 |
|
||||
| Home | Login | |||||
| About | Editorial board | Articles
|
NSI Publications
|
Search | Instructions | Online Submission | Subscribe | Videos | Etcetera | Contact |
|
||||||||||||||||||||
|
|
|
Acute subdural haematoma : a reappraisal.
Correspondence Address:
Acute subdural haematoma is a well-entrenched nosological entity implying subdural collection of blood following acute head injury. Pathologically, it is usually associated with or, for that matter, secondary to cerebral contusion and laceration. Based on cumulated experience, clinical and pathological studies it is proposed that, for too long the neurosurgeons have put emphasis on the clot rather than the totality of the pathological anatomy and that they have focused their therapeutic strategy on removal of the accumulated blood, unmindful of the associated parenchymatous lesion. Not surprisingly, such attempts have been associated with a very high mortality. On the basis of nearly four decades of personal experience and critical review of the literature, evidence has been provided that to reduce the mortality associated with this condition, it is necessary to evolve a strategy, not only to evacuate the blood but comprehensively deal with the associated parenchymatous lesions and the cascade of secondary insult to the underlying brain.
'Words have crept into the medical vocabulary and have often been retained long after the ideas for which they stood have been refuted'. (Raymond Adams) Nosologically a correct term, acute subdural haematoma (ASDH), which is generally the tip of an ice berg, has misled a generation of neurosurgeons in evolving an appropriate management. Based on the corollary of extradural haematoma, the 'logical' treatment of this lesion was considered to be burr-hole evacuation of the clot. If one uses the borrowed expression from Adams, this label 'embalmed' a fallacious theory.[1] It is not surprising therefore, that Klun and Fettich[2] in a review of the largest series of ASDH called it as one of the 'unsolved problems of neurosurgery'. Much fatalistic nihilism developed over this problem as mortality rates of 80 - 90 percent were reported.[3],[4],[5],[6],[7],[8],[9] Many neurosurgeons doubted the utility of surgery, specially within the first hours of injury.[10],[11] Pathophysiological Considerations : ASDH and Cerebral Contusion Following a review of approximately 3000 cases of head injury, among which all fatal cases were submitted to autopsy, Kristiansen and Tandon,12 in 1960 concluded that acute subdural haematoma is seldom an isolated lesion. It was pointed out that, 'the pathogenesis of subdural haemorrhage makes it nearly inevitable that a small contusion in the pathological sense is co-existent'. And further more, '.......there are patients in whom the extensive associated contusion and laceration may be the dominant feature, the SDH only a relatively unimportant co-existent pathological finding. In between these two grades are cases in which the two lesions are independently significant and add to the severity of each other'. This study had also revealed that a thin layer of subdural blood was consistently seen in association with traumatic intracerebral haematoma which, from a pathological stand point, was difficult to differentiate from cerebral contusion with haemorrhage. Thus, in many patients the cerebral contusion/laceration may, on one hand be associated with a layer of subdural clot and on the other, extend into the brain parenchyma to manifest as an intracerebral haematoma. However, many neurosurgeons at that time, (and even today) were primarily concerned with removing the subdural clot through multiple burr-holes and lamented about the high fatality rates among these patients. Mc Laurin and Tutor,[13] unaware of our observations commented that, 'those patients in whom subdural collections develop and who expire within first few days, undoubtedly do so because of associated cerebral damage.' They could not find any case at autopsy 'which disclosed only venous rupture without accompanying parenchymal injury'. They concluded that, 'the neurological picture was primarily, if not completely the result of associated direct brain damage'. Courville and Blomquist,[14] in their autopsy study on traumatic intracerebral haemorrhage had already documented this. Jamieson and Yelland,[15] in a study of 553 cases of traumatic SDH subdivided these into simple SDH, SDH with contusion, and complicated haematomas. The last category was observed after severe acceleration-deceleration and carried the highest mortality rates. Richard and Hoff[16] described gross disruption of the brain underlying the haematoma in 61% of the fatal cases of ASDH. Britt and Hamilton[17] found massive cerebral oedema or extensive contusion at necropsy in the majority of their cases of ASDH. Adams et al,[18] summarising their years of clinical and pathological studies, concluded that 'the intradural haematomas comprise 'pure' subdural haematomas caused by the tearing of bridging veins in the subdural space, 'burst' lobes (intracerebral and subdural haematomas in continuity) and intracerebral haematomas (mainly in the frontal and temporal lobes in relation to contusion). Jennett and Teasdale[19] observed that the relative contribution of the subdural and intracerebral components