were eligible for inclusion in this study based on the known presence or absence of coexisting intracranial aneurysms. The dAVF-only and dAVF+ cohorts comprised 978 (93.8%) and 65 (6.2%) patients, respectively. The dAVF+ cohort had a total of 96 aneurysms (mean 1.48 aneurysm/patient). The total mean clinical and radiographic follow-ups for the dAVF+ cohort were 3 ± 3.29 and 2.38 ± 3.01 years, respectively.
Table 1 compares patient demographics between dAVF-only and dAVF+ cohorts. The two cohorts were comparable, with the exception of higher rates of illicit drug use (p = 0.02) and smoking (p < 0.001) in the dAVF+ cohort. Table 2 shows all the aneurysms and the specific locations, categorized into FRA or NFRA, and intra- or extradural. Fifty-five patients (5.3%) had 84 NFRAs, most of which (84.5%) were intradural. There were no ruptured extradural aneurysms. Ten patients (1%) had 12 FRAs, half of which were intradural and half were extradural. The ICA was the most common location for aneurysms, with 33 (34.4%) being either “paraclinoid” (i.e., including the cavernous and clinoid segments, and carotid cave aneurysms) or supraclinoid (i.e., ophthalmic segment), 11 (11.5%) in the anterior communicating artery and A1 segment, and 11 in the middle cerebral artery bifurcation and M1 segment.
Demographics for patients with dAVF-only versus those with dAVF and aneurysm
| dAVF Only (n = 978) | dAVF+ (n = 65) | p Value | |
|---|---|---|---|
| Age | 59.5 ± 14.5 | 60.8 ± 11.3 | 0.596 |
| Female | 428/978 (43.8) | 35/65 (53.8) | 0.123 |
| CAD | 73/960 (7.6) | 6/63 (9.5) | 0.623 |
| DM | 121/962 (12.6) | 10/64 (15.6) | 0.442 |
| HTN | 392/962 (40.7) | 33/64 (51.6) | 0.115 |
| Cancer | 119/961 (12.4) | 11/63 (17.5) | 0.241 |
| Illicit drugs | 12/823 (1.5) | 3/52 (5.8) | 0.020 |
| Smoking | 275/781 (35.2) | 32/54 (59.3) | <0.001 |
CAD = coronary artery disease; DM = diabetes mellitus; HTN = hypertension. Values represent the number of patients/total number of patients with data (%) or mean ± SD unless stated otherwise. Boldface type indicates statistical significance; p values do not adjust for multiple comparisons.
Locations of all 96 aneurysms and the 16 ruptured aneurysms
| Location | Total | Ruptured | |
|---|---|---|---|
| Unrelated (n = 84) | |||
| Extradural (n = 13) | ICA (cavernous segment) | 10 | 0 |
| ICA (cervical, petrous, lacerum segments) | 2 | 0 | |
| ICA (clinoidal segment) | 1 | 0 | |
| Intradural (n = 71) | ICA (ophthalmic segment & “supraclinoid”) | 14 | 1 |
| M1/M2/MCA bifurcation | 11 | 0 | |
| A1/ACoA | 11 | 5 | |
| ICA (communicating segment) | 10 | 3 | |
| V4/VB junction | 6 | 2 | |
| ICA, “paraclinoid” | 5 | 0 | |
| Anterior choroidal artery | 4 | 1 | |
| Midbasilar | 2 | 1 | |
| ICA (terminus) | 2 | 0 | |
| A3, A4, & “pericallosal” | 2 | 0 | |
| ICA ("cave") | 2 | 0 | |
| PICA | 1 | 1 | |
| Basilar tip | 1 | 0 | |
| Posterior choroidal artery (FRA for tumor) | 1* | 0 | |
| FRA (n = 12) | |||
| Extradural (n = 6) | Ophthalmic artery | 3 | 0 |
| ICA (cavernous segment) | 1 | 0 | |
| Inferolateral trunk | 1 | 0 | |
| Posterior meningeal artery | 1 | 0 | |
| Intradural (n = 6) | PICA | 2 | 1 |
| AICA | 1 | 1 | |
| ACoA | 1 | 0 | |
| SCA | 1 | 0 | |
| PCA | 1 | 0 |
ACoA = anterior communicating artery; AICA = anterior inferior cerebellar artery; MCA = middle cerebral artery; PCA = posterior cerebral artery; PICA = posterior inferior cerebellar artery; SCA = superior cerebellar artery; VB = vertebrobasilar.
Coded as “multiple aneurysms” but counted as 1 for purposes of table.
