- Email: [email protected]
The three phases of vasospasm
The Three Phases of Vasospasm John P. Kapp, M.D., Ph.D., Walter R. Neill, M.D., Charles L. Neill, M.D., Lucien R. Hodges, M.D., and Robert R. Smith, M.D. We propose the theory that prolonged cerebral vasospasm involves three phases: (1) the initial muscular contraction of the arterial wall; (2) a secondary injury to the artery that consists of endothelial desquamation with adherence of platelets to the denuded internal elastic lamina and mural thrombus formation; and (3) the repair process, which is the proliferative endarteropathy that has been observed in autopsy specimens. Cerebral ischemia can be the end product of any of these three conditions. We have postulated a possible subcycle in the overall scheme by which adherence of platelets to the denuded internal elastic lamina of the artery provide a continuously replenishing supply of spasmogenic chemical factors to the mural receptors and stimulate prolonged contraction of the muscular layer. We propose that this cycle may be interrupted by the administration of heparin. To test this hypothesis, the records of 112 consecutive patients who received systemic heparin in conjunction with carotid ligation were compared with the results of carotid ligation reported in the Cooperative Study, in which heparin was not used. The incidence of ischemic complications in the group of patients receiving heparin was 6%, as compared to 23% in the control group, with a concomitant reduction in mortality from 16% to 10%. The incidence of recurrent subarachnoid hemorrhage was slightly less in the patients receiving heparin than in patients in the Cooperative Study. We conclude that the data provide support for our hypothesis of the mechanism of prolonged cerebral vasospasm and cerebral ischemia associated with subarachnoid hemorrhage, and that systemic heparin may be used with relative safety in patients in whom the aneurysm is protected by partial carotid ligation. Kapp JP, Neill WR, Neill CL, HodgesLR, Smith RR: The three phases of vasospasrn. SurgNeurol 18:40-45, 1982 The literature concerning cerebral vasospasm has grown to be voluminous over the past twenty years. At least sixteen From the Departmentof Neurosurgery,Universityof Mississippi Medical Center, Jackson, MS. Address reprint requests to Dr. John P. Kapp, Department of Neurosurgery,Universityof MississippiMedicalCenter, 2500North State St., Jackson, MS 39216. Key words: carotid ligation; cerebral aneurysm; cerebral artery; cerebral embolism; cerebral ischemia; vasospasm; heparin; subarachnoid hemorrhage. 40
substances normally present in circulating blood have been reported to cause vasospasm [ 11]. The list of remedies that have been reported to prevent or reverse vasospasm occupies almost seven pages in tabular form [12]. Yet the clinical problem remains unsolved and several observations remain unexplained. Some patients whose angiograms demonstrate severe arterial narrowing have no neurological deficit. Others who have much less severe arterial narrowing develop catastrophic deficits. Not uncommonly, spasm may persist for three weeks. Yet virtually all of the known spasmogenic compounds are inactivated rapidly by various enzymes and survive only briefly in free form in biological fluids. Logic tends to oppose the theory that spasm is a prolonged or delayed response to a single stimulus. To explain the observed clinical facts, we must invoke a cycle that, once initiated, feeds on itself, that is, a self-perpetuating cycle. Observations published within the last ten years allow us to view vasospasm as consisting of three separate components, or phases, that follow each other in sequence. These three phases, the stimulus, the injury, and the repair, fall into an obvious self-perpetuating cycle. Examination of this cycle may allow us to understand the phenomenon of vasospasm as seen in patients with subarachnoid hemorrhage. Interruption of this cycle, which we believe can be done with available drugs, should prevent the development of ischemic complications in patients with ruptured aneurysms. Hypothesis The first phase, the stimulus, is the initial vasoconstriction that occurs when blood escapes from the lumen of the vessel into the subarachnoid space. This is probably a nonspecific phenomenon and is mediated by mechanical factors and several chemical substances that become spasmogenic when they are released from their normal site of sequestration within formed blood elements and gain access to the exterior surface of the vessel. Studies involving fractionation of blood into plasma, platelets, and cells suggest that most of the spasmogenic activity can be found in the platelet fraction [6]. Presumably these spasmogenic substances, serotonin, prostaglandin F~, thromboxane A2, and at least one additional platelet factor [7], are inactive until platelet clumping or lysis occurs, and until they gain access to the nonendothelialized surface of the vessel. The second phase is morphological and has been de-
0090-3019/82/070040-06501.25 © 1982 by Little, Brown and Company (Inc.)
