Cardiopulmonary Hemodynamic Changes During Acute Subdural Hematoma Evacuation

Neurol Med Chir (Tokyo) 46, 219¿225, 2006

Tomonori TAMAKI,YojiNODE,YasuhiroYAMAMOTO*,andAkiraTERAMOTO

Departments of Neurosurgery and *Critical Care Medicine and Emergency, Nippon Medical School, Tokyo

Abstract

The aim of this study was to clarify the mechanism of hemodynamic changes leading to intraoperative hypotension during evacuation of acute subdural hematoma. To our knowledge, little data is available about the mechanism of hemodynamic changes during surgical interventions to decrease intracranial pressure after severe head injury. The influence of preoperative hypotension on intraoperative hypotension was examined. Hemodynamic studies (pulmonary artery catheterization) were carried out in 15 patients before and after acute subdural hematoma evacuation. All patients were assessed for hemodynamic parameters, evacuated hematoma volume, and intracranial pressure measurements. Comparison between just before and after evacuation of the hematoma showed that the mean arterial pressure, pulmonary arterial pressure, systemic vascular resistance, pulmonary vascular resistance, central venous pressure, and pulmonary capillary wedge pressure all decreased after hematoma evacuation. However, the cardiac index was unchanged after hematoma evacuation. Mean arterial blood pressure is dependent on the cardiac index and vascular resistance, so the decrease in arterial blood pressure during hematoma evacuation was the result of a decline in vascular resistance. The influence of preoperative blood pressure on intraoperative hemodynamic changes was analyzed by dividing the patients into two groups, the preoperative hypotension group and preoperative nonhypotension group. The decrease in mean arterial blood pressure was more marked in the preoperative hypotension group than in the preoperative nonhypotension group. Intraoperative hypotension during evacuation of acute subdural hematoma is caused by a decrease in vascular resistance. Preoperative hypotension is a also risk factor for intraoperative hypotension.

Key words: severe head injury, pulmonary artery catheter, systemic vascular resistance, cardiac index, acute subdural hematoma

Introduction decrease in oxygen delivery because of the high metabolic rate and absence of substrate storage.11,16) Patients with severe head injury often require early The effect of prehospital hypotension on the long- surgical intervention which carries the risk of term outcome after head injury is well documented, developing hypoxia or hypotension.12,17,18) Evidence but the importance of intraoperative hypotension is of secondary ischemic brain damage was found in unclear.5,21) Cerebral perfusion pressure (CPP) is 66% of the patients who died of head injury, important to maintain in patients with head injuries suggesting that ischemia may be very common in the intensive care unit, so a similar policy in the during hospital treatment.18) Intraoperative hypoten- operating room would seem prudent.14) Accord- sion occurs in 32% of patients with severe head ingly, the hemodynamic mechanism of intraopera- trauma initially without hypotension, and intraoper- tive hypotension is important to understand. In ative hypotension is inversely correlated with the certain physiological states, mean arterial blood Glasgow Outcome Scale.10) Hypotension reduces the pressure (MABP) is dependent on the cardiac index delivery of substrates (oxygen and glucose) to the and systemic vascular resistance.4) Therefore, the brain tissue, which is especially vulnerable to a decrease in MABP during surgical intervention for

Received March 7, 2005; Accepted December 2, 2005 Author's present address: T. Tamaki, M.D., Department of Neurosurgery, Nippon Medical School Nagayama Hospital, Tama, Tokyo, Japan.