Table 3 compares the presentations and dAVF angioarchitecture between dAVF-only and dAVF+ cohorts, and then between FRA and NFRA/dAVF-only cohorts. Patients in the dAVF-only cohort were more likely to present with dAVF symptoms (p < 0.001). Patients in the dAVF-only cohort also had a higher rate of dAVF rupture, although this did not achieve statistical significance. Patients in the dAVF+ cohort were more likely to have dAVFs located along the convexity or with superior sagittal sinus drainage (p = 0.002). Conversely, patients in the dAVF-only cohort were more likely to have dAVFs in the region of the cavernous sinus (i.e., cavernous-carotid fistula, p = 0.038). The most common dAVF location for both cohorts was the transverse-sigmoid junction. Pial arterial supply to the dAVF was more common in the dAVF+ cohort, whereas conventional dAVF supply from the external carotid artery (ECA) and dural branches of the internal carotid artery (ICA) was more commonly detected in the dAVF-only cohort. Similarly, conventional venous sinus drainage (i.e., to the transverse-sigmoid sinus junction) was more common in the dAVF cohort. Rates of cortical venous drainage and Borden type distributions were similar between the two cohorts, although the dAVF-only cohort had a higher rate of venous ectasia (p = 0.048). These relationships were all very similar when comparing FRA with NFRA/dAVF-only groups. Table 4 demonstrates the independent predictors of coexisting aneurysms in dAVF patients. Current smoking status, convexity location, and pial arterial supply were independent predictors of the presence of aneurysms, while venous ectasia was an independent predictor of the absence of aneurysms. Table 4 displays the details for the nonimputed multivariate model as well (complete-case analysis).
Presentation and angioarchitecture details for patients with dAVF only versus those with dAVF and aneurysm
| No. of Patients/Total No. of Patients w/ Data (%) | p Value | No. of Patients/Total No. of Patients w/ Data (%) | p Value | |||
|---|---|---|---|---|---|---|
| dAVF Only (n = 978) | dAVF+ (n = 65) | FRA (n = 10) | dAVF & Unrelated (n = 1033) | |||
| Symptomatic on presentation from dAVF | 811/977 (83.0) | 23/65 (35.4) | <0.001 | 5/10 (50) | 829/1032 (80.3) | 0.035 |
| Ruptured dAVF at presentation | 242/974 (24.8) | 12/65 (18.5) | 0.297 | 4/10 (40) | 250/1029 (24.3) | 0.27 |
| Location* | ||||||
| Transverse-sigmoid junction | 355/970 (36.6) | 19/64 (29.7) | 0.286 | 3/10 (30) | 373/1024 (36.4) | >0.99 |
| Convexity/SSS | 95/970 (9.8) | 15/64 (23.4) | 0.002 | 2/10 (20) | 108/1024 (10.5) | 0.289 |
| Tentorial | 146/970 (15.1) | 10/64 (15.6) | 0.858 | 2/10 (20) | 154/1024 (15.0) | 0.653 |
| Cavernous sinus | 115/970 (11.9) | 2/64 (3.1)†† | 0.038 | 1/10 (10) | 108/1024 (10.5) | >0.99 |
| Petrosal | 33/970 (3.4) | 5/64 (7.8) | 0.08 | 1/10 (10) | 37/1024 (3.6) | 0.313 |
| Anterior cranial fossa | 51/970 (5.3) | 7/64 (10.9) | 0.083 | 1/10 (10) | 57/1024 (5.6) | 0.44 |
| Arterial supply to dAVF† | ||||||
| Occipital artery | 553/957 (57.8) | 22/62 (35.5) | 0.001 | 4/9 (44.4) | 571/1010 (56.5) | 0.514 |
| Other ECA‡ | 364/958 (38.0) | 13/63 (20.6) | 0.007 | 3/10 (10) | 374/1011 (37.0) | 0.753 |
| Small ICA§ | 366/959 (38.2) | 17/63 (27.0) | 0.082 | 2/10 (20) | 381/1012 (37.6) | 0.336 |
| Pial artery¶ | 188/957 (19.6) | 20/64 (31.2) | 0.036 | 7/10 (70) | 201/1011 (19.9) | 0.001 |
| Venous drainage of dAVF** | ||||||
| Transverse-sigmoid junction | 415/957 (43.4) | 18/65 (27.7) | 0.014 | 2 (20) | 431/1012 (42.6) | 0.204 |
| CVD | 627/967 (64.8) | 42/65 (64.6) | >0.99 | 7/10 (70) | 662/1023 (64.7) | >0.99 |
| Venous ectasia | 321/927 (34.6) | 13/60 (21.7) | 0.048 | 2/10 (20) | 332/977 (34.0) | 0.509 |
| Borden type | 0.411 | 0.349 | ||||
| I | 327/966 (33.9) | 23/65 (35.4) | 3/10 (30) | 347/1021 (34) | ||
| II | 167/966 (17.3) | 4/65 (6.2) | 0 | 171 (16.7) | ||
| III | 472/966 (48.9) | 38/65 (58.5) | 7/10 (70) | 503 (49.3) | ||
SSS = superior sagittal sinus.
Boldface type indicates statistical significance; p values do not adjust for multiple comparisons.
Only the 3 most common locations and those with near significance (p < 0.1) are included in the table. Eleven locations were analyzed in total.
