One of the most controversial issues in critical care of severely head-injured patients is the management of mean arterial blood pressure (MAP) or cerebral perfusion pressure (CPP). Although there is convincing evidence that hypotension is deleterious to the injured brain, it is unknown whether increasing CPP to more than the minimum level that provides adequate tissue perfusion improves outcome or whether it is detrimental (1).
CPP-directed management, advocated by Rosner et al. (2), includes the use of vasopressors to maintain a CPP of 70 mm Hg or higher. If cerebral autoregulation is preserved, this type of management may offer protection against secondary ischemic insults and help to control increased intracranial pressure (ICP) by preventing the occurrence of a vasodilatory cascade. However, in a prospective randomized trial, Robertson et al. (3) found that a CPP-targeted treatment protocol prevented secondary ischemic insults but did not reduce intracranial hypertension (ICHT) or improve clinical outcome. Furthermore, the use of vasopressors and fluid loading to augment MAP may predispose to cardiorespiratory complications (3–6). Nonetheless, CPP-directed management is widely used, and it has been included in international head injury guidelines as a treatment option (supported by Class III evidence and expert opinion) (7,8).
Another therapeutic concept, proposed by clinicians from Lund (Sweden), emphasizes a reduction in microvascular hydrostatic pressures to minimize edema formation in the brain (9). These authors postulate that if cerebral vasomotor responses are impaired and the blood-brain barrier is disrupted, increased MAP can be transmitted to the capillary bed, and vascular engorgement, transcapillary fluid filtration, and ICHT may be aggravated (10). Consequently, this type of management includes measures to reduce MAP, allowing CPP to decrease to 50 mm Hg.
Considering these therapeutic extremes, a rational approach might be to individually target CPP management depending on cerebrovascular autoregulation characteristics. As a first step toward this approach, we explored the effects of varying levels of CPP on ICP, cerebral pressure-flow autoregulation, and brain tissue oxygenation in severely head-injured patients without intracranial mass lesions during the first days after trauma. We analyzed separately data from patients with normal and increased ICP, because the presence of ICHT is the most important factor for guiding clinical decision making.
This study was approved by our institutional ethics committee, and written, informed consent was obtained from the patients’ next of kin. Patients with acute severe closed-head injuries (Glasgow Coma Scale score ≤8 or deteriorating to ≤8) who were admitted to the neurosurgical intensive care unit of our hospital were included in the study. Patients with intracranial mass lesions requiring surgery were not included. All patients were managed according to a protocol consistent with international guidelines (7,8). Propofol was used for sedation, with IV morphine 20–40 mg/d as an analgesic adjunct. Patients who had a CPP <70 mm Hg were routinely given a continuous IV infusion of norepinephrine to maintain their CPP above this threshold. When pyrexia evolved, cooling measures were taken to maintain temperature <38.0°C. Second-range therapy for ICHT included the sequential use of mannitol and moderate hyperventilation (guided by jugular venous oximetry). Third-range therapy included the use of large-dose barbiturates (guided by electroencephalography).
A multimodality neuromonitoring setup was used. ICP was measured by using an intraparenchymal pressure transducer (Ventrix; Camino Neurocare, San Diego, CA) inserted through a three-lumen intracranial bolt in the right frontal region. Brain tissue partial pressure of oxygen (Ptio2) was measured with a Clark-type polarographic microcatheter probe (Licox oxygen probe; GMS, Kiel, Germany). Temperature was measured from a thermocouple inserted through the same bolt. Care was taken to insert the Licox oxygen probe into a noncontused nonhemorrhaged area. Jugular bulb venous oxygen saturation (Sjo2) was measured continuously by using a fiberoptic catheter (5.5F Opticath Oximetrix; Abbott Critical Care Systems, Chicago, IL) inserted retrogradely into the right internal jugular vein. The catheter was calibrated at regular intervals. Proper positioning of the catheter tip was verified by a lateral skull radiograph. Peak flow velocity in the middle cerebral artery was measured bilaterally by pulsed Doppler ultrasound (2-MHz probe; Multi-Dop T; DWL, Sipplingen/Bodensee, Germany). The probes were secured by a metal head frame to obtain a constant insonation angle. Data were sampled at 250 Hz and subsequently averaged and stored at 0.1 Hz for offline analysis by using the POLY Physiological Analysis Package (Inspektor Research Systems, Amsterdam, The Netherlands).
