Acute subdural hematomas: an age-dependent clinical entity

Transcription

1 J Neurosurg 71: , 1989 Acute subdural hematomas: an age-dependent clinical entity MATTHEW A. HOWARD III, M.D., ALAN S. GROSS, B.S., RALPH G. DACEY, JR., M.D., AND H. RICHARD WINN, M.D. Department of Neurological Surgery, University of Washington School of Medicine, Seattle, Washington v, Reports prior to 1980 describe overall mortality rates for acute subdural hematomas (SDH's) ranging from 40% to 90% with poor outcomes observed in all age groups. Recently, improved results have been reported with rapid diagnosis and surgical treatment. A relatively large number of older patients (34 patients over 65 years old) were treated recently at Harborview Medical Center, enabling a retrospective comparison with similarly treated younger patients (33 patients aged 18 to 40 years). Clinical information and computerized tomography morphometric data were obtained. Patients in the younger group were most often injured in motor-vehicle accidents (15 cases), whereas falls were most frequent in the older group (19 cases). Patients in both groups were rapidly resuscitated in the field; more than 30% were treated within 1 hour after the time of injury. Injury severity, determined by the admission Glasgow Coma Scale score, was similar for the two groups. Mean acute SDH volume was significantly larger in the older patients than in the younger group (mean standard deviation: vs cu cm), as was the amount of midline shift ( vs cm). Surgical treatments were similar, but outcomes were dramatically different for the younger and older patients. Mortality rates were more than four times higher in older patients than in younger ones (74% vs. 18%). Three older patients and 25 younger patients were functional survivors. Old age, a larger SDH volume, and a larger midline shift all correlated with a poor outcome. The results of this study suggest that the pathophysiology of acute SDH varies with age, and that currently employed resuscitation and treatment methods have differentially improved the outcome for younger patients. KEY WORDS " subdural hematoma 9 head injury 9 outcome A CUTE subdural hematomas (SDH's) are highly lethal injuries. Reports from the pre-computerized tomography (CT) era cite mortality rates ranging from 65% to 90%, with poor outcomes observed in all age groups. 3-5'12'13'15'16'2~176 More recently, mortality rates have decreased, presumably as a result of improved ability to rapidly diagnose and treat head-injured patients. 7'27'29 Seelig, eta/., 27 reported a 40% overall mortality rate in patients treated surgically for acute SDH and a 25% mortality rate for patients operated on less than 3 hours following their injury. Patient age was not a significant predictive factor in that series. Patients with acute SDH admitted to Harborview Medical Center (HMC) are treated in an aggressive and systematic fashion. In recent years, a large number of elderly patients (over 65 years old) have been treated, enabling us to make useful comparisons between this older group and a group of young patients (aged 18 to 40 years) with acute SDH. A review of patient histories and morphometric analysis of admission CT scans of the head revealed significant clinical and radiographic differences between the two age groups. The implications of these findings for understanding the pathophysiology and treatment of acute SDH are discussed. Clinical Material and Methods The medical records were reviewed for 49 patients aged 18 to 40 years and for 54 patients aged over 65 years who were admitted to HMC between 1983 and 1986 and who were subsequently discharged with a diagnosis of SDH or acute SDH. Inclusion criteria included: 1) a medical record delineating the admission Glasgow Coma Scale (GCS) score, treatment, and outcome to a point at least 2 months from the time of treatment, and 2) an acute SDH thicker than 0.5 cm diagnosed on review of the CT scan. The study included 858 J. Neurosurg. / Volume 71/December, 1989

2 Acute subdural hematomas and age TABLE 1 Reasons for excluding 37 patients from the study Reason for Exclusion Older Younger Group Group chronic subdural hematoma 18 1 no subdural hematoma (or < 0.5 cm) no medical record total patients FIG. 1. Bar graph showing the mechanisms of injury causing acute subdural hematomas in young (striped bars) and old patients (solid bars). The number of patients in each injury category is reflected as a percentage score on the vertical axis. MVA = motor-vehicle accident; PED vs CAR = pedestrian struck by car. TABLE 2 Time from head injury to emergency room admission Time Interval Older Group Younger Group < 1 hr l to < 3 hrs to 6 hrs 2 1 > 6 hrs unknown total cases "young" and 34 "old" patients. Twenty-one old patients and 16 young patients were not included for reasons described in Table 1. Admission GCS score was determined based on physical examination findings as described by Teasdale and Jennett. 