Nine arterial pedicles analyzed in total.
Includes any branches of the ECA other than middle meningeal, occipital, and ascending pharyngeal arteries.
Includes inferolateral trunk, meningohypophyseal trunk, small meningeal branches, and marginal tentorial branch.
Includes branches from the anterior, middle, and posterior cerebral arteries, and the superior, anterior inferior, and posterior inferior cerebellar arteries.
Only categories with statistical significance (p < 0.05) are included in table. Six venous sinuses were analyzed in total.
Both patients with associated aneurysm were indirect fistulas.
Univariate and multivariate predictors of associated aneurysms in dAVF patients using forward stepwise logistic regression starting with the set of best univariate predictors
| Variable | Multiple-Imputation Analysis | Complete-Case Analysis | ||||||
|---|---|---|---|---|---|---|---|---|
| Univariate | Multivariate Forward Selection | Univariate | Multivariate Forward Selection | |||||
| OR | p Value | OR | p Value | OR | p Value | OR | p Value | |
| Smoking | 0.002 | 0.003 | 0.001 | 0.002 | ||||
| Past (vs never) | 1.91 | 0.087 | 2.05 | 0.064 | 2.18 | 0.018 | 2.20 | 0.018 |
| Current (vs never) | 3.43 | 0.002 | 3.20 | 0.004 | 3.79 | <0.001 | 3.47 | 0.001 |
| Illicit drugs | 0.095 | 0.321 | — | 0.078 | 0.393 | |||
| Nonstimulants (vs no) | 2.51 | 0.178 | 4.51 | 0.024 | ||||
| Stimulants (vs no) | 1.23 | 0.763 | 0.00 | >0.99 | ||||
| Anterior cranial fossa location | 2.22 | 0.060 | 0.072 | 2.21 | 0.062 | 0.222 | ||
| Convexity/SSS location | 2.78 | 0.001 | 2.68 | 0.002 | 2.82 | 0.001 | 2.99 | 0.001 |
| Cavernous location | 0.24 | 0.048 | 0.185 | 0.24 | 0.049 | 0.202 | ||
| Petrosal location | 2.33 | 0.090 | 0.102 | 2.41 | 0.078 | 0.183 | ||
| Pial artery feeders | 1.83 | 0.033 | 1.97 | 0.021 | 1.86 | 0.028 | 0.264 | |
| Venous ectasia | 0.54 | 0.057 | 0.47 | 0.022 | 0.52 | 0.043 | 0.051 | |
| Cox-Snell R2 | 3.1% | 2.8% | ||||||
Boldface type indicates statistical significance.
Table 5 and Fig. 1 show details of the dAVF+ cohort. Multiple aneurysms were seen in 24.6% of the dAVF+ patients. Most (57.3%) aneurysms were small (< 7 mm), and only 1 (1.0%) was giant (≥ 25 mm). Most aneurysms (78.1%) were saccular. The majority of patients (88%) were diagnosed with an aneurysm and dAVF simultaneously. Seven patients (11%) were diagnosed with an aneurysm first and subsequently developed de novo dAVFs during the follow-up period. Of these 7 patients, 5 had adequate angiography (i.e., all vessels imaged) at the time of aneurysm diagnosis, suggesting that the dAVFs developed de novo. For these 5 patients, the average time to dAVF diagnosis was 34.3 ± 27 months. No patient developed de novo aneurysms. A significant proportion (40%) of patients in the dAVF+ cohort had aneurysm(s) ipsilateral to the side of the dAVF.
Overview of patients with a dAVF and aneurysm
| Category | Value |
|---|---|
| No. of patients (%; n = 65) | |
| No. of aneurysms | |
| 1 | 48 (73.8) |
| 2 | 9 (13.8) |
| 3 | 3 (4.6) |
| 4 | 1 (1.5) |
| 5 | 2 (3.1) |
| 6 | 1 (1.5) |
| Multiple | 1 (1.5) |
| No. of aneurysms (%; n = 96) | |
| Maximum aneurysm dimension, mm | |
| 0–3.9 | 33 (34.4) |
| 4–6.9 | 22 (22.9) |
| 7–12.9 | 25 (26.0) |
| 13–24.9 | 5 (5.2) |
| ≥25 | 1 (1.0) |
| Missing | 10 (10.4) |
| Morphology | |
| Saccular | 75 (78.1) |
| Fusiform &/or dissecting | 14 (14.6) |
| Missing | 7 (7.3) |
| Treatment | |
| None | 43 (44.8) |
| Endovascular | 35 (36.5) |
| Surgery | 15 (15.6) |
| Both | 1 (1.0) |
| Missing | 2 (2.1) |
| Follow-up | |
| Treated | |
| Stable | 35 (36.5) |
| No follow-up | 9 (9.4) |
| Recurrence/progression of residual/re-treatment | 7 (7.3) |
| Untreated | |
| No follow-up | 19 (19.8) |
| Stable | 16 (16.7) |
| Progression/rupture | 2 (2.1) |
| Spontaneous regression | 6 (6.2) |
| Missing | 2 (2.1) |
Characteristics of aneurysms found in patients with dAVFs. A: Order of diagnoses for the aneurysm and dAVF (for 65 patients in dAVF+ cohort). Note that there were no cases of de novo aneurysm development. B: Rupture status of aneurysms (includes all 96 aneurysms). C: Location of aneurysm(s) in relationship to the location of the dAVF (for 65 patients). D: Categorization of aneurysms (for all 96 aneurysms). Figure is available in color online only.
