Classification of supratentorial arteriovenous malformations. A score system for evaluation of operability and surgical strategy based on an analysis of 66 cases.

A classification of arteriovenous malformations (AVM) is proposed, which is based on a retrospective analysis of the records and results of radical operation in 57 patients between 1983 and 1990. It represents the new developments and more recent technical facilities which influence operability of supratentorial AVMs. Predictability of outcome has been settled upon three groups of factors: anatomical, haemodynamical, and clinical. Anatomical factors are localisation and sectorisation of AVM, determination, caliber and straightening of feeding arteries. Haemodynamical factors are volume of AVM and vascular autoregulation, circulatory velocity of red blood cells in the main arteries of the neck and brain tissue cellular steal. Clinical factors are age, previous rupture of AVM, associated diseases and malformations of vital organs. Each of these factor groups has been divided into parameters to which a code number from 0 to 5 according to the severity of the considered parameter has been attributed. When a contraindication for radical surgery was clearly obvious, as, for example, an AVM with extension to the upper brain stem, number 10 was given. This grading has been done by a team of four persons (3 neurosurgeons including the senior author, and one biophysicist). By adding up all code numbers an Operability Score for a given patient is defined by the number of points, with a minimum of 3 and a maximum of 69. In cases with a score higher than 30 surgery is not advisable. A score between 21 and 30 indicates that always several staged operations are required, whilst only some of the patients with a score between 11 and 20 may require two stage operative treatment. The AVM in patients with a score under 10 can always be radically excised in a single stage operation. We have been able to demonstrate that the Operability Score allows a reliable prediction of outcome, thus giving indications and contraindications for surgery, and also for the surgical strategy. Moreover, we have explained why surgery had been refused in 9 additional cases during the same period. The causes of 8 fatalities out of 57 surgical cases are analyzed. These 57 cases represent a systematically explored series of the senior author; his experience is based upon 295 personal AVM cases (1958-1990).