3~ Intubated patients were assigned one point for the GCS verbal category. Additional clinical data collected included mechanism of injury, time from injury to emergency room (ER) admission, time from ER admission to start of surgery, time from treatment to death, and cause of death. Because there was some uncertainty as to the exact time of injury, we separated the interval between injury and ER admission into four categories: less than 1 hour, 1 to 3 hours, 3 to 6 hours, and more than 6 hours. If the medical record did not adequately describe the injury-to-er admission interval, patients were placed in the "unknown" time interval category (see Table 2). The Glasgow Outcome Scale (GOS) category, as described by Jennett and Bond, ~4 was determined in surviving patients from either a report of a recent clinic visit or a telephone interview. The five GOS categories are: good; moderately disabled; severely disabled; vegetative; and dead. All outcome scores were obtained at least 2 months after treatment. "Functional survival" is defined as a GOS grade of good or moderately disabled. Fifty-two admission CT scans (22 in the young group and 30 in the old group) were available for analysis of acute SDH volume, size of midline shift, and presence or absence of intracerebral hematoma. Acute SDH volume and magnitude of midline shift were calculated using a Zeiss Videoplan image analysis system.* Univariant analyses comparing the young and old groups were performed tbr all categories of clinical and radiographic data. Each variable was analyzed using * Zeiss Videoplan image analysis system manufactured by Carl Zeiss, Inc., Thornwood, New York. Student's t-test. Multivariant analysis was performed using a logistic regression method to determine the independent significance of various clinical and CT scan features in predicting outcome. Clinical Data Results There were 33 patients (18 men and 15 women) in the younger group and 34 patients (20 men and 14 women) in the older group. Mechanism of Injury. The mechanisms of injury in this series are shown in Fig. 1. The most frequent type of injury sustained by young patients was associated with a motor-vehicle accident -- the cause of injury in 15 of the 33 patients. Falls accounted fbr only four of the injuries sustained by young patients, and all of these patients were intoxicated (ethyl alcohol level of > 0.10 mg/dl on admission) and had fallen or been pushed from a height. In contrast, 19 of 34 elderly patients sustained simple falls and three were involved in a motor-vehicle accident. The difference in types of injury sustained by the two groups was statistically significant (p < 0.001). Time to Treatment. All unconscious patients were initially evaluated and resuscitated by paramedics in the field. Of the 44 patients with an admission GCS score of 8 or less, all but two old and four young patients were intubated prior to arrival at HMC. Eleven (33%) of the 33 young patients and 15 (44%) of the 34 old patients were admitted to HMC less than 1 hour after injury. Table 2 shows the number of patients with longer intervals between injury and admission. An accurate interval could not be determined for six (18%) of the older and seven (21%) of the younger patients. J. Neurosurg. / Volume 71/December,

3 M. A. Howard III, et al. TABLE 3 Interval (rain)from emergency room admission to start of surgery* Time Elapsed Older Group Younger Group mean range standard deviation median no. of cases * Values were missing for four patients in the older group and 11 in the younger group. TABLE 5 Initial surgical treatment in this series" Initial Procedure* Older Group Younger Group craniotomy ICP monitor/ventriculostomy 3 2 burr holes 3 3 twist drill 1 4 none no. of cases * ICP = intracranial pressure. TABLE 4 Admission Glasgow Coma Scale (GCS) scores Injury Severity Older Group Younger Group minor (GCS score 13-15) 7 9 moderate (GCS score 9-12) severe (GCS score 3-8) no. of cases Differences in time elapsed before treatment between age groups were not significant (p > 0.4). Exact time intervals between ER admission and start of surgery were available for 22 of the 33 young patients and 30 of the 34 old patients (Table 3). Young patients underwent surgery an average (_+ standard deviation) of minutes from the time of ER admission, and old patients 198 _+ 127 minutes. The difference in time elapsed before surgery between the age groups was not significant. Neurological Examination on Admission. The injury severity distribution reflected in the admission GCS scores was not significantly different for the two groups (p > 0.4, Table 4). The majority of patients presented with a severe head injury (GCS score 3 to 8: 21 young, 23 old), and the remainder fell within the moderate (GCS