Table 5 and Fig. 1 show details of the dAVF+ cohort. Multiple aneurysms were seen in 24.6% of the dAVF+ patients. Most (57.3%) aneurysms were small (< 7 mm), and only 1 (1.0%) was giant (≥ 25 mm). Most aneurysms (78.1%) were saccular. The majority of patients (88%) were diagnosed with an aneurysm and dAVF simultaneously. Seven patients (11%) were diagnosed with an aneurysm first and subsequently developed de novo dAVFs during the follow-up period. Of these 7 patients, 5 had adequate angiography (i.e., all vessels imaged) at the time of aneurysm diagnosis, suggesting that the dAVFs developed de novo. For these 5 patients, the average time to dAVF diagnosis was 34.3 ± 27 months. No patient developed de novo aneurysms. A significant proportion (40%) of patients in the dAVF+ cohort had aneurysm(s) ipsilateral to the side of the dAVF.
Most patients presented with an unruptured aneurysm (80%), and 17% had a ruptured aneurysm at presentation, while 1% experienced rupture during follow-up (rupture status was missing in 2% of patients). Only 1 AICA and 1 PICA (2/16 = 12.5%) in the ruptured aneurysm cohort were FRAs, similar to the representation of FRAs within the unruptured aneurysm cohort (10/78 = 12.8%).
Aneurysms were managed conservatively in 44.8%, treated with endovascular modalities in 36.5%, surgically in 15.6%, and with both surgery and endovascular techniques in 1.0% of the cases. Of the treated aneurysms (51), 68.6% remained stable and 13.7% had recurrence or re-treatments at the time of final follow-up. Of the untreated aneurysms (43), most (44.2%) had no follow-up or were stable during follow-up (37.2%). One patient (1.5%) had a fusiform basilar trunk aneurysm that ruptured during follow-up. Six (6.3%) of the aneurysms (in 4 patients) resolved spontaneously; 4 of these were FRAs that resolved after dAVF treatment, 1 was an aneurysm associated with a brain tumor feeding artery and resolved after removal of the tumor, and 1 was a small (2 mm) NFRA aneurysm that also resolved after dAVF treatment. Figure 2 illustrates a case example of an FRA that resolved after dAVF treatment.
Example of spontaneous regression of FRA in a treated ruptured Borden type III tentorial dAVF. A and B: Initial axial (A) and sagittal (B) CT angiograms showing an aneurysm (arrow in A, asterisk in B) on a branch vessel of left cavernous. C: Cerebral angiography, lateral view, left ICA injection, confirming aneurysm (arrow) as well as tentorial dAVF (asterisk). D: Three-dimensional reconstructed angiogram showing meningohypophyseal trunk (arrowhead) and inferolateral trunk arterial feeders to tentorial dAVF (asterisk), the inferolateral trunk harboring an FRA (arrow). E: Posttreatment intraoperative angiogram (which included 3 sequential stages of transarterial Onyx embolization to the middle meningeal artery, meningohypophyseal trunk, and occipital arteries, with a final craniotomy for clip ligation), showing resolution of previous dAVF but persistent sluggish filling of the FRA. F: Follow-up angiography at 9 months confirms obliteration of dAVF with subsequent regression of the FRA.
Discussion
It is now well recognized that patients with intracranial dAVFs can also harbor a coexisting underlying intracranial aneurysm. Cagnazzo et al. recently published a systematic review of all intracranial aneurysms identified in patients with dAVF,7 pooling data from 26 studies (3 case series5,8,9 and 23 case reports6,10–31), which included 43 patients with 62 aneurysms, with 23 (37.1%) being FRAs. The included patients derived from an international cohort, including patients from the United States, Japan, China, Italy, Germany, Canada, and beyond. Interestingly, only 1 patient was from a CONDOR contributing center.20 However, the included patient reported was not an adult and thus was not included in the current CONDOR database. Therefore, we present an entirely new large database that is distinct from this recent review, representing a different perspective that includes a “control group” with dAVF patients without concomitant aneurysms for comparisons of features and characteristics.