vary, but there is a sizeable subdural haematoma in most of the cases. Based on aetiopathogenesis, Miller and Statham20 classified ASDH into three types: i) haematoma associated with laceration of the brain, usually at temporal pole, in which there is a mixture of intracerebral contusion and haemorrhage and an acute subdural haematoma, a condition known by some neurosurgeons as 'burst temporal lobe'; ii) ASDH that results from tearing of a bridging vein between the surface of the brain and one of the main venous sinuses; iii) haematoma due to bleeding from a small artery on the surface of the brain that may have been injured by an overlying fracture of the skull. Amongst 5000 cases of head injury which included 209 cases of extradural haematoma, 257 cases of temporal lobe lesions acting as expanding mass, there were only 25 cases with ASDH without any detectable parenchymatous lesion. The possibility of the associated parenchymatous lesion, having been overlooked at surgery cannot be ruled out, since many patients in this series belonged to pre-CT era and not all fatal cases had autopsy.[1] The significance of the associated brain lesions in patients with ASDH is now being generally recognized. The availability of CT scan has been greatly responsible for establishing this fact. In a series of 109 consecutive head injured patients with the CT scan diagnosis of ASDH, there were only 8 cases with pure ASDH, 49 patients had contusion, while others had extradural haematoma, intracerebral haematoma, subarachnoid haemorrhage and brain swelling.[21] Wilberger et al[22] concluded that the patho-physiological evidence supports the belief that in ASDH 'the extent of primary underlying brain injury is more important than the subdural clot itself, in dictating the outcome'. Most other authors[23],[24],[25],[26] found mortality and morbidity related to the extent of underlying brain damage rather than the extent of the subdural blood clot. Yet, till very recently, the surgical attention was primarily directed to the removal of compression due to the surface blood clot. In view of the above, it could be safely concluded that while a 'simple' ASDH due to tearing of the bridging vein/veins without concomitant or associated cerebral pathology, no doubt exists, it is a rarity than a rule. Thus, in a vast majority of such cases, the aetiopathogenesis, clinical picture and prognosis cannot be explained simply on the basis of a surface clot producing focal compression of the brain. Acute subdural haematoma and diffuse axonal injury While cerebral contusion, laceration or intracerebral haemorrhage are obvious on CT scans and to the naked eye at operation or autopsy, diffuse axonal injury (DAI), which may occasionally be suspected on the basis of CT findings, can only be observed in carefully conducted pathological studies. The significance of this lesion in the clinical manifestations and outcome of patients with ASDH has, therefore, not received due attention. Considering the biomechanism of both DAI and ASDH, both being caused by rotational acceleration/deceleration, the coexistence of the two is not unexpected. Infact, Adams et al[18] and Gennarelli et al[27],[28] reported that 11% of all patients with DAI have associated intracranial haematomas. A detailed study of six fatal cases of ASDH by Sahuquillo-Barris et al[29] revealed widespread evidence of DAI in all. These authors concluded that 'in some patients with a head injury, acute SDH may be only an epiphenomenona of a primary impact lesion of variable severity; that is a diffuse axonal injury. In these cases, the final outcome is fundamentally dependent on the severity of the subjacent diffuse axonal injury'. It is, therefore, understandable that the patients with ASDH who remain unconscious from the moment of head injury, as many do, clinically reflect the severity of this primary damage rather than the compressive effect of any associated subdural blood. Acute subdural haematoma and ischaemic neural damage Ischaemic brain damage is the commonest neuropathological abnormality found in patients, who die of ASDH.[30] The pathological observations of ischaemic brain damage in fatal head injury patients,[31],[32],[33] angiographic evidence of cerebral vasospasm in patients with severe head injury[34] and post-traumatic temporal lobe lesions35 (generally associated with overlying subdural blood), failed to attract due attention of the neurotraumatologists. The latter, in search for a cause for brain swelling, seem to have been, preoccupied with the concept of 'luxury perfusion'.[36],[37],[38],[39] More recently, studies on experimental model of ASDH in rats have unequivocally established the development of ischaemic brain damage underlying the subdural blood.[40],[46] The possibility of some 'species difference' in this response has been pointed out by Shaver et al,[42] who observed extensive areas of the white matter necrosis under the haematoma in an infant piglet model of ASDH. Zones of necrosis were also present in the cortex though much less than in the rodent model. While the precise mechanism by which ASDH causes underlying brain ischaemia remains uncertain, there is evidence to suggest that neither focal nor generalised increased intracranial pressure could adequately explain the condition. The possibility of direct effect of some vasoactive substances released by the blood clot, being responsible for the ischaemia, seems attractive.