Overview
Dural arteriovenous fistulas, often abbreviated as dAVFs, are abnormal connections between arteries and veins within the dura, the tough outer covering of the brain and spinal cord. Although they are relatively uncommon, they can have serious consequences when they drain blood in a way that raises venous pressure or places fragile veins at risk of rupture. The danger is not uniform across all dAVFs. In most cases, the pattern of venous drainage and the angiographic grade are the main drivers of hemorrhage risk.
This large international registry study from the CONDOR collaboration examined whether systemic cardiovascular factors, such as smoking, hypertension, diabetes, hyperlipidemia, atrial fibrillation, and antithrombotic medication use, also influence bleeding risk, treatment success, and early recovery. The findings suggest that venous anatomy remains the dominant predictor of hemorrhagic presentation, while most cardiovascular comorbidities do not independently change the odds of bleeding or short-term functional outcome.
Why dAVFs Matter Clinically
Dural arteriovenous fistulas are important because their clinical behavior can range from incidental and benign to rapidly progressive or life-threatening. Some patients develop pulsatile tinnitus, headache, visual symptoms, ocular congestion, cranial nerve dysfunction, seizures, or focal neurological deficits. Others present with intracranial hemorrhage.
The risk of hemorrhage is strongly tied to how blood exits the fistula. Lesions with cortical venous reflux, deep venous drainage, or other high-risk venous patterns are more likely to bleed. This is why angiographic classification systems are so important in clinical decision-making. They help determine whether a lesion should be observed, treated endovascularly, surgically, or with radiosurgery.
At the same time, many patients with dAVFs also live with common cardiovascular conditions. Since these disorders and their treatments can alter blood flow, clotting, and vessel integrity, it is reasonable to ask whether they modify the natural history of a fistula. Until now, that relationship has not been well defined.
Study Design and Population
This study used data from the Consortium for Dural Arteriovenous Fistula Outcomes Research registry, a large multicenter international database that retrospectively collected cases from 14 centers in 4 countries between 1990 and 2017. The analysis included 1,350 adults with intracranial dAVFs.
Investigators reviewed demographic information, cardiovascular comorbidities, antithrombotic therapy, angiographic anatomy, treatment approach, and follow-up outcomes. The main outcomes were:
1. Hemorrhagic presentation at diagnosis
2. Angiographic obliteration after treatment
3. Functional status at 90 days after treatment
The study used both univariable and multivariable logistic regression models. This approach is important because it helps distinguish whether a factor appears related to an outcome on its own versus whether the relationship disappears after accounting for stronger predictors such as lesion grade and drainage pattern.
Major Findings on Hemorrhagic Presentation
Among the 1,350 patients, 375 experienced hemorrhage, representing 27.8% of the cohort. The factors most strongly associated with hemorrhagic presentation were high-grade dAVF classification and male sex. These findings align with the broader understanding that lesion anatomy is the primary determinant of bleeding risk.
A notable finding was that antithrombotic therapy was associated with lower odds of hemorrhagic presentation. This observation should be interpreted cautiously. It does not necessarily mean that antithrombotic treatment protects against hemorrhage. More likely, this association reflects patient selection, treatment indication, or confounding factors that may differ across centers and clinical contexts. The study did not find an independent relationship between other cardiovascular risk factors and bleeding.
In practical terms, common systemic cardiovascular diseases did not appear to add much predictive value beyond the angiographic features of the fistula itself. This is clinically relevant because it suggests that risk stratification should remain focused mainly on venous anatomy and fistula grade rather than relying heavily on comorbidity burden.
Treatment and Angiographic Obliteration
Angiographic obliteration was achieved in 621 of 845 treated patients, or 73.5%. This is a favorable overall result and indicates that modern management can often eliminate the fistula when treatment is feasible.
The factors independently associated with successful obliteration included hemorrhagic onset, high dAVF grade, and surgical treatment. The association between hemorrhagic presentation and obliteration may reflect more aggressive treatment of higher-risk lesions or closer procedural attention when a patient has already bled. High-grade lesions were also more likely to be definitively treated, although they can be more complex and dangerous.
Surgical treatment showed a strong association with angiographic cure. In real-world practice, surgery is often reserved for selected lesions that are accessible or that remain residual after embolization, and this may contribute to its high success rate. Endovascular embolization remains a cornerstone of treatment for many dAVFs, but not every lesion is equally amenable to catheter-based closure. Radiosurgery may also have a role in selected cases, though it usually acts more slowly and is not ideal when immediate protection from hemorrhage is needed.
Smoking and embolization showed only nonsignificant trends after adjustment, meaning these variables were not robust independent predictors in the final statistical models. This again highlights that anatomy and treatment strategy are more important than most systemic cardiovascular factors when the goal is complete occlusion.