Importantly, the risk of patients with intracranial dAVFs harboring a concomitant NFRA intracranial aneurysm appears similar (5.3%) when compared with the general population (quoted rates range from 1% to 5%32,33). A small subset of dAVF+ (1%) do harbor FRAs, half of which are intradural. Overall, these rates sum to 6.3%, which is lower than the previously reported rate of 13%8 or 21%5 in dAVF patients. Dural AVF patients with and without concomitant aneurysms shared similar risk factors.34,35 The prevalence of aneurysm in dAVF patients appears to be quite different from that of associated aneurysm in AVM patients, which has been estimated to be 20.2%.36 This likely relates to higher rates of intracranial arterial feeders in AVMs compared with dAVFs. The proportion of dAVF+ patients presenting with aneurysmal hemorrhage is different from that of aneurysm rupture in patients with AVM; 17% of dAVF+ presented with aneurysm rupture, 18.5% of dAVF+ presented with fistula rupture, and 24.8% of the dAVF-only cohort presented with fistula rupture, while 64% of patients with AVM-associated aneurysms presented with rupture (compared with 50% without aneurysm), 49% of which (approximately 32% total) were related to the aneurysm.36 Thus, while the presence of aneurysm in dAVF patients does increase the overall risk that the patient will present with hemorrhage of any kind, the overall rate of hemorrhage at presentation is much lower than that of AVM-associated aneurysms.
Contrary to prior reports, angioarchitectural features of the dAVF+ cohort were distinctly different from those of the dAVF cohort. The transverse-sigmoid junction was the most common dAVF location for both cohorts. However, there was a higher proportion of fistulas located along the convexity or near the superior sagittal sinus and a lower proportion of CCFs in the dAVF+ cohort. Convexity/superior sagittal sinus location emerged as an independent predictor of coexisting aneurysms. While typical arterial contributors to dAVF (ECA and dural ICA branches) were more commonly identified in the dAVF-only cohort, pial intracranial feeders were more common in the dAVF+ cohort. There did not seem to be differences in venous drainage between the two cohorts, except that the dAVF-only cohort more commonly had transverse-sigmoid junction venous drainage (consistent with this cohort having a more typical fistula location/appearance). Finally, while rates of CVD and Borden grade distributions were similar between the two groups, it is interesting to note that dAVF+ patients less often had venous ectasia, which held true after adjusting for other covariates. This warrants further histological and microbiological analyses, as there may be different underlying endothelial dysfunction in the arterial and venous systems that contributes to preferential aneurysm formation.
FRAs
Only 12 (12.5%) of 96 aneurysms in the study were FRAs, which was substantially lower than the rate of 71.4% reported in AVM-associated aneurysms.36 The FRAs in this study can be categorized into those that are located on 1) meningeal feeding arteries (e.g., the inferolateral trunk, posterior meningeal arteries), 2) pial feeding arteries that otherwise supply parenchyma (such as the ophthalmic segment of the ICA when the ophthalmic artery supplies the dAVF), or 3) other intracranial branches not supplying dura or brain (such as the intraorbital ophthalmic artery). Accordingly, FRAs can be grouped into intra- and extradural locations, with 50% in each category in this series. When triaging FRAs, one must acknowledge the less aggressive natural history associated with extradural aneurysms, as they do not carry a risk of subarachnoid hemorrhage. Importantly, FRAs had a similar rate of rupture at presentation compared to NFRAs (2 [16.7%] of 12 FRAs were ruptured, 15 [17.9%] of 84 NFRAs were ruptured or ruptured during follow-up).
Regression
Similar to distal FRAs in the AVM literature,37 we found that 4 FRAs spontaneously regressed after treatment of the dAVF, 1 after treatment of a brain tumor (this aneurysm was associated with arterial supply to tumor, not dAVF), and 1 resolved spontaneously. This likely stems from a similar mechanism; after the underlying pathology (i.e., arteriovenous shunt) is eliminated, previously prominent arterial feeders likely have a reduction in blood flow, promoting stasis and eventually thrombosis within the aneurysm.38 Figure 2 shows serial angiography for an example of an FRA that spontaneously regressed 9 months after dAVF treatment.
Limitations
This study, despite representing the largest of its type, has limitations. The retrospective design implies that there may be potential confounding bias. For example, it is not clear why venous ectasia seems to predict dAVF over dAVF+, even in multivariable modeling. This may be due to an intimate relationship between venous ectasia and another variable. Second, the data were collected by multiple people at multiple centers. The architects of CONDOR did include a validation process for all chart reviewers prior to data collection to decrease the potential spectrum of variability, but this remains a concern. Third, follow-up varied significantly from patient to patient and from institution to institution. While the medical records of all patients with dAVF and aneurysm were investigated more thoroughly after identification, this analysis was performed retrospectively and not in a prospective longitudinal fashion with specific attention to aneurysm or dAVF development. Finally, while we suppose that most patients here had simultaneous diagnosis of dAVF and aneurysm (based on retrospective chart analysis and available angiographic data), we are limited in the imaging performed before presentation, so the precise timing of dAVF and/or aneurysm development is not totally clear. Future studies might aim to follow a group of patients with either intracranial aneurysm or dAVF (not both) prospectively, to delineate the precise rates and timing of de novo lesion development.