219 220 T. Tamaki et al. severe head injury may be related to these left ventricular stroke work index, right ventricular parameters. stroke work index, and CPP using standard formu- The present study investigated the hemodynamic lae. changes which occurred during evacuation of acute The hemodynamic parameters and ICP were subdural hematoma by pulmonary artery catheteri- measured immediately before hematoma evacua- zation in patients with and without preoperative tion, defined as pre-evacuation data, and immedi- hypotension. ately after hematoma evacuation, defined as post- evacuation data. ICP monitoring and pulmonary Material and Methods artery catheter measurements were carried out at the same time. Intraoperative hypotension was This study included 15 patients with severe head defined as a decrease in MABP by more than 10 injury (Glasgow Coma Scale º8) admitted to our mmHg during surgery (post-evacuation MABP － institution within 60 minutes of injury between pre-evacuation MABP). Preoperative hypotension January and June 1997. On admission, all patients was defined as a pre-evacuation systolic blood pres- hadisolatedheadinjurywithnoevidenceof sure of less than 90 mmHg. The influence of pulmonary contusion or aspiration of gastric con- preoperative MABP on intraoperative hemodynam- tents. All patients underwent endotracheal intuba- ic changes was analyzed by classifying the patients tion and mechanical ventilation was continued into the preoperative hypotension group and during evaluation in the Emergency Department. preoperative nonhypotension group. Data were Arterial blood gas analysis was performed regularly analyzed statistically using Student's paired t-test and mean PaCO2 was maintained at 32.8 ± 0.8 and significance was accepted at p º 0.05. Correla- mmHg. All patients were given acetated Ringers tions were determined by calculating Pearson's solution at 0.01–0.03 ml/mg/hr immediately after correlation coefficients. Regression used the least admission. If the systolic blood pressure was less squares method. than 90 mmHg 20 minutes after starting infusion, dopamine was administered at 0.005 ml/mg/min. Results The definitive intracranial diagnosis was estab- lished based on computed tomography findings The clinical characteristics of the patients are given evaluated using traumatic coma data bank catego- in Table 1. The mean time from injury to hematoma ries. All patients had acute subdural hematoma in evacuation was 115 ± 32 minutes. Ten of the 15 the traumatic coma data bank category of mass patients experienced intraoperative hypotension. lesion.8) All patients underwent immediate surgery However, there was almost no intraoperative bleed- with burr holes under local anesthesia. A fiberoptic ing in all patients. Eight patients received dopamine subdural intracranial pressure (ICP) monitor support because of the marked decrease in systolic (Camino ICP monitoring system; Integra Neu- blood pressure. Hemodynamic data obtained during roScience, Andover, England) was inserted. Then a hematoma evacuation are shown in Table 2. The silicone drain tube was positioned and the acute mean time between pre-evacuation and post-evacua- subdural hematoma was evacuated slowly.20) The tion measurement of parameters was 14.3 minutes. evacuated hematoma volume was measured. In the preoperative hypotension group, the MABP, The radial artery and pulmonary artery pressures, mean pulmonary artery blood pressure, systemic central venous pressure, and pulmonary capillary vascular resistance, pulmonary vascular resistance, wedge pressure were obtained directly from pres- central venous pressure, pulmonary capillary wedge sure transducers connected to arterial catheters and pressure, and ICP all significantly decreased after a pulmonary artery catheter, which was positioned evacuation. In the preoperative nonhypotension by the Seldinger technique. Cardiac output was group, only central venous pressure showed a sig- measured in triplicate using a cardiac output nificant decrease during surgery (Fig. 1). There was computer (COM-1 9310; Baxter Edwers Critical a weak correlation between the intraoperative Care, Santa Ana, Calif., U.S.A.) after a 5-ml bolus of change in MABP and the intraoperative hematoma 5% dextrose (º109C) was injected into the right evacuation volume (r ＝ 0.29, p ＝ 0.037), but there ventricle at the end-expiratory phase of the respira- was no correlation between the intraoperative tory cycle. The heart rate was monitored from the changes in MABP and ICP (r ＝ 0.036, p ＝ 0.501) R-wave of the electrocardiogram. From the meas- (Fig. 2). ured data, we calculated the MABP, mean pulmonary artery blood pressure, cardiac index, systemic vascular resistance, pulmonary vascular resistance,

Neurol Med Chir (Tokyo) 46, May, 2006 Hemodynamic Changes During Hematoma Evacuation 221