Kapp et al: Vasospasm 41
Fig. 1. Scanning electron micrograph of the luminal surface of a normal cat basilar artery. The internal elastic lamina is covered by a reneer of endothelial cells in a tight pavement pattern. (x2,900.)
scribed by investigators both in the United States and Japan [5, 8]. When examined with the scanning electron microscope, the luminal surface of a normal cerebral artery is covered by a veneer of endothelial cells that fit together to provide complete coverage of the subintimal connective tissue or internal elastic lamina (Fig. 1). As the smooth muscle cells within the walls contract in response to subarachnoid blood, the internal elastic lamina corrugates and the endothelial cells protrude into the lumen of the vessel, finally becoming detached from the vessel wall (Fig. 2). Platelets and other formed blood elements in a fibrin mesh then adhere to the subirLtimal connective tissue and form mural thrombi (Fig. 3). At this stage, injury to the vessel is complete. The third phase is the constrictive endarteropathy or proliferative vasculopathy [2-4, 9] (Fig. 4). This is the repair process and probably encompasses both the organization of mural thrombi and the re-endothelialization of the luminal surface. Microscopically, necrosis of the medial layer and thickeninlg of the subintimal stratum of the artery are seen. The endothelial layer appears intact in later
stages, although small round cells may appear clumped to the endothelial cell. Thickening in the subendothelial strata seems to consist of tight connective tissue, presumably collagen, and at least two types of cells. The cellular elements may be smooth muscle cells or fibroblasts. Similar morphological changes have been seen following puncture, cold and heat injuries, and isolation of vessels between ligatures. In subarachnoid hemorrhage, the structural changes are much more widespread and, in addition to causing reduction in luminal diameter through proliferation of wall tissue, there is also reduction due to the loss of compliance [2]. Thus, the vessel does not expand elastically at normal intraluminal pressures. This proposed sequence of events following subarachnoid hemorrhage is summarized in Figure 5. At the center of the cycle of events is ischemia of the brain, which is the end product of any of the three phases. In addition, there may be a subcycle that may explain the phenomenon of pro. longed vasospasm. If one assumes that the various spasmogenic chemicals bound in platelets are effective stimulators of smooth muscle contraction on the luminal surface of the vessel as well as on the adventitial surface, but are normally bound in platelets and separated from the muscle cell receptors by the intimal cells, then aggregations of platelets adherent to the denuded vessel wall would be
42 SurgicalNeurology Vol 18 No 1 July 1982
expected to maintain a high and replenished concentration of chemical stimulators in intimate contact with the smooth muscle (or neural) receptors in the media of the vessel. Three pathways lead to brain ischemia: the initial vasoconstriction, the embolization of platelet aggregates from the denuded vessel wall, and the structural stenosis of the vessel as the platelet thrombi organize and proliferative endarteropathy develops. In time, the endothelium re-forms, the healing process is completed, and the artery becomes normal. Where can the cycle be stopped? Most research up to this time has concentrated on the search for a drug that acts at the first phase, the initial chemically-induced vasoconstriction. If our hypothesis is correct, a drug that acts on the second phase (platelet aggregation and adherence to the vessel wall that has been denuded of endothelium) may effectively interrupt the cycle and should dramatically reduce the incidence of ischemic complications in patients with subarachnoid hemorrhage. Materials and Methods Several years ago, one of us (W. N.) noted the high incidence of ischemic complications in patients with ruptured intracranial aneurysms treated by carotid ligation. On the
Fig. 2. Scanning electron micrograph of a cat basilar artery 30 minutes after spasm had been induced by tearing a small branch of the artery. The internal elastic lamina is corrugated. Endothelial cells promale into the lumen and some are detached. Platelets and other formed blood elements in a fibrin mesh adhere to the denuded internal elastic lamina. (×3,500.)
assumption that ischemic complications were caused by emboli from the carotid clamp, heparin was given during the time the clamp was being closed. The patients received systemic heparin, which was begun the day after the carotid clamp was applied and partially closed, and continued until the day after the clamp was entirely closed. The usual heparin dosage was 5,000 units administered subcutaneously, with the dosage subsequently adjusted to keep the Lee-White clotting time between 17 and 30 minutes (at 2 hours after the heparin dose). The records of 112 consecutive patients thus treated were reviewed regarding mortality, incidence of permanent ischemic complications, and incidence of recurrent subarachnoid or intracerebral hemorrhage during hospitalization, as determined by lumbar puncture or at postmortem examination. This group included 69 patients with aneurysms of the internal carotid artery, 19 patients with aneurysms of the middle cerebral artery, and 24 patients with aneurysms of the anterior
Fig. 3. Scanning electron micrograph of a cat basilar artery 2 hours after spasm had been induced by tearing a small branch of the artery. The internal elastic lamina is corrugated and covered by a layer of platelets. A mural thrombus h~lsformed at the area of maximum constriction. (×300.)