[44] This finds confirmation in the experimental studies by Chen et al.[30] Using microdialysis techniques, they had demonstrated that ASDH was associated with massive release of glutamate, to six times the normal levels, for 20 minutes in the underlying cortex. They demonstrated 54 percent reduction in the ischaemic damage in animals pretreated with a glutamate antagonist. This was confirmed by Bullock et al[45] and Kuroda et al,[46] who found profound increase in the levels of extracellular excitatory aminoacids (EAA) within the ischaemic zone under the SDH. They postulated that the EAA induced increase in local cerebral glucose utilization resulted in a breakdown of the metabolism and cerebral blood flow coupling. Inglis et al[47] had earlier demonstrated reduction of ischaemic brain damage associated with tissue hypermetabolism in ASDH following use of a glutamate antagonist. It is important to note that Kuroda and Bullock[40] failed to observe any significant effect on the underlying area of cerebral ischaemia following removal of the subdural blood in their rodent model of ASDH. They hypothesised that clotted blood, in contact with the cortical surface, initiates 'this focal progression of ischaemia and oedema, which is not reversed by removal of the haematoma...'. Duhaime et al,[41] in a well planned experiment to elucidate the relative role of increased pressure, vasoactive effects or toxicity of the blood itself, concluded that blood in prolonged contact with the cortical surface in the absence of an increased pressure, ischaemia, or other insults is insufficient to cause underlying infarction. Schroder et al48 documented reversal of ischaemia immediately after removal of an ASDH in two patients. However, they believed that compression of the microcirculation could be responsible for the ischaemia. A more recent CT study on ASDH on 31 patients by Domenicucci et al[49] provides an interesting angle to this whole issue. Amongst these 31 patients, they identified a group of 5, whose preoperative CT scans showed intact subarachnoid space underlying the haematoma and absence of blood in the cerebrospinal fluid, though all of them had appreciable midline shift. They classified these five cases as 'intradural' haematoma and recorded a mortality of only 20% as compared to the overall mortality of 68% for the whole series. They considered the good result in these patients with undamaged arachnoid barrier to be due to the shielding of the cortex from the neurotoxic and vasoactive substances preventing the 'ischaemic and oedemigenic' response in the brain tissue. It is known that such vasoactive and neurotoxic substances are released not only by the subarachnoid and subdural blood but also by the brain damaged by an impact injury or ischaemia-induced by hypoperfusion and hypotension occurring at the time of injury. The relative contribution of each one of these mechanisms in initiating and maintaining the vicious cycle of increasing brain damage and oedema, so commonly associated with ASDHS, is difficult to judge. Acute SDH and brain swelling One of the least understood and most devastating features of ASDH is the associated brain swelling which frequently acquires malignant proportion following evacuation of the haematoma. Among 178 patients with ASDH, Zumkeller et al,[50] recorded a mortality of 50% when midline shift exceeded the thickness of haematoma by 3 mm in the CT scan. However, if the midline shift exceeded haematoma thickness by 5 mm, the survival rate was only 25%. At surgery itself, often quite unpredictably, the brain herniates through the craniotomy following evacuation of the clot, making dural closure impossible. In some cases, a post-operative CT scan shows cerebral hemispheric enlargement, with persistent or even increased midline shift in absence of any residual haematoma. The increased intracranial pressure and the progressive downhill course, unresponsive to the routine antioedema measures, ends in a fatal outcome in nearly 70-80 percent of such patients.[51] Britt and Hamilton17 found 'massive cerebral oedema or extensive contusion' at necropsy in the majority of their cases of ASDH. While this phenomenon is not unknown following evacuation of extradural haematomas, it is far more frequent in cases of ASDH.[52],[53] Thus, Lanksch et al[54] found cerebral oedema in 8% of 118 patients with extradural haematoma as compared to 26% of 168 patients with ASDH. The long standing controversy regarding its nomenclature - brain oedema versus brain swelling, and its pathophysiology - vascular hyperperfusion versus angiogenic or cytotoxic oedema, has not yet been unambiguously resolved. Unilateral cerebral swelling of an entire hemisphere, not specifically attributable to contusion, intracerebral haematoma or infarction was observed in 17 of 151 patients of ASDH.