score 9 to 12: three young, four old) or minor (GCS score 13 to 15: nine young, seven old) head-injury categories. Neurological status in the young patients did not deteriorate significantly between the time of admission and the start of emergency surgery. In contrast, while being evaluated prior to surgery, two of the seven older patients who presented with an admission GCS score suggesting minor closed head injury (GCS score _> 13) suddenly developed symptoms and signs consistent with uncal herniation. Treatment. After initial resuscitation in the field, all unconscious patients were oxygenated, hyperventilated, and hemodynamically stabilized in the ER. Patients who were hemodynamically stable with signs of brainstem dysfunction or abnormal motor movements received mannitol. Patients with systemic injuries requiring immediate surgery were brought directly to the operating room without a CT scan; acute SDH was diagnosed in these patients with exploratory burr holes. Six patients had burr holes placed (three old, three young); five patients early in the series underwent twistdrill craniotomies (Table 5). All other patients received CT scans prior to surgery. Neurosurgical treatment of patients with acute SDH diagnosed by CT scan most often consisted of a wide craniotomy, hematoma evacuation, and placement of an intracranial pressure (ICP) monitor (48 patients). Five patients with significant head injuries and a small SDH causing minimal mass effect underwent placement of an ICP monitor or ventriculostomy alone. Six patients had no significant treatment; these patients were either brain-dead on admission (one patient) or were normal neurologically with a small acute SDH noted on CT scan (five patients). These five latter patients were observed closely, and follow-up CT scans were obtained. The distribution of types of surgical treatment did not vary significantly between the two age groups (p > 0.05). Outcome. The outcome in this series is summarized in Fig. 2. The mortality rate was significantly higher in the older group compared to the young group (74% vs. 18%, respectively; p < 0.001). Of the elderly patients who died, 64% died as a direct consequence of severe brain injury and the remainder of systemic complications. Most patients in the latter group were severely impaired neurologically when systemic complications developed. However, two elderly patients who presented with minor closed head injuries and had not deteriorated neurologically developed fatal cardiac complications while convalescing, 16 and 38 days after their injuries. Mean time from injury to death for the older group was days. The principal cause of death in young patients was brain injury; their mean time from injury to death was days. Outcomes for elderly survivors were generally poor. Eight were severely disabled or vegetative, and only three were functional survivors (moderately disabled or good outcome). The only elderly patient in the series with a good outcome presented with a minor closed head injury. In contrast, two of the 27 young survivors were severely disabled or vegetative, and the remainder were functional survivors. Even in the group of young patients who presented with a severe closed head injury 860 J. Neurosurg. / Volume 71/December, 1989

4 Acute subdural hematomas and age comes could not be proved because very few patients had one of these features without the other two. Consequently, none of these three variables correlated with outcome independently (p = 0.125, logistic regression with three variables: acute SDH volume, midline shift, and age). FIG. 2. Bar graph depicting the distribution of Glasgow Outcome Scale scores as a function of admission Glasgow Coma Scale (GCS) scores for young (striped bat:t) and old (solid bars) patient groups with acute subdural hematomas. Three categories of admission GCS scores are represented, and the number of patients in each category is represented on the vertical axis. SD/VEG = severely disabled or vegetative; MD = moderately disabled. (GCS score 3 to 8), 67% had either a good outcome (five patients) or were only moderately disabled (nine patients). CT Scan Data Fifty-two CT scans were available for analysis (30 in the young group and 22 in the old). Acute SDH volumes were larger in the old group (mean _ cu cm) than in the young group (mean cu cm). Midline shift was also greater in the old group (mean cm) than in the young group (mean cm). Both differences are statistically significant (p < 0.01). The incidence of intracerebral hematoma or contusion was 50% and 54% in the old and young groups, respectively. Statistical Analysis The results of univariant analyses have been described above. Multivariable analysis using a logistic regression method was employed to determine which clinical and CT scan variables correlated significantly with outcome. Three characteristics (age over 65 years, a large acute SDH volume, and large midline shift) correlated significantly with poor outcomes when analyzed individually (p < 0.001). The other variables analyzed, including time from injury to ER admission and time from ER admission to start of surgery, did not influence outcome (p > 0.05). The independent value of old age, large acute SDH volume, or large midline shift in predicting poor out- Discussion Acute SDH is a common finding in patients who have sustained severe closed head injuries. Reports in the last decade suggest that outcomes in this patient population have improved as a result of more rapid diagnosis and neurosurgical treatment. 