Table 1 Clinical characteristics of the patients

Intraopera- Intraopera- Intraopera- Evacuated Case Age Initial Initial tive MABP Intraopera- tive ICP tive CPP hematoma No. (yrs) Sex SBP GCS CT findings change tive change change volume Course (mmHg) score (mmHg) hypotension (mmHg) (mmHg) (ml)

172F 1555 ASDH,CC, －11.3 ＋－13.0 1.7 55 dead SAH 253M 1504 ASDH,CC 5.0 －－16.0 21.0 40 alive 323M 1673 ASDH,CC, 3.0 －－20.0 23.0 40 alive SAH 458F 1865 ASDH,AEDH, －30.3 ＋－13.0 －16.7 95 dead CC, SAH 524M 1205 ASDH,CC, －19.3 ＋－6.0 －13.3 30 dead SAH 6 46 F 156 3 ASDH, SAH －14.0 ＋－13.0 －1.0 55 alive 7 64 M 95 5 ASDH, SAH －14.3 ＋－9.0 －4.7 65 alive 838F 656 ASDH,AEDH, －23.3 ＋－9.0 －12.4 85 dead CC, SAH 945M 563 ASDH －23.0 ＋－8.0 －15.0 75 dead 10 66 F 45 7 ASDH, SAH －9.0 －－14.0 5.0 95 dead 11 43 M 67 5 ASDH, CC, 0 －－9.0 9.0 50 dead SAH 12 44 F 80 6 ASDH, CC, －18.7 ＋－19.0 －0.3 65 alive SAH 13 35 F 67 4 ASDH, CC －22.0 ＋－3.0 －19.0 30 dead 14 37 M 58 3 ASDH 8.0 －－4.0 12.0 35 dead 15 61 F 65 3 ASDH, CC, －16.3 ＋－8.0 －8.3 65 dead SAH

Intraoperative cerebral perfusion pressure (CPP) change＝post-evacuation CPP － pre-evacuation CPP, intraoperative intracranial pressure (ICP) change＝post-evacuation ICP － pre-evacuation ICP, intraoperative mean arterial blood pressure (MABP) change＝post-evacuation MABP － pre-evacuation MABP. AEDH: acute epidural hemorrhage, ASDH: acute subdural hemorrhage, CC: cerebral contusion, CT: computed tomography, GCS: Glasgow Coma Scale, SAH: traumatic subarachnoid hemorrhage, SBP: systolic blood pressure.

Table 2 Comparison of hemodynamic parameters during acute subdural hematoma evacuation

Pre-evacuation Post-evacuation p Value

Mean arterial blood pressure (mmHg) 68.9±34.2 54.8±38.0 0.0012* Mean pulmonary arterial blood pressure (mmHg) 19.5±6.1 17.5±5.4 0.0038* HR (beat/min) 101±33 110±33 0.0558 Cardiac index (l/min/m2)2.76±1.21 2.76±1.25 0.8646 Central venous pressure (mmHg) 3.5±1.7 2.3±1.0 0.0001* Pulmonary capillary wedge pressure (mmHg) 12.9±4.0 11.1±3.9 0.043* Pulmonary vascular resistance (dyne・sec・cm－5)413.6±116.0 372.7±103.4 0.0043* Systemic vascular resistance (dyne・sec・cm－5) 1242.2±236.6 890.6±311.5 0.0001* Left ventricular stroke work index (g・m/HR) 0.45±0.47 0.33±0.30 0.0956 Right ventricular stroke work index (g・m/HR) 0.11±0.11 0.09±0.07 0.1563 Intracranial pressure (mmHg) 30.7±10.8 19.0±8.2 º0.0001* Cerebral perfusion pressure (mmHg) 38.2±27.0 33.0±33.1 0.134

Data are expressed mean±standard deviation. *Significant difference between pre-evacuation and post-evacuation (pº0.05). HR: heart rate.

Discussion subjects included many patients with preoperative hypotension and a poor Glasgow Coma Scale score. Intraoperative hypotension occurred in 10 of the 15 Evacuation of acute subdural hematoma used a patients with acute subdural hematoma. This is a single burr hole under local anesthesia, and there higher incidence of intraoperative hypotension was little intraoperative bleeding. There was no compared with other reports, possibly because our relationship between intraoperative hypotension

Neurol Med Chir (Tokyo) 46, May, 2006 222 T. Tamaki et al.