Conclusions
The prevalence of intracranial aneurysms in dAVF patients is 6.2%, most of which are unrelated to the dAVF. A minority of concomitant aneurysms are located on arteries supplying the dAVF (defined as FRAs). Aneurysms in dAVF patients seem to rupture at lower rates than those associated with AVMs. Compared with dAVF patients without aneurysm, those with dAVFs with aneurysms harbor unique angioarchitecture features. Spontaneous regression of aneurysms has been observed with dAVF treatment.
Acknowledgments
We thank Jason Barber, MS, for his assistance with statistical analysis, and Sharon Durfy, PhD, for assistance with manuscript preparation.
Appendix
CONDOR Collaborators
Washington University School of Medicine: Gregory J. Zipfel, MD; Akash P. Kansagra, MD, MS; Ridhima Guniganti, MD; Jay F. Piccirillo, MD; Hari Raman, MD; and Kim Lipsey.
Mayo Clinic: Giuseppe Lanzino, MD; Enrico Giordan, MD; Waleed Brinjikji, MD; Roanna Vine, RN; Harry J. Cloft, MD; David F. Kallmes, MD; Bruce E. Pollock, MD; and Michael J. Link, MD.
University of Virginia Health System: Jason Sheehan, MD, PhD; Ching-Jen Chen, MD; Mohana Rao Patibandla, MCh; Dale Ding, MD; Thomas Buell, MD; and Gabriella Paisan, MD.
University of Washington: Louis J. Kim, MD, MBA; Michael R. Levitt, MD; Isaac Josh Abecassis, MD; R. Michael Meyer IV, MD; and Cory Kelly.
University of Southampton: Diederik Bulters, FRCS(SN); Andrew Durnford, MA, MSc, FRCS; Jonathan Duffill, MBChB; Adam Ditchfield, MBBS; John Millar, MBBS; and Jason Macdonald, MBBS.
University of Florida: W. Christopher Fox, MD; Adam J. Polifka, MD; Dimitri Laurent, MD; Brian Hoh, MD; Jessica Smith, MSN, RN; and Ashley Lockerman, RN.
University of Pittsburgh: Bradley A. Gross, MD; L. Dade Lunsford, MD; and Brian T. Jankowitz, MD.
University of Iowa Hospitals and Clinics: Minako Hayakawa, MD, PhD; Colin P. Derdeyn, MD; Edgar A. Samaniego, MD; Santiago Ortega Gutierrez, MD, MS; David Hasan, MD; Jorge A. Roa, MD; James Rossen, MD; Waldo Guerrero, MD; and Allen McGruder.
University of Illinois at Chicago: Sepideh Amin-Hanjani, MD; Ali Alaraj, MD; Amanda Kwasnicki, MD; Fady T. Charbel, MD; Victor A. Aletich, MS, MD (posthumous); and Linda Rose-Finnell.
University of Groningen, University Medical Center Groningen: J. Marc C. van Dijk, MD, PhD; and Adriaan R. E. Potgieser, MD, PhD.
University of Miami: Robert M. Starke, MD, MSc; Eric C. Peterson, MD; Dileep R. Yavagal, MD; Samir Sur, MD; and Stephanie H. Chen, MD.
Tokushima University: Junichiro Satomi, MD, PhD; Yoshiteru Tada, MD, PhD; Yasuhisa Kanematsu, MD, PhD; Nobuaki Yamamoto, MD, PhD; Tomoya Kinouchi, MD, PhD; Masaaki Korai, MD, PhD; Izumi Yamaguchi, MD, PhD; and Yuki Yamamoto, MD.
University of California, San Francisco: Adib Abla, MD; Ethan Winkler, MD, PhD; Ryan R. L. Phelps, BA; Michael Lawton, MD; and Martin Rutkowski, MD.
Brigham and Women’s Hospital: Rose Du, MD, PhD; Pui Man Rosalind Lai, MD; M. Ali Aziz-Sultan, MD; Nirav Patel, MD; and Kai U. Frerichs, MD.
Disclosures
Dr. Levitt reports non–study-related clinical or research effort from the NINDS, NIH, AHA, Stryker, Medtronic, and Philips Volcano; consultant fees from Medtronic; equity interest in Synchron, Cerebrotech, and eLoupes; and is an advisor for Metis Innovative. Dr. Gross reports consultant fees from MicroVention and Medtronic. Dr. Lanzino reports being a consultant for Nested Knowledge and Superior Medical Editors. Dr. Starke reports funding support from NREF, Joe Niekro Foundation, Brain Aneurysm Foundation, Bee Foundation, and the NIH; and consultant fees from Penumbra, Abbott, Medtronic, and Cerenovus. Dr. Alaraj reports funding support from the NIH, and consultant fees from Cerenovus and Siemens. Dr. Samaniego reports serving as a proctor and consultant for MicroVention. Dr. Derdeyn reports being a consultant for Penumbra, Genae, and NoNo; and receiving support of non–study-related clinical or research effort from Siemens Healthineers. Dr. Kansagra reports consultant fees from MicroVention, Medtronic, and Penumbra. Dr. Kim reports stock ownership in SPI Surgical.