7,27,29 A review of our experience with patients suffering from acute SDH who were systematically treated at a regional trauma center reveals divergent outcomes in young (18- to 40-year-old) versus old (over 65-year-old) patients. Differences in clinical presentation and CT findings provide possible explanations for this observation and offer insights into why improved outcomes were seen only in young patients. Approximately 30% of patients with severe closed head injuries have acute SDH. 9 Studies published from the pre-ct era report overall mortality rates ranging from 40% to 90%. 3-5'~2"13'~5'16'2~ 25,28,30 In general, outcomes in these earlier reports were slightly better for patients younger than 40 years old (40% to 76% mortality) than for patients older than 60 years (69% to 100% mortality). Time from injury to treatment varied considerably between studies, and neurosurgical treatment methods ranged from burr hole drainage to hemicraniectomy. More recently, emphasis has been placed on rapidly transporting head-injured patients to a regional neurotrauma center where medical resuscitation is initiated, a CT scan of the head is quickly obtained, and a craniotomy for hematoma evacuation is performed. 27'29 Seelig, et al., :7 reported a mortality rate of 30% for patients treated less than 4 hours posttrauma versus a 90% mortality rate for those treated more than 4 hours from the time of injury. In the study reported by Seelig, et al., intubation and hyperventilation were performed after arrival at the hospital. Age did not influence outcome in that study. The patients with acute SDH reviewed in the current series were also treated in a systematic fashion. Time to initiation of medical treatment was short. In contrast to other studies, unconscious patients in this series were intubated and resuscitated in the field by paramedics. Many patients were hyperventilated and hemodynamically stabilized less than 30 minutes from the time of injury. At least 39% of injured patients were hospitalized within 1 hour of injury. Mean times from injury to ER admission and from ER admission to surgery were similar for the two age groups. In this setting, significant age-related differences were observed in the type of injuries sustained, the CT scan findings, and neurological outcomes. J. Neurosurg. / Volume 71/December,

5 M. A. Howard III, et al. The most common mechanism of injury sustained by young adults was a high-speed motor-vehicle accident (45%). The paramedic field reports often described signs of severe neurological dysfunction, such as abnormal posturing and abnormal respiratory patterns, immediately after the injury. Sudden neurological deficits following this type of low strain rate injury, as described by Gennarelli, et al., TM are likely to have resulted from immediate direct brain injury. The relatively small acute SDH volume observed in young patients (21.6 vs cu cm for elderly patients) may in some cases have been an epiphenomenon as described by Sahuquillo-Barris, eta/. 26 As a group, the young patients were rapidly treated and transported to the hospital. Field intubation and resuscitation are likely to have restored adequate cerebral perfusion pressure and oxygenation, and created hypocapnia. These medical measures would, in theory, provide immediate and effective treatment for the direct-injury component of the brain injury which occurred immediately upon impact. 2'19 Following hospital admission, CT scans were obtained and hematoma evacuation was initiated an average of 3 hours and 11 minutes after arrival in the ER. The mortality rate for young patients was only 18%, and 66% of young patients were functional survivors. Encouraging results were noted even in those young patients presenting with a severe head injury (14 of the 21 with GCS scores 3 to 8 were functional survivors). Shorter times from ER admission to surgical treatment did not correlate with improved outcomes. This may be because the principle pathological process in many of the young patients was an immediate direct parenchymal injury; in these cases important medical treatment was initiated quickly in the field, allowing time for the acute SDH