RESUMO
OBJECTIVE: Cerebral cavernous malformations (CCMs) are vascular lesions with an overall risk of rupture from 2% to 6% per year, which is associated with significant morbidity and mortality. The diagnostic incidence is increasing, so it is of paramount importance to stratify patients based on their risk of rupture. Data in the literature seem to suggest that specific medications, particularly antithrombotic and cardiovascular agents, are associated with a reduced risk of bleeding. However, the effect of the patient coagulative status on the cumulative bleeding risk remains unclear. The aim of this study was to assess the impact of different radiological, clinical, and pharmacological factors on the bleeding risk of CCMs and to assess the predictive power of an already validated scale for general bleeding risk, the HAS-BLED (hypertension, abnormal renal/liver function, stroke, bleeding history or predisposition, labile international normalized ratio, elderly, drugs/alcohol concomitantly). METHOD: This was a multicenter retrospective observational study. The authors collected imaging, clinical status, and therapy data on patients with bleeding and nonbleeding CCMs. Univariate analysis and subsequent multivariate logistic regression were performed between the considered variables and bleeding or nonbleeding status to identify potential independent predictors of bleeding. RESULTS: The authors collected data on 257 patients (46.7% male, 25.3% with bleeding CCMs). Compared with patients with nonbleeding lesions, those with bleeding CCMs were younger, less frequently had hypertension, and less frequently required antiplatelet drugs and beta-blockers (all p < 0.05). Bleeding lesions, however, had significantly higher median volumes (1050 mm3 vs 523 mm3 , p < 0.001). On multivariate analyses, after adjusting for age, history of hypertension and diabetes, and use of antiplatelet drugs or beta-blockers, lesion volume ≥ 300 mm3 was the only significant predictor of bleeding (adjusted OR 3.11, 95% CI 1.09-8.86). When the diagnostic accuracy of different volume thresholds was explored, volume ≥ 300 mm3 showed a limited sensitivity (36.7%, 95% CI 24.6%-50.0%), but a high specificity 78.2% (95% CI 71.3%-84.2%), with an area under the curve of 0.57 (95% CI 0.51-0.64). CONCLUSIONS: This study supports previous findings that the CCM volume is the only factor influencing the bleeding risk. Antithrombotic agents and propranolol seem to have a protective role against the bleeding events. A high HAS-BLED score was not associated with an increased bleeding risk. Further studies are needed to confirm these results.