Author Contributions
Conception and design: Kim, Abecassis, Meyer, Levitt. Acquisition of data: Abecassis, Meyer. Analysis and interpretation of data: Abecassis, Meyer, Levitt. Drafting the article: Abecassis. Critically revising the article: Abecassis, Meyer, Levitt. Reviewed submitted version of manuscript: all authors. Approved the final version of the manuscript on behalf of all authors: Kim. Study supervision: Kim.
Supplemental Information
References
- 1↑
D’Aliberti G, Talamonti G, Cenzato M, et al. Arterial and venous aneurysms associated with arteriovenous malformations. World Neurosurg. 2015;83(2):188–196.
- 2↑
Yu JL, Yang S, Luo Q, et al. Endovascular treatment of intracranial ruptured aneurysms associated with arteriovenous malformations: a clinical analysis of 14 hemorrhagic cases. Interv Neuroradiol. 2011;17(1):78–86.
- 3↑
Amin-Hanjani S, Goodin S, Charbel FT, Alaraj A. Resolution of bilateral moyamoya associated collateral vessel aneurysms: Rationale for endovascular versus surgical intervention. Surg Neurol Int. 2014;5(4)(suppl 4):S155–S160.
- 4↑
He L, Gao J, Thomas AJ, et al. Disappearance of a ruptured distal flow-related aneurysm after arteriovenous malformation nidal embolization. World Neurosurg. 2015;84(5):1496.e1–1496.e6.
- 5↑
Gross BA, Ropper AE, Du R. Cerebral dural arteriovenous fistulas and aneurysms. Neurosurg Focus. 2012;32(5):E2.
- 6↑
Onu DO, Hunn AW, Harle RA. A rare association of cerebral dural arteriovenous fistula with venous aneurysm and contralateral flow-related middle cerebral artery aneurysm. BMJ Case Rep. 2013;2013 bcr2013200764.
- 7↑
Cagnazzo F, Peluso A, Vannozzi R, et al. Arterial aneurysms associated with intracranial dural arteriovenous fistulas: epidemiology, natural history, and management. A systematic review. Neurosurg Rev. 2019;42(2):277–285.
- 8↑
Suzuki S, Tanaka R, Miyasaka Y, et al. Dural arteriovenous malformations associated with cerebral aneurysms. J Clin Neurosci. 2000;7(suppl 1):36–38.
- 9↑
Kikuchi K, Kowada M. Anterior fossa dural arteriovenous malformation supplied by bilateral ethmoidal arteries. Surg Neurol. 1994;41(1):56–64.
- 10
Andersson T, Kihlström L, Söderman M. Regression of a flow-related ophthalmic artery aneurysm after treatment of a frontal DAVS. A case report. Interv Neuroradiol. 2004;10(3):265–268.
- 11
Chen Z, Zhu G, Feng H, et al. Dural arteriovenous fistula of the anterior cranial fossa associated with a ruptured ophthalmic aneurysm: case report and review of the literature. Surg Neurol. 2008;69(3):318–321.
- 12
Gács G, Viñuela F, Fox AJ, Drake CG. Peripheral aneurysms of the cerebellar arteries. Review of 16 cases. J Neurosurg. 1983;58(1):63–68.
- 13
Gilard V, Curey S, Tollard E, Proust F. Coincidental vascular anomalies at the foramen magnum: dural arteriovenous fistula and high flow aneurysm on perimedullary fistula. Neurochirurgie. 2013;59(6):210–213.
- 14
Ishikawa T, Houkin K, Tokuda K, et al. Development of anterior cranial fossa dural arteriovenous malformation following head trauma. Case report. J Neurosurg. 1997;86(2):291–293.
- 15
Kaech D, de Tribolet N, Lasjaunias P. Anterior inferior cerebellar artery aneurysm, carotid bifurcation aneurysm, and dural arteriovenous malformation of the tentorium in the same patient. Neurosurgery. 1987;21(4):575–582.
- 16
Kan P, Stevens EA, Warner J, Couldwell WT. Resolution of an anterior-inferior cerebellar artery feeding aneurysm with the treatment of a transverse-sigmoid dural arteriovenous fistula. Skull Base. 2007;17(3):205–210.
- 17
Kawaguchi S, Sakaki T, Okuno S, et al. Peripheral ophthalmic artery aneurysm. Report of two cases. J Neurosurg. 2001;94(5):822–825.
- 18
Kirsch M, Henkes H. A ruptured intraorbital ophthalmic artery aneurysm, associated with a dural arteriovenous fistula: combined transarterial and transvenous endovascular treatment. Minim Invasive Neurosurg. 2011;54(3):128–131.