Fig. 4. Photomicrograph of autopsy specimen of cerebral artery, from a patient with subarachnoid hemorrhage, showing typical features of proliferative vasculopathy. (H&E; × 176.) 43
44
SurgicalNeurology Vol 18 No 1 July 1982
NORMALARTERY
S
RE-ENOOTHELIALIZATION
PROLIFERATIVEENDARTEBOPATHY WITH STRUCTURAL VASCULARSTENOSIS
ligation in the Cooperative Study [10], the early mortality was 16% (24 patients), and the incidence of recurrent subChemica~fstimulutionarachnoid hemorrhage was 8% (12 patients). Sixty-seven Muscle ceOsor percent (8 cases) of the patients with recurrent subdventitial receptors arachnoid hemorrhages in the group not receiving heparin VASOCONSTRICTION died. The major difference observed between this group and curs is the incidence of ischemic complications, which was 23% (34 patients). ruptures Mechanical stimulation
-•aAneurysm ENDOTHELIAL DESQUAMATION I
Discussion We use this data to support our synthesis of the pathogenesis of cerebral vasospasm as a phenomenon with three phases. We are aware that there are other plausible INTERNALELASTICLAMINA explanations. Emboli may really originate at the site of the ORGANIZATIONOF PLATELET carotid clamp. Heparin may exert its beneficial effect by FIBRINMURALTHROMBI improving the microcirculation through a process unrelated Fig. 5. Diagram of proposed mechanisms for cerebral ischemia in pato the changes that have been seen in the walls of the larger tients with cerebral vasospasm and subcycle of proposed events that arteries. Nevertheless, the use of heparin does appear to remay explain prolonged vasospasm. duce the incidence of ischemic complications in patients with subarachnoid hemorrhage. In addition, the use of communicating-anterior cerebral artery complex. Twenty- heparin while the aneurysm is protected by partial carotid nine patients were grade I, 42 grade II, 37 grade III, 4 occlusion does not appear to increase the incidence of regrade IV, and 0 grade V (Botterell classification) [1]. In 58 bleeding from the aneurysm. Although all recurrent patients the carotid clamp was applied less than one week hemorrhages in the heparinized patients were fatal, the after the hemorrhage and in 54 more than one week after overall early mortality rate in the patients in the Cooperathe hemorrhage. Sixty-three patients were below 50 years tive Study was greater than the mortality rate in our paand 49 were above 50 years of age. There were 56 males tients who received heparin. Thus it appears that heparin and 56 females. may be used with relative safety in patients who have A control series of patients with subarachnoid hemor- undergone partial carotid ligation to protect the aneurysm. rhage whose intracranial aneurysms were treated by carotid Using the treatment concept of partial carotid occlusion ligation without the use of heparin was derived from pub- plus heparin anticoagulation, additional improvement in lished data in the Cooperative Study of Aneurysmal Sub- outcome may be achievable. Many of our patients received arachnoid Hemorrhage [10]. This other series included 125 heparin only 3 or 4 days of the first two weeks following patients in whom carotid ligation was accomplished, 21 hemorrhage. Early diagnosis and earlier institution of patients in whom carotid occlusion was attempted but not treatment with a longer period of heparin administration tolerated (data from text), and 5 patients who died during may contribute. A more dynamic form of control of presclosure of the carotid artery. Deaths, ischemic complica- sure in the carotid artery may also reduce the incidence of tions, and recurrent hemorrhages occurring within 27 days both ischemic complications and recurrent hemorrhages. A of carotid occlusion were summarized. One early death from clamp or ligature reduces the head of pressure on the an unrelated cause was included, although the exact time aneurysm but does not provide protection from increases or interval from hemorrhage to death could not be ascertained decreases in systemic blood pressure. A microprocessorfrom the report. Only those patients with recurrent sub- controlled carotid artery-occluding device that will arachnoid hemorrhages occurring during carotid occlusion maintain a constant pressure on the vascular bed containand those patients who did not tolerate carotid occlusion ing the aneurysm in the face of wide swings in systemic blood pressure is being developed and hopefully will be were considered for comparison with our series. ready for clinical testing in the near future. Results We are not advocating a return to carotid ligation for the In our 112 patients the overall mortality was 10% (11 pa- definitive treatment of intracranial aneurysms. It does aptients); the incidence of permanent ischemic complications pear, however, that heparin in combination with localized was 6% (7 patients); and the incidence of recurrent intra- hypotension in the artery bed harboring a ruptured cranial hemorrhage was 5% (6 patients). All recurrent aneurysm may be a rational, effective, and safe form of hemorrhages were fatal in the patients receiving heparin. acute management of patients with ruptured intracranial For comparison, in the 151 patients treated by carotid aneurysms. We plan a prospective study in patients receiv-
/ %''LATELETAD,ERENCETO
PLATELET• FiRm EMBOLI
Kapp et al: Vasospasm
ing heparin to examine the incidence of angiographically demonstrated arterial spasm, labile hypertension, ischemic complications, and recurrent hemorrhage during the first two weeks after subarachnoid hemorrhage. Finally, we suggest that the most productive avenue for basic research in cerebral vasospasm may be directed n o t at the blood vessel wall, but at the platelet.