Fig. 1 Comparison of hemodynamic parameters during acute subdural hematoma evacuation. Open circle indicates preoperative nonhypotension group (n ＝ 7) and solid circle indicates preoperative hypotension group (n ＝ 8). Data are expressed mean ± standard deviation. *Significant difference between pre-evacuation and post-evacuation (p º 0.05). HR: heart rate, NS: no significant difference between pre-evacuation and post-evacuation, Pre: pre- evacuation data, Post: post-evacuation data. and hypovolemia, so the onset of intraoperative nificantly altered. The preoperative nonhypotension hypotension was clearly caused by reduction in ICP group showed a slight decrease in MABP (not or evacuation of the hematoma. significant), but no other significant changes in The cardiac index and systemic vascular hemodynamic parameters apart from central venous resistance are known to modulate MABP.4) The pressure during hematoma evacuation. The preoper- occurrence of intraoperative hypotension in our ative hypotension group showed major changes to patients during evacuation of acute subdural the circulatory system during hematoma evacuation: hematoma resulted from a rapid decrease in the a marked decrease in MABP caused by decreased systemic vascular resistance (i.e., systemic vasodila- systemic vascular resistance, a decrease in the tion), whereas the cardiac function was not sig- mean pulmonary artery blood pressure caused by

Neurol Med Chir (Tokyo) 46, May, 2006 Hemodynamic Changes During Hematoma Evacuation 223

central circulation regulatory areas.1) An acute, rapid rise in the ICP to a level in excess of the systolic arterial pressure causes systemic hypertension and decreased heart rate (the Cushing response), and the pressor component of the Cushing response is wholly or partly related to cerebral ischemia.1–3,6,13,20,21) Hemodynamic changes related to a decrease in ICP during neurosurgical intervention may arise as an inverse Cushing response. If the Cushing response affected intraoperative blood pressure, patients without preoperative hypotension might develop severe intraoperative hypotension, because of more severe brain stem ischemia, rather than patients with preoperative hypotension. However, the changes in cardiac function after severe head trauma can be explained by a number of factors, such as the Cushing response, neurogenic shock, and heart failure induced by catecholamines.9,19,21) Catechola- mines cause myocardial necrosis and propranolol has a cardioprotective effect, but excessive levels of circulating catecholamines may induce heart failure in patients with severe head injury.9) Many patients with acute subdural hematoma suffer acute brain swelling intraoperatively and Fig. 2 Upper: Relationship between intraoperative control of the ICP becomes impossible, resulting in mean arterial blood pressure (MABP) death. Marked hypertension (systolic pressure À140 change and intraoperative intracranial mmHg) before surgery and intraoperative hypoten- pressure (ICP) change (MABP change ＝ sion (systolic pressure º90 mmHg) were risk factors －16.60 － 0.40 × ICP change, R2 ＝ 0.036). for acute brain swelling in patients with acute Lower: Relationship between intraoperative subdural hematoma.15) In our study, two of three MABP change and evacuated hematoma patients with these criteria died. Intraoperative volume (MABP change ＝ 5.19 － 0.29 × hypotension should be vigorously managed with 2 hematoma volume, R ＝ 0.313). Regres- fluids initially and with pharmacological support sion lines formed by the least squares (especially vasoconstrictor agents like alpha- method. agonists) to achieve rapid reversal of cerebral hypoperfusion. Furthermore, hematoma evacuation decreased pulmonary vascular resistance, and a should be performed slowly to achieve steady ICP decrease in central venous pressure and pulmonary reduction. Maintenance of the CPP should be the capillary wedge pressure caused by generalized goal of resuscitation in patients with severe head vasodilation. The intraoperative change in MABP injury, in both the preoperative and the intraopera- was correlated with the evacuated hematoma tive phases. Neurosurgical interventions, such as volume rather than the change in ICP. The reason hematoma evacuation and external decompression for this is unknown since the evacuated hematoma craniectomy, are aimed at reducing ICP and volume does not directly reflect the true hematoma elevating CPP in patients with severe head injury, volume because of obstruction to the drain tube or but reducing the ICP and evacuating a hematoma sequestration of parts of the hematoma by acute may cause deterioration in the hemodynamic brain swelling.7,15) parameters. An experimental head injury model suggested that heart rate and blood pressure responses were References probably initiated by deformation or local ischemia in the central nervous system regions that control 1) Brown RS, Mohr PA, Carey JS, Shoemaker CW: the circulatory system.6) AdecreaseinMABP Cardiovascular changes after cerebral injury and induced by increased ICP may be due to a decline of increased intracranial pressure. Surg Gynecol Obstet sympathetic stimulation because of damage to the 125: 1205–1211, 1967