Functional Recovery at 90 Days
At 90 days after treatment, 934 patients, or 88.9%, were functionally independent. Functional status was measured using the modified Rankin Scale, a standard neurological outcome measure where lower scores indicate less disability.
The strongest predictor of 90-day functional outcome was the baseline modified Rankin Scale score, meaning how disabled the patient was before treatment. This is a common and important finding in neurology: patients who start with better function tend to recover better, especially when the lesion is treated before major neurological injury occurs.
Neither cardiovascular comorbidities nor treatment modality independently influenced short-term functional status after adjustment. This suggests that the patient’s neurological condition at presentation, rather than systemic vascular disease, largely determines early recovery. It also supports the value of timely diagnosis and treatment before hemorrhage or neurological decline creates lasting disability.
What These Results Mean for Patients and Clinicians
For clinicians, the study reinforces a simple but important message: when managing dAVFs, the anatomy of venous drainage should remain the primary focus of hemorrhage risk assessment. While cardiovascular health is crucial for overall medical care, common risk factors such as hypertension, diabetes, or hyperlipidemia did not independently predict bleeding in this cohort.
For patients, the results may be reassuring in one sense: having a cardiovascular risk factor does not appear to dramatically change dAVF hemorrhage risk once the lesion’s anatomy is considered. However, this should not reduce the importance of controlling blood pressure, quitting smoking, and managing other cardiovascular conditions, since those steps still matter for overall vascular and neurological health.
The finding that antithrombotic therapy was associated with lower odds of hemorrhagic presentation is interesting but should not be used to justify starting these medications solely to influence dAVF behavior. Decisions about antiplatelet or anticoagulant therapy should continue to be based on established indications such as atrial fibrillation, coronary disease, stroke prevention, or venous thromboembolism, while carefully balancing bleeding risk.
Clinical Context and Mechanistic Considerations
Why might venous anatomy matter more than systemic cardiovascular disease? dAVFs become dangerous when arterial blood is shunted into veins under abnormal pressure. If the venous outlet is restricted, redirected, or forced into cortical veins, the pressure can rise dramatically. This can lead to venous hypertension, vessel wall injury, congestion, and eventual hemorrhage.
Systemic factors like hypertension or smoking may theoretically worsen vascular stress, but in this registry they did not independently predict hemorrhage after accounting for the lesion itself. That suggests the local hemodynamic environment created by the fistula is the dominant determinant of clinical behavior. In other words, the fistula’s drainage pattern drives the risk more than the patient’s background cardiovascular profile.
This also helps explain why treatment effectiveness and outcomes depend so heavily on the lesion’s architecture and the chosen intervention. A technically successful repair that eliminates dangerous venous reflux can rapidly reduce risk, even in patients with multiple comorbidities.
Strengths of the Study
This work has several strengths. It is large, multicenter, and international, which improves generalizability compared with single-center series. The use of prospectively maintained databases helps improve data quality, even though the analysis itself was retrospective. The investigators also examined multiple clinically meaningful outcomes: hemorrhagic presentation, radiographic cure, and short-term functional recovery.
Another strength is the attempt to evaluate both anatomical and systemic factors in the same dataset. That design makes it possible to determine whether cardiovascular risk factors offer additional prognostic information beyond lesion anatomy.
Limitations to Keep in Mind
As with any retrospective registry study, there are limitations. Causality cannot be proven, and confounding by indication is always possible, especially for medication use such as antithrombotic therapy. Treatment choices were not randomized, so associations with cure rates may partly reflect case selection.
The cohort spans a long period, from 1990 to 2017, during which diagnostic imaging, embolic materials, surgical techniques, and perioperative care all evolved substantially. That means some treatment patterns may not reflect current practice exactly. In addition, the analysis focused on early functional outcome at 90 days, so longer-term recovery, recurrence, and quality-of-life outcomes were not fully assessed.
Finally, while the registry included many patients, certain cardiovascular subgroups may still have been too small to detect subtle effects. A lack of statistical significance does not always mean a factor has no biological relevance; it may simply mean the effect is modest or difficult to measure in this setting.
Bottom Line
In this large international cohort of patients with intracranial dural arteriovenous fistulas, hemorrhagic presentation was driven mainly by venous angioarchitecture and high-grade fistula classification. Male sex and antithrombotic therapy were also associated with hemorrhagic presentation, but most other cardiovascular comorbidities were not independently linked to bleeding. Angiographic cure was common, especially after surgery, and 90-day recovery depended most strongly on the patient’s baseline neurological status.
The practical takeaway is clear: when evaluating dAVFs, clinicians should prioritize the lesion’s venous anatomy and neurological presentation. Systemic cardiovascular disease remains important for overall care, but it appears to play a much smaller role than the fistula’s structure in determining hemorrhage risk and early outcomes.