We express our gratitude to Drs. Otis R. Blaumanisand Patricia A. Grady, Department of Neurology, University of Maryland, Baltimore, for supplying the scanning electron micrographs (Figs. 1, 2, and 3).
References 1. Botterell EH, Lougheed WM, Scott JW, Vandewater SL: Hypothermia, and interruption of carotid, or carotid and vertebral circulation, in the surgical management of intracranial aneurysms. J Neurosurg 13:1-42, 1956 2. Clower BR, Haining JL, Smith RR: Pathophysiological changes in the cerebral artery after subarachnoid hemorrhage, in Wilkins RH (ed): Cerebral Arterial Spasm. Baltimore: Williams & Wilkins, 1979, pp 124-131 3. Conway LW, McDonald LW: Structural changes of the intradural arteries following subarachnoid hemorrhage. J Neurosurg 37:715723, 1972
45
4. Cromptom MR: The pathogenesis of cerebral infarction following the rupture of cerebral berry aneurysms. Brain 87:491-510, 1964 5. Grady PA, Blaumanis OR, Nelson ER: Morphology and flow dynamics of focal arterial constriction, in Wilkins RH (ed): Cerebral Arterial Spasm. Baltimore: Williams & Wilkins, 1979, pp 107-112 6. Kapp J, Mahaley MS Jr, Odom GL: Cerebral arterial spasm. Part 2. Experimental evaluation of mechanical and humoral factors in pathogenesis. J Neurosurg 29:339-349, 1968 7. Kapp J, Mahaley MS Jr, Odom GL: Cerebral arterial spasm. Part 3. Partial purification and characterization of a spasmogenicsubstance in feline platelets. J Neurosurg 29:350-355, 1968 8. Ohta T, Kajikawa H, Funatsu N, Yoshikawa Y, Someda K: Cerebral vasospasm and its relaxation responses to vasodilators: pathological study of severe prolonged vasospasm, in Wilkins RH (ed): Cerebral Arterial Spasm. Baltimore: Williams & Wilkins, 1979, pp 132-138 9. Peerless SJ, KassellNF, Kiyohide K, Hunter IG: Cerebral vasospasm: acute proliferative vasculopathy? II. Morphology, in Wilkins RH (ed): Cerebral Arterial Spasm. Baltimore: Williams & Wilkins, 1979, pp 88-96 10. Perret GE, Nibbelink DW: Randomized treatment study. Carotid ligation, in Sails AL, Nibbelink DW, Torner JC (eds): Aneurysmal Subarachnoid Hemorrhage. Baltimore: Urban & Schwarzenberg, 1981, pp 121-143 11. White RP: Overview of the pharmacology of vasospasm, in Wilkins RH (ed): Cerebral Arterial Spasm. Baltimore: Williams & Wilkins, 1979, pp 228-236 12. Wilkins RH: Attempted prevention or treatment of intracranial arterial spasm. A survey, in Wilkins RH (ed): Cerebral Arterial Spasm. Baltimore: Williams & Wilkins, 1979, pp 542-555
Editorial: The Future T h e future does n o t just: happen. For the most part it is something that someone: has made happen. If one is to mold the future, one must see the problems and t h e n ponder their solution. Sir Isaac N e w t o n said, "If I have seen further it is by standing upon the shoulders of giants." Every neurological surgeon stands upon the shoulders of giants. If any one of t h e m does n o t see the problems that must confront t h e m everyday, it is because they will n o t look. If they will do n o t h i n g about these problems, it is because they will not t h i n k and act. A l m o s t one h u n d r e d ,¢ears ago, Sir Victor Horsley was asked by one of his colleagues why he did n o t stop doing his operations since the results were so poor. He replied wisely, "If I do n o t continue those who come after me will do no better." It is n o t enough, however, just to continue to make the same mistakes. O n e must endeavor to correct them. Look at one aspect of neurological s u r g e r y - - b r a i n tumors. Over the years many improvements have been made. W e operate upon them with greater certainty and with more ease. Some of' our patients are now cured while others live longer and in a better condition than formerly.