[18] Its aetiopathogenesis continues to be illunderstood. It is, therefore, not surprising that its management remains a dismal chapter in neurotraumatology. Considering the voluminous amount of literature on the subject, no attempt will be made to discuss it in any details. However, a brief reference to some of the studies will justify the above statements. In a series of experiments on ASDH, which included measurements of brain tissue water, blood-brain barrier studies with labelled sucrose and electron microscopy observations, Gennarelli[55] failed to find evidence of cerebral oedema in the swollen brain. He attributed it to increased cerebral blood volume as had already been suggested by Langfitt et al[52] from their experimental studies and Bruce et al[38] from the clinical data of patients with so-called syndrome of 'malignant brain oedema'. On the other hand, Tornheim and McLaurin,[56] in their experimental studies, Galbraith et al[57] from measurement of water content of white matter after head injury in man and Yoshino et al,[58] following dynamic CT studies in fatal head injury patients concluded that acute oedema formation is a common cause of brain swelling in cases of fatal head injury. Kuroda and Bullock,[40] measured local cerebral blood flow, before and after removal of ASDH in the rat led, and conclude that focal ischaemic and oedema rather than vascular engorgement or hyperaemia was the major cause of hemispheric swelling, seen after removal of haematoma. It is true that cerebral circulatory disturbances are far more common in patients with severe head injury than is generally recognized. Earlier cerebral blood flow (CBF) studies[36],[37],[38],[39] highlighted the incidence of vasoparalysis leading to hyperaemia or 'luxury perfusion' resulting in brain swelling. More recent studies[40],[59],[60] established the high incidence of impaired CBF, decoupling of CBF and metabolism, failure of autoregulation, resulting in cerebral ischaemia and infarction with consequent brain oedema. The role of local pressure by the haematoma and the release of vasoactive and cytotoxic substances in this process, have already been described above. Effects of compression by experimental intracranial mass lesions on cerebral blood flow have been elucidated by Jakobsson et al.[61],[62] A resolution of this controversy is not just of an academic interest but has direct bearing on the therapeutic management of these patients. The currently recommended 'aggressive treatment' of all severe head injury patients which includes artificial respiration, high doses of corticosteroids, intravenous mannitol, and even use of barbiturates,63-66 not unexpectedly failed to reduce the mortality drastically.[67],[68] In more recent years, attention has been directed to elucidate the molecular mechanisms underlying these secondary 'autodestructive' processes. Excitotoxic substances like glutamate and aspartate,[30[,[69],[70],[71] free radicals,[72],[73] and platelet activating factor[74] have been implicated. A cascade of events initiated by the primary traumatic event is complicated by secondary events of local and generalised increase in intracranial pressure, abnormalities of cerebral perfusion and ischaemia. These are undoubtedly overlapping events as also self perpetuating vicious circles. Removal of the subdural blood is, thus, only one and not necessarily the most important component of any therapeutic strategy aimed at reducing the currently unacceptable mortality rates of ASDH, which remain 50 per cent or above. Timing of surgery A very high mortality associated with surgery for ASDH, as mentioned earlier, prompted a number of neurosurgeons to delay operations on such patients. Futility of operating within 24 hours was a common refrain till 1970s. A landmark paper by Seeling et al[75] claimed a dramatic reduction in mortality to 30%, if the subdural haematoma was evacuated within 4 hours of injury, as compared to a mortality rate of 85%, if the operation was delayed. Inspite of very concerted efforts on the part of others, in achieving this 4 hour target, there has been no clear confirmation of such excellent results in the management of ASDH. Stone et al[76] while claiming that direct admission to a trauma centre yielded improved results, failed to find any difference in mortality between patients operated upon within 4 hours and those undergoing surgery after 4 hours from injury. In a selected group of patients of ASDH with admission GCS scores of 11 to 15, Croce et al[77] found no correlation between timing of surgery and neurologic outcome. In contrast to this, Dent et al[78] reported an unselected series of 211 patients of ASDH (GCS 3 to 15), 83 of whom were managed with craniotomy (average GCS 7,8). Thirty five percent of operated patients had their haematoma, evacuated within 4 hours and 65% after 4 hours. Early operated patients had a significantly lower incidence of functional survival (early 24% v.s. delayed 51%). Hatashita et al,[83] in their series of 60 patients with ASDH observed a mortality of 62% for patients with GCS scores of 4-6 operated within 4 hours, in contrast to 33% for those operated upon from 4 to 10 hours. All patients with GCS of 3 died while all patients with GCS of 7 or more achieved functional recovery. Wilberger et al[22] recorded a mortality rate of 59% for patients operated on within 4 hours as compared to 69% for those operated on after 4 hours. However, they observed that 'the results do not support the crucial role of the timing of operative intervention established by Seeling et al,[75] in 1981'. While the debate regarding the benefits of early surgery (within 4 hours) continues unabated, it is generally agreed now that denial of operation during the first 24 hours or delaying surgery in the hope of improving the outcome has no basis at all. As advocated by us earlier, we believe that surgery should be performed as early as possible, specially in patients with GCS of 8 or below, with more than five millimeter shift of midline, without waiting for deterioration in the clinical condition or rise of intracranial pressure. Failure to observe progressive improvement and not evidence of deterioration is in itself an indication for surgery.