Ondra SL, Troupp H, George Ed, Schwab K. The natural history of symptomatic arteriovenous malformations of the brain: a 24-year follow-up assessment. J Neurosurg 73: 387-391, 1990 [PubMed]
| review | risk factors for haemorrhage | AVM calculator | disclaimer | ANS aneurysm calculator |
Ondra SL et a.l found that for brain arteriovenous malformations (AVM) the overall:
The study was based on a subset (166 of 260) of patients with symptomatic AVM. The allocation into the 'natural history group' does not appear to have been random; the authors have noted that large AVMs and AVMs in eloquent areas were more likely to have been allocated to the natural history group than the treatment group.
The study was undertaken in the pre-CT era (!942-1975) and haemorrhage was defined mainly on clinical grounds or on radiological (presumably palin X-ray and angiographic) evidence of mass effect.This has two implications:
1. Not all the patients with AVM might have been diagnosed as having AVM: patients with small bleeds from AVM might not have been referred to neurosurgeons. The authors state that during that period, there was a high threshold for patients to be considered for angiography.
2. Small bleeds might have not been diagnosed in those patients in the 'natural history group' during the study period.
The symptomatic patients in the natural history group were not homogenous: 114 (71 %) had presnted with haemorrhage; 38 had seizures and 8 had headaches or vague symptoms.
The study did not look at haemorrhage, morbidity and mortality rates associated with now known risk factors for such events. Risk factors for haemorrhage from AVM:
( Stapf C at al. Predictors of hemorrhage in patients with untreated brain arteriovenous malformation. Neurology 2006; 66: 1350-1355).
Staph C et al found that when an AVM had presented with no haemorrhage and if its locations was not deep and if there is no deep venous drainage the annual risk of haemorrhage was 0.9 %. However, when all these three risk factors are present the risk of haemorrhage was as high as 34.4%.
There is conflicting findings on whether the presence of aneurysm (intra-nidal or feeding artery) and size of avm has any significant effect on re-haemorrhage rates ( reviewed by Friedlander RM. Arteriovenous malformations of the brain. New England Journal of Medicine, 2007;356(26): 2704-12).
Based on the findings of Ondra et al and the multiplicative law of probability formula outlined by Kondziolka D et al [Kondziolka D, McLaughlin MR, Kestle JRW.Simple Risk Predictions for Arteriovenous Malformation Hemorrhage. Neurosurgery, 1995; Volume 37(5): 851-855] we have created the following program to estimate the life-time risk of haemorrhage, morbidity and mortality rates for arteriovenous malformation. Please note that the normal life-expectancy values used in this calculation are based on United Kingdom (UK) data.
This article is for only neurosurgeons, neuroradiologists and radiotherapist with particular interest in managing patients with arteriovenous malformation. The following review of the paper is not exhaustive. Therefore, please refer to the original paper and make your own critical appraisal of the study. The contents of the page have been created to stimulate discussion on the natural history of arteriovenous malformation and clinical decisions should not be made based on any part of this article.
E-mail: g_narenthiran@hotmail.com
Copyright 2007, G Narenthiran