O n the other hand, so far as cerebral gliomas are concerned, we are doing no better today t h a n Sir W i l l i a m Macewen and Victor Horsley did a hundred years ago. Practically all patients with cerebral gliomas still die of their tumors. W e are still not curing them. Cerebral gliomas do not pose an impossible problem. There are no malignant gliomas, and they do n o t metastasize. T h e y remain localized. If such a tumor were removed completely, it would be cured. Cerebral gliomas are the most c o m m o n of all brain tumors. T h e y are our problem. A hundred years is too long to be still faced with a problem without solving it. It is a problem for neurological surgeons since it is unlikely that these tumors will be cured by anything except complete removal. W e must face up to this problem. It is a many-faceted one, and its solution will require much thought and considerable effort by many people. It is time we got off our backsides and did something about it.
Paul C. Bucy, M . D . , Editor
substances normally present in circulating blood have been reported to cause vasospasm [ 11]. The list of remedies that have been reported to prevent or reverse vasospasm occupies almost seven pages in tabular form [12]. Yet the clinical problem remains unsolved and several observations remain unexplained. Some patients whose angiograms demonstrate severe arterial narrowing have no neurological deficit. Others who have much less severe arterial narrowing develop catastrophic deficits. Not uncommonly, spasm may persist for three weeks. Yet virtually all of the known spasmogenic compounds are inactivated rapidly by various enzymes and survive only briefly in free form in biological fluids. Logic tends to oppose the theory that spasm is a prolonged or delayed response to a single stimulus. To explain the observed clinical facts, we must invoke a cycle that, once initiated, feeds on itself, that is, a self-perpetuating cycle. Observations published within the last ten years allow us to view vasospasm as consisting of three separate components, or phases, that follow each other in sequence. These three phases, the stimulus, the injury, and the repair, fall into an obvious self-perpetuating cycle. Examination of this cycle may allow us to understand the phenomenon of vasospasm as seen in patients with subarachnoid hemorrhage. Interruption of this cycle, which we believe can be done with available drugs, should prevent the development of ischemic complications in patients with ruptured aneurysms. Hypothesis The first phase, the stimulus, is the initial vasoconstriction that occurs when blood escapes from the lumen of the vessel into the subarachnoid space. This is probably a nonspecific phenomenon and is mediated by mechanical factors and several chemical substances that become spasmogenic when they are released from their normal site of sequestration within formed blood elements and gain access to the exterior surface of the vessel. Studies involving fractionation of blood into plasma, platelets, and cells suggest that most of the spasmogenic activity can be found in the platelet fraction [6]. Presumably these spasmogenic substances, serotonin, prostaglandin F~, thromboxane A2, and at least one additional platelet factor [7], are inactive until platelet clumping or lysis occurs, and until they gain access to the nonendothelialized surface of the vessel. The second phase is morphological and has been de-
0090-3019/82/070040-06501.25 © 1982 by Little, Brown and Company (Inc.)
Kapp et al: Vasospasm 41
Fig. 1. Scanning electron micrograph of the luminal surface of a normal cat basilar artery. The internal elastic lamina is covered by a reneer of endothelial cells in a tight pavement pattern. (x2,900.)
