Treatment of Spinal Cord Vascular Malformations by Surgical Excision

J. Neurosurg. / Volume 30 / April, 1969

H. KRAYENBOHL, M. G. YA~ARGIL, M.D., AND H. G. McCLINTOCK* Section v] Neurosurgery, Kantonsspital, The University o] Ziirich, Ziirich, Switzerland

ECENT developments have now made called attention to an increase in symptoms direct surgical attack the treatment during pregnancy with subsidence after of choice for spinal cord vascular delivery, z~ Newman has stated that he be- malformations. We are reporting 17 cases lieves the increase in symptoms in such cases treated with surgical excision, the last 11 of may be due to "venous congestion" from the which were operated on under the operating distended uterus and interestingly suggests microscope. the possibility of some "hormonal factor act- There is much confusion in the literature ing on the vessel walls. ''22 Although none of concerning the histological nomenclature our cases was a child, several authors have used to describe varieties of spinal vascular reported the occurrence in children and even malformations. This confusion is partly the in infants?, ~, 10,22,23 result of the lack of opportunity for ade- quate microscopic study of the entire lesion. Clinical Picture We prefer to follow the classification of History. The clinical history is usually one Bergstrand, et al.2 who divided these malfor- of three types. There can be 1) a slow mations into: 1) angioma cavernosum, 2) progression of neurological symptoms and angioma racemosum, and 3) angioreticu- signs, 2) progression followed with regres- loma. Some vascular malformations will sion or a stationary period, or 3) a sudden show characteristics of more than one group, apoplectic onset. making classification difficult. From the sur- Neurological Examination. This may gical standpoint the extent of the lesion and range from an entirely normal neurological the amount, if any, of intramedullary in- examination up to that of a complete trans- volvement are of much greater importance verse myelitis. It is interesting that a rather than the specific type of malformation pres- high percentage of patients have some im- ent. The 17 cases we are reporting are pri- pairment in bladder control. In many cases, marily of the angioma racemose type, some the complaints and neurological findings are largely venous, others predominantly arterial. such as to suggest the diagnosis of herniated It has been shown that spinal cord vascu- disc, poliomyelitis, spinal cord tumor, multi- lar malformations are much more likely to ple sclerosis, or even hysteria. The correct occur in certain areas of the cord? ~ In only diagnosis is much more apt to be made in one of our cases was the lesion confined to those cases having a sudden onset, bloody the cervical area. The remainder were lo- spinal fluid, and neurological findings indi- cated in the thoracic or lumbosacral region, cating some sort of myelopathy. Matthews re- with the latter site the most common loca- ports the occasional presence of a spinal tion. Often the malformation was so exten- bruit as a diagnostic sign of spinal vascular sive as to involve the entire thoracic and malformation. 19 lumbar cord with even occasional extension Spinal Fluid. Nearly all writers on the into the cauda equina. Only two of our cases subject agree that in the presence of spinal were females, although an exacerbation of vascular malformations the spinal fluid pro- symptoms in women during menstruation tein is almost always increased. There may has been reported, and several authors have be a slight increase in the spinal fluid white cell count. The Queckenstedt test is usually Received for publication May 20, 1968. normal except in the occasional case having * Present address: 1855 Gaylord Street, Denver, a "spinal block" from a very large malforma- Colorado 80206. tion. 427 428 H. Krayenbiihl, M. G. Yasargil and H. G. McClintock Radiological Findings raphy and the cause is obscure, contrast Plain X-Ray Films of the Spine. These will medium should then be instilled from the seldom show anything to indicate the vascu- opposite end of the vertebral column in lar nature of the lesion. Rarely there may be order to delineate the extent of the block. widening of the pedicles at the site of a large Occasionally this will demonstrate the cause malformation. Even more rarely an asso- of the block to be a vascular malformation. ciated vascular involvement of the spinal Most all vascular malformations of the cord vertebrae may by seen, but this is more often are located on the posterior rather than the present in association with an extradural ventral surface. It is for this reason that vascular lesion than with an intradural le- these lesions are better demonstrated myelo- sion. Peculiarly, a high percentage of pa- graphically with the patient lying in the su- tients with spinal cord vascular malforma- pine position. In searching for such vascular tions show considerable osteoarthritic verte- lesions it is essential that at least 10 cc of bral changes. contrast medium be employed. It would Myelography. The very typical myelo- seem that the value of air myelography is graphic picture of the "worm-like" mass of very limited in diagnosing this particular dis- vessels is well known (Fig. 1).8,2s Whenever ease. a complete block is demonstrated on myelog- Spinal Angiography. Angiographic dem- onstration of the abnormal vasculature of the spinal cord represents a great advance in not only the diagnostic area but also in plan- ning the surgical attack. In recent years several authors have shown how well these vascular malformations can be demonstrated with one of several angiographic techniques. 2,7,~-15 Retrograde femoral catheterization, brachial catheterization, and vertebral angiography are being utilized. The use of multiple casette changers and subtrac- tion technique is essential. We wonder if, with further perfection in angiographic technique and the use of stereoscopy in subtrac- tion, it may not even become possible to reveal the presence and extent of any signifi- cant intramedullary extension of these vascular lesions? Our own experience with spinal angiography is limited, but we well recognize the value of the procedure (Fig. 2). Detailed preoperative spinal angiographic studies offer the best information as to the extent of the vascular malformation, and postopera- tive angiographic studies will reveal the "completeness" of removal. In contrast to myelography, angiography can give precise information as to the nature and location of the "feeder" arteries, which can be of great help in the surgical removal. 2 Treatment In years past the treatment of these vascular spinal malformations has, for the most Fro. 1. Typical picture on Pantopaque myelo- part, consisted of decompressive laminec- gram of the "worm-like" mass of vessels, taken tomy and deep x-ray therapy. Decompres- with patient in the supine position. sion is of little or no value and then only in