The Epidemiology of Brain Arteriovenous Malformations

To the Editor: I read with great interest the critical review of arteriovenous malformations (AVMs) by Berman et al. (3), who argued that the previously reported estimates of the prevalence and incidence of AVMs are of questionable accuracy. They recalculated the overall detection rate as 1.1:100,000 person-years and the detection rate for symptomatic AVMs as 0.94:100,000 person-years. They also recalculated the prevalence rate as 10.3:100,000 people, which is much lower than the 500: to 600:100,000 prevalence rate obtained from autopsy data. Despite the stated objective of their study—to establish best estimates for both the incidence and prevalence of brain AVMs—and its importance, the authors did not consider the possibility of ethnic and geographical differences in the incidence and prevalence of AVMs. They also stated the ratio of aneurysms to AVMs using data from the Cooperative Study of Intracranial Aneurysms and Subarachnoid Hemorrhage (13). In that study, the incidence of AVMs was approximately one-seventh that of saccular aneurysms. The overall ratio of AVMs to aneurysms was 1:6.5, but the United States ratio was 1:5.3, and the ratio in England was 1:13.8 (13). It is important to mention that several reports in the literature suggest that, in some parts of the world, the ratios of the incidences of AVMs and aneurysms are very different from those reported in the United States and England, with AVMs in some areas being more common than aneurysms. A study of spontaneous subarachnoid hemorrhage in Qatar found more AVMs than aneurysms (12). The ratio of AVMs to aneurysms reported was 14:13, and the authors argued that the incidence of AVMs could be higher because only bleeding AVMs were considered in the study. In Kenya, AVMs were found to outnumber aneurysms (15). In Saudi Arabia, the AVM-to-aneurysm ratio was reported to be 10:11 when only Saudi citizens were considered and migrant workers were excluded (2). The AVM-to-aneurysm ratio was reported to be 1:1 among the Chinese (16) and 4:1 in Singapore (17). These observations may be due to a low incidence of aneurysms or to a high incidence of AVMs in the study populations. That genetic factors predispose some individuals to developing certain types of AVMs lends support to the existence of ethnic and geographical differences with regard to the incidence of AVMs. Intrafamilial AVMs are rare but well documented (1, 5, 18). Mexican Americans and Hispanics are known to be prone to developing cavernous angiomas (4, 6, 8). AVMs are known to complicate syndromes such as Sturge-Weber syndrome, von Hippel-Lindau syndrome, and Rendu-Osler-Weber syndrome. Cerebral AVMs also have been reported in association with autosomal dominant polycystic kidney disease (14), intracranial aneurysms (10), and intrafamilial aneurysms (7). Teleangiectasis has been documented in multiple generations of the same family, and some generations of the same family have manifested vascular malformations (9, 11). It seems that the incidence and prevalence of AVM are complex and might be influenced by geography and ethnicity. Therefore, the incidence and prevalence rates that Berman et al. (3) reported may not be universally applicable. No information was provided in their study about the racial or ethnic backgrounds of the patients they analyzed. Samuel C. Ohaegbulam