scribed by investigators both in the United States and Japan [5, 8]. When examined with the scanning electron microscope, the luminal surface of a normal cerebral artery is covered by a veneer of endothelial cells that fit together to provide complete coverage of the subintimal connective tissue or internal elastic lamina (Fig. 1). As the smooth muscle cells within the walls contract in response to subarachnoid blood, the internal elastic lamina corrugates and the endothelial cells protrude into the lumen of the vessel, finally becoming detached from the vessel wall (Fig. 2). Platelets and other formed blood elements in a fibrin mesh then adhere to the subirLtimal connective tissue and form mural thrombi (Fig. 3). At this stage, injury to the vessel is complete. The third phase is the constrictive endarteropathy or proliferative vasculopathy [2-4, 9] (Fig. 4). This is the repair process and probably encompasses both the organization of mural thrombi and the re-endothelialization of the luminal surface. Microscopically, necrosis of the medial layer and thickeninlg of the subintimal stratum of the artery are seen. The endothelial layer appears intact in later
stages, although small round cells may appear clumped to the endothelial cell. Thickening in the subendothelial strata seems to consist of tight connective tissue, presumably collagen, and at least two types of cells. The cellular elements may be smooth muscle cells or fibroblasts. Similar morphological changes have been seen following puncture, cold and heat injuries, and isolation of vessels between ligatures. In subarachnoid hemorrhage, the structural changes are much more widespread and, in addition to causing reduction in luminal diameter through proliferation of wall tissue, there is also reduction due to the loss of compliance [2]. Thus, the vessel does not expand elastically at normal intraluminal pressures. This proposed sequence of events following subarachnoid hemorrhage is summarized in Figure 5. At the center of the cycle of events is ischemia of the brain, which is the end product of any of the three phases. In addition, there may be a subcycle that may explain the phenomenon of pro. longed vasospasm. If one assumes that the various spasmogenic chemicals bound in platelets are effective stimulators of smooth muscle contraction on the luminal surface of the vessel as well as on the adventitial surface, but are normally bound in platelets and separated from the muscle cell receptors by the intimal cells, then aggregations of platelets adherent to the denuded vessel wall would be
42 SurgicalNeurology Vol 18 No 1 July 1982
expected to maintain a high and replenished concentration of chemical stimulators in intimate contact with the smooth muscle (or neural) receptors in the media of the vessel. Three pathways lead to brain ischemia: the initial vasoconstriction, the embolization of platelet aggregates from the denuded vessel wall, and the structural stenosis of the vessel as the platelet thrombi organize and proliferative endarteropathy develops. In time, the endothelium re-forms, the healing process is completed, and the artery becomes normal. Where can the cycle be stopped? Most research up to this time has concentrated on the search for a drug that acts at the first phase, the initial chemically-induced vasoconstriction. If our hypothesis is correct, a drug that acts on the second phase (platelet aggregation and adherence to the vessel wall that has been denuded of endothelium) may effectively interrupt the cycle and should dramatically reduce the incidence of ischemic complications in patients with subarachnoid hemorrhage. Materials and Methods Several years ago, one of us (W. N.) noted the high incidence of ischemic complications in patients with ruptured intracranial aneurysms treated by carotid ligation. On the
Fig. 2. Scanning electron micrograph of a cat basilar artery 30 minutes after spasm had been induced by tearing a small branch of the artery. The internal elastic lamina is corrugated. Endothelial cells promale into the lumen and some are detached. Platelets and other formed blood elements in a fibrin mesh adhere to the denuded internal elastic lamina. (×3,500.)
assumption that ischemic complications were caused by emboli from the carotid clamp, heparin was given during the time the clamp was being closed. The patients received systemic heparin, which was begun the day after the carotid clamp was applied and partially closed, and continued until the day after the clamp was entirely closed. The usual heparin dosage was 5,000 units administered subcutaneously, with the dosage subsequently adjusted to keep the Lee-White clotting time between 17 and 30 minutes (at 2 hours after the heparin dose). The records of 112 consecutive patients thus treated were reviewed regarding mortality, incidence of permanent ischemic complications, and incidence of recurrent subarachnoid or intracerebral hemorrhage during hospitalization, as determined by lumbar puncture or at postmortem examination. This group included 69 patients with aneurysms of the internal carotid artery, 19 patients with aneurysms of the middle cerebral artery, and 24 patients with aneurysms of the anterior
Fig. 3. Scanning electron micrograph of a cat basilar artery 2 hours after spasm had been induced by tearing a small branch of the artery. The internal elastic lamina is corrugated and covered by a layer of platelets. A mural thrombus h~lsformed at the area of maximum constriction. (×300.)