[1] Type of surgery Believing the subdural clot to be the real culprit responsible for the clinical condition of the patient, the standard surgical procedure advocated in earlier years was to deal with it through burr-hole evacuation. This not only proved to be futile in majority of cases, it also prevented the surgeon from appreciating the associated brain lesions required to be attended to. Classifying patients with ASDH with contusion as a distinct entity already in 1960, we advocated, 'in the management of these cases, attention must be paid to both components of the lesion'. It was recommended that 'severely damaged and lacerated tissue may be sucked away and bleeding points cliped or coagulated'. It was recognized that while burr-hole evacuation was adequate for 'uncomplicated subdural haematomas', larger opening was required to deal with ASDH with contusion since there was a higher frequency of clotted and mixed haematomas, which are not easy to remove through a single burr-hole. In addition, burr-holes did not permit adequate attention to the underlying damaged brain.[12] Increasing recognition of this problem made us abandon the policy of using burr-hole evacuation for all cases of ASDH (as also for patients with pulped temporal/frontal lobes) and routinely use, either an osteoplastic craniotomy or a large Scoville trephine. This brought down our overall mortality for such patients to 43%.[35] Following this practice, it was only a rare occasion when we had to face the problem of dural closure. This is now the generally accepted policy, since several studies have confirmed improved prognosis following craniotomy as compared to burrhole evacuation.[1],[15],[51] The frequent occurrence of severe brain swelling, and raised intracranial pressure unresponsive to standard therapy prompted many surgeons to recommend bifrontal or hemicranial removal of bone.[84],[85] Phuenpatham et al[21] found removal of bone essential in 28 out of 83 patients owing to brain bulging out through the craniotomy opening, making the standard closure impossible. Zumkeller et al[50] while favouring craniotomy for evacuation of haematoma reserved craniectomy for massive brain swelling. Rosenorn and Gjerris[9] failed to find any difference in prognosis for the different surgical treatments. Fell et al,[7] in their series of 144 patients with ASDH, found no difference in mortality between patients treated with multiple burr-holes and subtemporal craniectomy (41 percent) and large craniotomies (45 per cent).
On the basis of the current knowledge, it is reasonable to conclude the following i) Isolated ASDH, acting as a compressive lesion, is an uncommon clinicopathological entity. ii) Majority of patients with ASDH have associated focal (contusion / laceration / intracerebral haematoma) or global (diffuse axonal injury, subarachnoid haemorrhage) involvement of the brain or both of these. iii) In addition, ischaemia underlying the haematoma and ipsilateral hemispherical brain swelling (? hyperaemic, ? oedematous), which may be self perpetuating, are consistent findings in most patients with surgically significant haematoma. iv) A combination of these factors, if not attended to promptly, may lead to uncontrollable rise in intracranial pressure with its adverse consequences including foraminal herniations, brainstem compression and haemorrhages. v) The molecular basis underlying the cascades of events, resulting in secondary damage, are now better elucidated and understood. vi) It is now established that even with successful and early evacuation of the clot and the current aggressive intensive care management, the mortality still remains high, (50% or above). Functional survival is even much lower. Some of the routinely used ICU practices not only fail to arrest or reverse the secondary changes, but may aggravate the same. vii) The future hope for further reduction in morbidity and mortality lies in preventing, arresting or reverting the molecular events responsible for the secondary ischaemia and cytotoxic oedema. viii) It is obvious that the term ASDH does not reflect the true pathology or pathogenesis of the lesion/s determining the clinical picture or outcome. ix) It is felt that this term has outlived its utility in determining the appropriate therapeutic strategy. However, realising the difficulty of abolishing a long standing term, it is recommended that it could be subclassified as SDH with or without associated parenchymatous pathology which can be established preoperatively with the current diagnostic modalities. This would help to shift the focus of attention from the haematoma to the brain itself and permit better comparison of different therapeutic modalities.
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|||||