Neurol Med Chir (Tokyo) 46, May, 2006 224 T. Tamaki et al.

2) Cushing H: Concerning a definite regulatory injury. Acta Neurochir (Wien) 54: 243–250, 1980 mechanism of the vasomotor centers which controls 18) Shackford SR, Mackersie RC, Davis JW, Wolf PL, blood pressure during cerebral compression. Bull Hoyt DB: Epidemiology and pathology of traumatic Johns Hopkins Hosp 12: 290, 1901 deaths occurring at a Level I Trauma Center in a 3) Fitch W, McDowall DG: Systemic vascular responses regionalized system: the importance of secondary to increased intracranial pressure. 1. Effects of brain injury. JTrauma29: 1392–1397, 1989 progressive epidural balloon expansion on in- 19) Shigemori M, Tokutomi T, Yuge T, Shojima K, tracranial pressure: and systemic circulation. J Matsuo H, Moriyama T, Kuramoto S: [Treatment of Neurol Neurosurg Psychiatry 40: 833–842, 1977 acute subdural hematoma with low GCS score]. No 4) Forrester JS, Diamond G, Chatterjee K, Swan HJ: Shinkei Geka 15: 1305–1310, 1987 (Jpn, with Eng Medical therapy of acute myocardial infarction by abstract) application of hemodynamic subsets (first of two 20) Tamaki T, Isayama K, Yamamoto Y, Teramoto A: parts). NEnglJMed295: 1356–1362, 1976 Cardiopulmonary haemodynamic changes after 5) Graham DI, Ford I, Adams JH, Doyle D, Teasdale severe head injury. Br J Neurosurg 18: 158–163, 2004 GM, Lawrence AE, McLellan DR: Ischaemic brain 21) Youmans JR: Causes of shock with head injury. J damage is still common in fatal non-missile head Trauma 128: 204–209, 1964 injury. J Neurol Neurosurg Psychiatry 52: 346–350, 1989 6) Kransley JA, Koehler RC: Heart rate and rhythm and intracranial pressure. Am J Physiol 230: 1965–1700, Address reprint requests to: T. Tamaki, M.D., Department 1976 of Neurosurgery, Nippon Medical School Nagayama 7) Lang AD, Teasdale MG, Macpherson P, Lawrence A: Hospital, 1–7–1 Nagayama, Tama, Tokyo 206–8512, Diffuse brain swelling after head injury: more often Japan. malignant in adults than children? J Neurosurg 80: e-mail: tamakito＠nms.ac.jp 675–680, 1994 8) Marshall LF, Marshall SB, Klauber MR, Van Berkum Clark M, Eisenberg H, Jane JA, Luerssen TG, Commentary Marmarou A, Foulkes MA: The diagnosis of head injury requires a classification based on computed While the occurrence of intraoperative hypotension axial tomography. JNeurotrauma9 Suppl 1: during removal of acute subdural hematoma is known S287–292, 1992 to exert detrimental effects on the patient's outcome, 9) Neil-Dwyer G, Walter P, Cruickshank JM, Doshi B, little is known about the mechanism of hemodynamic O'Gorman P: Effect of propranolol and phentolamine on myocardial necrosis after subarachnoid haemor- changes leading to intraoperative hypotension. In this rhage. Br Med J 2(6143): 990–992, 1978 study, 15 patients with a diagnosis of acute subdural 10) Pietropaoli JA, Rogers FB, Shackford SR, Wald SL, hematoma underwent detailed examination of the Schmoker JD, Zhuang J: The deleterious effects of hemodynamic changes before and after hematoma intraoperative hypotension on outcome in patients evacuation. Systemic blood pressure at admission, as with severe head injuries. JTrauma33: 403–407, 1992 well as the mean arterial blood pressure, intracranial 11) Popp AJ, Gottlieb ME, Paloski WH, Rahm RL, Newell pressure, and cerebral perfusion pressure during JC,ShahDM,BourkeRS:Cardiopulmonary operation were monitored, and the correlations hemodynamics in patients with serious head injury. J between these values and the occurrence of intraoper- Surg Res 32: 416–421, 1982 ative hypotension, or the final outcome (alive or dead) 12) Randall MC, Theresa G, Barbara AB, Melville RK, were examined. This analysis disclosed a significant Lawrence FM: Neurogenic hypotension in patients with severe head injuries. JTrauma44: 958–963, 1998 association between the preoperative hypotension 13) Richardson TQ, Fermoso JF, Pugh GO: Effect of and the occurrence of intraoperative hypotension. acutely elevated intracranial pressure on cardiac Regarding the mechanism underlying the above output and other circulatory factors. JSurgRes115: association, the hemodynamic parameters consistent- 319–322, 1965 ly showed that hematoma evacuation caused a 14) Rosner MJ, Daughton S: Cerebral perfusion pressure decrease in vascular resistance in the face of the management in head injury. JTrauma30: 933, 1990 unchanged cardiac index. Therefore, it is logical to 15) Saito T, Kushi H, Makino K, Hayashi N: The risk assume that the lower the preoperative blood pressure factors for the occurrence of acute brain swelling in is, the greater the decrease in the intraoperative acute subdural hematoma. Acta Neurochir Suppl 86: arterial pressure becomes. Although this is an 351–354, 2003 important point in the clinical handling of acute 16) Saul TG, Ducker TB: Management of severe head injuries. Md State Med J 30: 45–48, 1981 subdural pressure, it seems desirable to further inves- 17)SchulteAM,EschJ,MurdayH,PfeiferG: tigate the possible relationship between the occur- Hemodynamic changes in patients with severe head rence of preoperative hypotension and the severity as