- 19
Kleinschmidt A, Sullivan TJ, Mitchell K. Intraorbital ophthalmic artery aneurysms. Clin Exp Ophthalmol. 2004;32(1):112–114.
- 20↑
Ko A, Filardi T, Giussani C, et al. An intracranial aneurysm and dural arteriovenous fistula in a newborn. Pediatr Neurosurg. 2010;46(6):450–456.
- 21
Kohyama S, Yamane F, Ishihara H, et al. Rupture of an aneurysm of the superior cerebellar artery feeding a dural arteriovenous fistula. J Stroke Cerebrovasc Dis. 2015;24(5):e105–e107.
- 22
Li M, Lin N, Wu J, Liang J, He W. Multiple intracranial aneurysms associated with multiple dural arteriovenous fistulas and cerebral arteriovenous malformation. World Neurosurg. 2012;77(2):398.e11–398.e15.
- 23
Martin NA, King WA, Wilson CB, et al. Management of dural arteriovenous malformations of the anterior cranial fossa. J Neurosurg. 1990;72(5):692–697.
- 24
Meneghelli P, Pasqualin A, Lanterna LA, et al. Surgical treatment of anterior cranial fossa dural arterio-venous fistulas (DAVFs): a two-centre experience. Acta Neurochir (Wien). 2017;159(5):823–830.
- 25
Murai Y, Yamashita Y, Ikeda Y, et al. Ruptured aneurysm of the orbitofrontal artery associated with dural arteriovenous malformation in the anterior cranial fossa—case report. Neurol Med Chir (Tokyo). 1999;39(2):157–160.
- 26
Muro K, Adel JG, Gottardi-Littell NR, et al. True aneurysm on the posterior meningeal artery associated with a dural arteriovenous fistula: case report. Neurosurgery. 2010;67(3):E876–E877.
- 27
Preul M, Tampieri D, Leblanc R. Giant aneurysm of the distal anterior cerebral artery: associated with an anterior communicating artery aneurysm and a dural arteriovenous fistula. Surg Neurol. 1992;38(5):347–352.
- 28
Reinard K, Basheer A, Pabaney A, et al. Spontaneous resolution of a flow-related ophthalmic-segment aneurysm after treatment of anterior cranial fossa dural arteriovenous fistula. Surg Neurol Int. 2014;5(14)(suppl 14):S512–S515.
- 29
Rumboldt Z, Beros V, Klanfar Z. Multiple cerebral aneurysms and a dural arteriovenous fistula in a patient with polyarteritis nodosa. Case illustration. J Neurosurg. 2003;98(2):434.
- 30
Sanchis JF, Orozco M, Cabanes J. Spontaneous extradural haematomas. J Neurol Neurosurg Psychiatry. 1975;38(6):577–580.
- 31
Sato K, Shimizu T, Fukuhara T, Namba Y. Ruptured anterior communicating artery aneurysm associated with anterior cranial fossa dural arteriovenous fistula—case report. Neurol Med Chir (Tokyo). 2011;51(1):40–44.
- 32↑
Brown RD Jr, Broderick JP. Unruptured intracranial aneurysms: epidemiology, natural history, management options, and familial screening. Lancet Neurol. 2014;13(4):393–404.
- 33↑
Salahuddin H, Siddiqui NS, Castonguay AC, et al. Recent trends in electively treated unruptured intracranial aneurysms. J Stroke Cerebrovasc Dis. 2019;28(7):2011–2017.
- 34↑
Müller TB, Vik A, Romundstad PR, Sandvei MS. Risk factors for unruptured intracranial aneurysms and subarachnoid hemorrhage in a prospective population-based study. Stroke. 2019;50(10):2952–2955.
- 35↑
Qureshi AI, Suarez JI, Parekh PD, et al. Risk factors for multiple intracranial aneurysms. Neurosurgery. 1998;43(1):22–27.
- 36↑
Cagnazzo F, Brinjikji W, Lanzino G. Arterial aneurysms associated with arteriovenous malformations of the brain: classification, incidence, risk of hemorrhage, and treatment-a systematic review. Acta Neurochir (Wien). 2016;158(11):2095–2104.
- 37↑
Redekop G, TerBrugge K, Montanera W, Willinsky R. Arterial aneurysms associated with cerebral arteriovenous malformations: classification, incidence, and risk of hemorrhage. J Neurosurg. 1998;89(4):539–546.
- 38↑
Martínez-Pérez R, Tsimpas A, Ruiz Á, et al. Spontaneous regression of a true intracanalicular fusiform ophthalmic artery aneurysm after endovascular treatment of an associated dural arteriovenous fistula. World Neurosurg. 2018;119:362–365.
- 39↑
Guniganti R, Giordan E, Chen CJ, et al. Consortium for Dural Arteriovenous Fistula Outcomes Research (CONDOR): rationale, design, and initial characterization of patient cohort. J Neurosurg. Published online September 10, 2021. doi:10.3171/2021.1.JNS202790