Fig. 4. Photomicrograph of autopsy specimen of cerebral artery, from a patient with subarachnoid hemorrhage, showing typical features of proliferative vasculopathy. (H&E; × 176.) 43
44
SurgicalNeurology Vol 18 No 1 July 1982
NORMALARTERY
S
RE-ENOOTHELIALIZATION
PROLIFERATIVEENDARTEBOPATHY WITH STRUCTURAL VASCULARSTENOSIS
ligation in the Cooperative Study [10], the early mortality was 16% (24 patients), and the incidence of recurrent subChemica~fstimulutionarachnoid hemorrhage was 8% (12 patients). Sixty-seven Muscle ceOsor percent (8 cases) of the patients with recurrent subdventitial receptors arachnoid hemorrhages in the group not receiving heparin VASOCONSTRICTION died. The major difference observed between this group and curs is the incidence of ischemic complications, which was 23% (34 patients). ruptures Mechanical stimulation
-•aAneurysm ENDOTHELIAL DESQUAMATION I
Discussion We use this data to support our synthesis of the pathogenesis of cerebral vasospasm as a phenomenon with three phases. We are aware that there are other plausible INTERNALELASTICLAMINA explanations. Emboli may really originate at the site of the ORGANIZATIONOF PLATELET carotid clamp. Heparin may exert its beneficial effect by FIBRINMURALTHROMBI improving the microcirculation through a process unrelated Fig. 5. Diagram of proposed mechanisms for cerebral ischemia in pato the changes that have been seen in the walls of the larger tients with cerebral vasospasm and subcycle of proposed events that arteries. Nevertheless, the use of heparin does appear to remay explain prolonged vasospasm. duce the incidence of ischemic complications in patients with subarachnoid hemorrhage. In addition, the use of communicating-anterior cerebral artery complex. Twenty- heparin while the aneurysm is protected by partial carotid nine patients were grade I, 42 grade II, 37 grade III, 4 occlusion does not appear to increase the incidence of regrade IV, and 0 grade V (Botterell classification) [1]. In 58 bleeding from the aneurysm. Although all recurrent patients the carotid clamp was applied less than one week hemorrhages in the heparinized patients were fatal, the after the hemorrhage and in 54 more than one week after overall early mortality rate in the patients in the Cooperathe hemorrhage. Sixty-three patients were below 50 years tive Study was greater than the mortality rate in our paand 49 were above 50 years of age. There were 56 males tients who received heparin. Thus it appears that heparin and 56 females. may be used with relative safety in patients who have A control series of patients with subarachnoid hemor- undergone partial carotid ligation to protect the aneurysm. rhage whose intracranial aneurysms were treated by carotid Using the treatment concept of partial carotid occlusion ligation without the use of heparin was derived from pub- plus heparin anticoagulation, additional improvement in lished data in the Cooperative Study of Aneurysmal Sub- outcome may be achievable. Many of our patients received arachnoid Hemorrhage [10]. This other series included 125 heparin only 3 or 4 days of the first two weeks following patients in whom carotid ligation was accomplished, 21 hemorrhage. Early diagnosis and earlier institution of patients in whom carotid occlusion was attempted but not treatment with a longer period of heparin administration tolerated (data from text), and 5 patients who died during may contribute. A more dynamic form of control of presclosure of the carotid artery. Deaths, ischemic complica- sure in the carotid artery may also reduce the incidence of tions, and recurrent hemorrhages occurring within 27 days both ischemic complications and recurrent hemorrhages. A of carotid occlusion were summarized. One early death from clamp or ligature reduces the head of pressure on the an unrelated cause was included, although the exact time aneurysm but does not provide protection from increases or interval from hemorrhage to death could not be ascertained decreases in systemic blood pressure. A microprocessorfrom the report. Only those patients with recurrent sub- controlled carotid artery-occluding device that will arachnoid hemorrhages occurring during carotid occlusion maintain a constant pressure on the vascular bed containand those patients who did not tolerate carotid occlusion ing the aneurysm in the face of wide swings in systemic blood pressure is being developed and hopefully will be were considered for comparison with our series. ready for clinical testing in the near future. Results We are not advocating a return to carotid ligation for the In our 112 patients the overall mortality was 10% (11 pa- definitive treatment of intracranial aneurysms. It does aptients); the incidence of permanent ischemic complications pear, however, that heparin in combination with localized was 6% (7 patients); and the incidence of recurrent intra- hypotension in the artery bed harboring a ruptured cranial hemorrhage was 5% (6 patients). All recurrent aneurysm may be a rational, effective, and safe form of hemorrhages were fatal in the patients receiving heparin. acute management of patients with ruptured intracranial For comparison, in the 151 patients treated by carotid aneurysms. We plan a prospective study in patients receiv-
/ %''LATELETAD,ERENCETO
PLATELET• FiRm EMBOLI
Kapp et al: Vasospasm
ing heparin to examine the incidence of angiographically demonstrated arterial spasm, labile hypertension, ischemic complications, and recurrent hemorrhage during the first two weeks after subarachnoid hemorrhage. Finally, we suggest that the most productive avenue for basic research in cerebral vasospasm may be directed n o t at the blood vessel wall, but at the platelet.
We express our gratitude to Drs. Otis R. Blaumanisand Patricia A. Grady, Department of Neurology, University of Maryland, Baltimore, for supplying the scanning electron micrographs (Figs. 1, 2, and 3).