Neurol Med Chir (Tokyo) 46, May, 2006 Hemodynamic Changes During Hematoma Evacuation 225 well as the topography of brain damage. Of the 15 patients, ten experienced intraoperative Takao ASANO,M.D. hypotension. Based upon the extensive hemodynamic Department of Neurosurgery monitoring, the authors concluded that the hypoten- Medical Research Center sion was due to the decreasing intracranial pressure Saitama Medical Center/School during hematoma evacuation. Their analysis deter- Saitama, Japan mined that this was due to a decrease in systemic vascular resistance at a time that cardiac function was not altered. This systemic vasodilation is interpreted The authors have investigated the phenomenon of by the authors as an inverse Cushing response and intraoperative hypotension during evacuation of warrants aggressive treatment with fluids and alpha- acute subdural hematomas in a study involving 15 agonists to maintain cerebral perfusion pressure. patients presenting with acute subdural hematoma, This important study lends further insight into the with Glasgow Coma Scale scores less than 8, admitted important hemodynamic factors influencing patients within 60 minutes of injury. These patients were with severe head injury. extensively evaluated for hemodynamic parameters, Daniel L. BARROW,M.D. including arterial blood pressure, pulmonary artery Department of Neurosurgery pressure, central venous pressure, pulmonary capilla- Emory University ry wedge pressure, heart rate, cardiac output, and Atlanta, Georgia, U.S.A. intracranial pressure.

Neurol Med Chir (Tokyo) 46, May, 2006