mass to be evacuated several hours later. Although direct comparisons cannot be made with earlier studies, these findings suggest that the current neurotrauma system has favorably influenced outcomes in young patients with acute SDH. The clinical histories for elderly patients with acute SDH were significantly different. The most common mechanism of injury for older patients was a simple fall (55%). A high incidence of falls causing severe head injuries in elderly patients has been reported previously by Marshall, et al. ~8 Frequently, falls were followed by a lucid interval followed by neurological deterioration. Two elderly patients who presented to the ER awake and alert following a minor head injury precipitously deteriorated while being evaluated. Neither patient had a good neurological outcome. The clinical pattern of delayed neurological deterioration after seemingly trivial injuries and the large acute SDH volumes noted on CT scans (mean 96.2 cu cm) suggest that the expanding SDH mass may be the most important pathophysiological process in this age group. Experimental and anatomical data indirectly support this theory. High strain rate head injuries, such as a head impacting a stationary floor, have been shown by Gennarelli, et a/., 1~ to be particularly effective in causing vascular disruption and SDH's in monkeys. Also, age-related brain atrophy causes stretching of bridging veins and is cited as a possible factor contributing to the predilection of elderly patients to develop SDH's. 6'22'33 Elderly patients with acute SDH were rapidly treated in the field and quickly transported to the ER. Time from injury to initiation of treatment did not differ significantly between the old and young age groups. Forty-four percent of elderly patients were evaluated within I hour after injury. After admission to the hospital, diagnostic studies were obtained and hematoma evacuation was initiated after a mean time interval of 198 _ minutes (similar to that observed with younger patients). Outcomes in the elderly group were poor, with a mortality rate of 74% and a 9% functional survival rate. The only elderly functional survivors presented with a minor closed head injury (GCS score 13 to 15). Even in this group, however, the mortality rate was 29%. Poor outcomes in the elderly group in the current study are similar to those reported recently in a smaller group of elderly patients with acute SDH by Amacher and Bybee,~ and appear relatively unchanged from pre reports. There are many possible explanations for the poor outcomes noted in elderly patients. Older patients may have a more destructive injury than young patients; acute SDH volumes in old patients were on average more than four times larger and caused twice the amount of midline shift as those of young patients. Acute SDH's in young patients with severe closed head injuries may often be an epiphenomenon, whereas in elderly patients the mass itself is probably the most important pathological process. With this in mind, rapid surgical decompression may be the most important aspect of treatment for elderly patients, and the time intervals from ER admission to start of surgery in the current series may have been too long. An important factor to consider is the possibility that aging brains have impaired regenerative capacity.'7 Experimental results have shown, for instance, that the temporal pattern and quantity of neurotrophic factors released after brain injuries in rats changes significantly with advanced age. 32 In the current series, no elderly patient whose neurological status deteriorated below a GCS score of 13 recovered to become a functional survivor. Because several clinical and radiographic features differ significantly between the two age groups, a single factor primarily responsible for the divergent outcomes cannot be identified. When interpreting the current results, methodological limitations should also be considered. Patients who were candidates for inclusion in the study were identified from a computer registry of discharge diagnoses. Errors during discharge summary dictation or computer data entry may have caused eligible patients to be excluded. Missing CT morphometric measurements and absent time interval data also create the potential 862 J. Neurosurg. / Volume 71/December, 1989