References 1. Botterell EH, Lougheed WM, Scott JW, Vandewater SL: Hypothermia, and interruption of carotid, or carotid and vertebral circulation, in the surgical management of intracranial aneurysms. J Neurosurg 13:1-42, 1956 2. Clower BR, Haining JL, Smith RR: Pathophysiological changes in the cerebral artery after subarachnoid hemorrhage, in Wilkins RH (ed): Cerebral Arterial Spasm. Baltimore: Williams & Wilkins, 1979, pp 124-131 3. Conway LW, McDonald LW: Structural changes of the intradural arteries following subarachnoid hemorrhage. J Neurosurg 37:715723, 1972
45
4. Cromptom MR: The pathogenesis of cerebral infarction following the rupture of cerebral berry aneurysms. Brain 87:491-510, 1964 5. Grady PA, Blaumanis OR, Nelson ER: Morphology and flow dynamics of focal arterial constriction, in Wilkins RH (ed): Cerebral Arterial Spasm. Baltimore: Williams & Wilkins, 1979, pp 107-112 6. Kapp J, Mahaley MS Jr, Odom GL: Cerebral arterial spasm. Part 2. Experimental evaluation of mechanical and humoral factors in pathogenesis. J Neurosurg 29:339-349, 1968 7. Kapp J, Mahaley MS Jr, Odom GL: Cerebral arterial spasm. Part 3. Partial purification and characterization of a spasmogenicsubstance in feline platelets. J Neurosurg 29:350-355, 1968 8. Ohta T, Kajikawa H, Funatsu N, Yoshikawa Y, Someda K: Cerebral vasospasm and its relaxation responses to vasodilators: pathological study of severe prolonged vasospasm, in Wilkins RH (ed): Cerebral Arterial Spasm. Baltimore: Williams & Wilkins, 1979, pp 132-138 9. Peerless SJ, KassellNF, Kiyohide K, Hunter IG: Cerebral vasospasm: acute proliferative vasculopathy? II. Morphology, in Wilkins RH (ed): Cerebral Arterial Spasm. Baltimore: Williams & Wilkins, 1979, pp 88-96 10. Perret GE, Nibbelink DW: Randomized treatment study. Carotid ligation, in Sails AL, Nibbelink DW, Torner JC (eds): Aneurysmal Subarachnoid Hemorrhage. Baltimore: Urban & Schwarzenberg, 1981, pp 121-143 11. White RP: Overview of the pharmacology of vasospasm, in Wilkins RH (ed): Cerebral Arterial Spasm. Baltimore: Williams & Wilkins, 1979, pp 228-236 12. Wilkins RH: Attempted prevention or treatment of intracranial arterial spasm. A survey, in Wilkins RH (ed): Cerebral Arterial Spasm. Baltimore: Williams & Wilkins, 1979, pp 542-555
Editorial: The Future T h e future does n o t just: happen. For the most part it is something that someone: has made happen. If one is to mold the future, one must see the problems and t h e n ponder their solution. Sir Isaac N e w t o n said, "If I have seen further it is by standing upon the shoulders of giants." Every neurological surgeon stands upon the shoulders of giants. If any one of t h e m does n o t see the problems that must confront t h e m everyday, it is because they will n o t look. If they will do n o t h i n g about these problems, it is because they will not t h i n k and act. A l m o s t one h u n d r e d ,¢ears ago, Sir Victor Horsley was asked by one of his colleagues why he did n o t stop doing his operations since the results were so poor. He replied wisely, "If I do n o t continue those who come after me will do no better." It is n o t enough, however, just to continue to make the same mistakes. O n e must endeavor to correct them. Look at one aspect of neurological s u r g e r y - - b r a i n tumors. Over the years many improvements have been made. W e operate upon them with greater certainty and with more ease. Some of' our patients are now cured while others live longer and in a better condition than formerly.
O n the other hand, so far as cerebral gliomas are concerned, we are doing no better today t h a n Sir W i l l i a m Macewen and Victor Horsley did a hundred years ago. Practically all patients with cerebral gliomas still die of their tumors. W e are still not curing them. Cerebral gliomas do not pose an impossible problem. There are no malignant gliomas, and they do n o t metastasize. T h e y remain localized. If such a tumor were removed completely, it would be cured. Cerebral gliomas are the most c o m m o n of all brain tumors. T h e y are our problem. A hundred years is too long to be still faced with a problem without solving it. It is a problem for neurological surgeons since it is unlikely that these tumors will be cured by anything except complete removal. W e must face up to this problem. It is a many-faceted one, and its solution will require much thought and considerable effort by many people. It is time we got off our backsides and did something about it.
Paul C. Bucy, M . D . , Editor