6 Acute subdural hematomas and age for sampling error. With these caveats in mind, several conclusions can be drawn. Acute SDH in young patients usually occurs in combination with a direct parenchymal brain injury and, in many cases, the acute SDH may be an epiphenomenon. Rapid in-the-field intubation and resuscitation followed by craniotomy several hours later has resulted in good outcomes in these patients. Outcome statistics in the young age group compare favorably with historic control data. Older patients typically develop large acute SDH after simple falls. Rapid medical management followed by surgical evacuation of the SDH are associated with poor outcomes similar to those reported in earlier studies. Among other possible explanations, poor outcomes in the older age group may reflect impaired recuperative capacity of the aging brain, a predisposition to develop a more lethal type of injury, or excessive time delay to hematoma evacuation. References 1. Amacher AL, Bybee DE: Toleration of head injury by the elderly. Neurosurgery 20: , Becker DP, Miller JD, Ward JD, et al: The outcome from severe head injury with early diagnosis and intensive management. J Neurosurg 47: , Britt RH, Hamilton RD: Large decompressive craniotomy in the treatment of acute subdural hematoma. Neurosurgery 2: , Cooper PR, Rovit RL, Ransohoff J: Hemicraniectomy in the treatment of acute subdural hematoma: a re-appraisal. Surg Neurol 5: Fell DA, Fitzgerald S, Moiel RH, et al: Acute subdural hematomas. Review of 144 cases. J Neurosurg 42: 37-42, Fogelholm R, Heiskanen O, Waltimo O: Chronic subdural hematomas in adults. Influence of patient's age on symptoms, signs, and thickness of hematoma. J Neurosurg 42:43-46, Galbraith JG: Subdural hematoma -- acute and chronic: a reappraisal. Clin Neurosurg 29:24-31, Gennarelli TA, Adams JH, Graham DI: Acceleration induced head injury in the monkey. I. The model, its mechanical and physiological correlates. Acta Neuropathol Suppl 7:23-25, Gennarelli TA, Spielman GM, Langfitt TW, et al: Influence of the type of intracranial lesion on outcome from severe head injury. J Neurosurg 56:26-32, Gennarelli TA, Thibault LE: Biomechanics of acute subdural hematoma. J Trauma 22: , Gennarelli TA, Thibault LE, Adams JH, et al: Diffuse axonal injury and traumatic coma in the primate, in Dacey RG Jr, Winn HR, Rimel RW, et al (eds): Trauma of the Central Nervous System. New York: Raven Press, 1985, pp Hernesniemi J: Outcome following acute subdural haematoma. Acta Neurochir 49: , Hernesniemi J: Outcome following head injuries in the aged. Acta Neurochir 49:67-79, Jennett B, Bond M: Assessment of outcome after severe brain damage. A practical scale. Lancet 1: , Klun B, Fettich M: Factors determining prognosis in acute subdural haematoma III. Acta Neuroehir Suppl 28: , Klun B, Fettich M: Factors influencing the outcome in acute subdural haematoma. A review of 330 cases. Acta Neurochir 71: , Long DM: Aging in the nervous system. Neurosurgery 17: , Marshall LF, Becker DP, Bowers SA, et al: The National Traumatic Coma Data Bank. Part 1: design, purpose, goals, and results. J Neurosurg 59: , Miller JD, Sweet RC, Narayan R, et al: Early insults to the injured brain. JAMA 240: , Moiel RH, Caram PC: Acute subdural hematomas. A review of eighty-four cases, a six year evaluation. J Trauma 7: , Morantz RA, Abaci RM, George AE, et al: Hemicraniectomy for acute extracerebral hematoma: an analysis of clinical and radiographic findings. J Neurosurg 39: , Perlmutter I: Subdural hematoma in older patients. JAMA 176: , Richards T, Hoff J: Factors affecting survival from acute subdural hematoma. Surgery 75: , Rogers LA, Sternbergh CA, Clark K: Subdural hematomas in the aged. South Med J 65: , Rosenorn J, Gjerris F: Long-term follow-up review of patients with acute and subacnte subdural hematomas. J Neurosurg 48: , Sahuquitlo-Barris J, Lamarca-Ciuro J, Vilalta-Castan J, et al: Acute subdural hematoma and diffuse axonal injury after severe head trauma. J Neurosurg 68: , Seelig JM, Becker DP, Miller JD, et al: Traumatic acute subdural hematoma. Majority mortality reduction in comatose patients treated within four hours. N Engl J Med 304: , Shigemori M, Syojima K, Nakayama K, et al: The outcome from acute subdural haematoma following decompressive hemicraniectomy. Acta Neuroehir 54:61-69, Stone JL, Lowe RJ, Jonasson O, et al: Acute subdural hematoma: direct admission to a trauma center yields improved results. J Trauma 26: , Talalla A, Morin MA: Acute traumatic subdural hematoma: a review of one hundred consecutive cases. J Trauma 11: , Teasdale G, Jennett B: Assessment of coma and impaired consciousness. A practical scale. Lancet 2:81-83, Whittemore SR, Nieto-Sampedro M, Needles DL, et al: Neuronotrophic factors for mammalian brain neurons: injury induction in neonatal, adult and aged rat brain. Dev Brain Res 20: , Yamashima T, Friede RL: Why do bridging veins rupture into the virtual subdural space? J Neurol Neurosurg Psychiatry 47: , 1984 Manuscript received February 27, Accepted in final form May 25, Address for Dr. Dacey: Division of Neurological Surgery, University of North Carolina School of Medicine, Chapel Hill, North Carolina. Address reprint requests to: Matthew A. Howard III, M.D., Department of Neurological Surgery, University of Washington, Harborview Medical Center, 325 Ninth Avenue, Seattle, Washington J. Neurosurg. / Volume 71 /December,