Cerebral hemodynamic responses to varying levels of CPP were assessed by pharmacological manipulation of MAP on Days 0, 1, and 2 after injury. Changes in ventilator settings and the use of mannitol or sedative bolus infusions were avoided immediately before and during testing. The intended MAP and CPP changes were sinusoidal; starting from a steady-state, CPP was first decreased to a target of 50–55 mm Hg by reducing any baseline IV norepinephrine infusion or by using a combination of phentolamine and esmolol IV push. CPP was then slowly increased to a target of 105–110 mm Hg by using a ramped IV norepinephrine infusion. For analysis, only data stored during the upward titration of CPP were used, to avoid problems with hysteresis due to signal response times. One hour after the MAP manipulation, a three-step CO2 reactivity test was performed to enable a general classification of a subject’s cerebrovascular responsiveness. For this, the ventilator settings were adjusted to obtain subsequent expired minute volumes of approximately 50%, 100%, and 150% of pretest values. All settings were maintained for 15 min. Sjo2 was not allowed to decrease to ischemic values (<50%) during MAP or Paco2 manipulation.
Mean cerebral blood flow velocity (CBFV) was calculated from the time-averaged middle cerebral artery peak flow velocities of the left and right hemispheres. In preparation for further data analysis, both CBFV and ICP data scatter was removed by using a repeated moving average smoothing function (T4253H smoothing; SPSS software; SPSS Inc., Chicago, IL). Ptio2 is a more stable signal, and smoothing was not necessary. After preprocessing, all data were resampled to obtain balanced datasets with Y-values at 5-mm Hg CPP intervals (Fig. 1, top). For each 5-mm Hg CPP interval, a static rate of regulation (sROR) was calculated as the ratio of the percentage change in cerebrovascular resistance by the percentage change in CPP, with cerebrovascular resistance estimated as the ratio of CPP to CBFV (Fig. 1, bottom) (11). Static ROR values close to 1 indicate perfect adaptation of cerebral resistance vessels to changes in CPP, whereas values close to 0 indicate the complete absence of a cerebrovascular response. CO2 responsiveness was calculated as the ratio of percentage change in CBFV to change in Paco2. A CO2 responsiveness >2.7%/mm Hg was considered normal (12).
There is a strong linear relationship between Ptio2 and arterial partial pressure of oxygen (Pao2) when Pao2 is in the normal or hyperoxic range (13). Therefore, a method of Ptio2 correction was used to compensate for variations in Pao2 within and between patients (14). First, baseline Ptio2 and Pao2 were recorded midway during the 1-h interval between MAP manipulation and CO2 reactivity testing. The inspired fraction of oxygen was then increased to 1.0, and after a 15-min period of equilibration, Ptio2 and Pao2 were again recorded. The Ptio2 and Pao2 values at the different inspired fraction of oxygen levels were used to calculate a linear regression equation relating Ptio2 to Pao2. By using this equation, all Ptio2 values were subsequently standardized to a Pao2 of 100 mm Hg.
Data were analyzed separately on the basis of the presence or absence of ICHT (ICP >20 mm Hg) at the start of the test protocol. Differences in pretest variables on subsequent assessment days were evaluated with nonparametric tests for dependent samples (Friedman’s test for continuous variables; Cochran’s test for discrete variables). As the primary analysis, individual response curves were constructed for each study session relating ICP, sROR, and Ptio2 to the induced changes in CPP (Fig. 1). Similarly, response curves to the induced changes in Paco2 were constructed. The sign test was used to assess the direction of changes in these variables from pretest values to the values attained at the upper and lower limits of the explored range within subjects. This test was considered appropriate because single patients inevitably contributed data to either the normal or high ICP groups more than once, and the sign test is the most conservative (least powerful) of available nonparametric paired-samples tests. Values are reported as medians and 25th–75th percentiles.
Three sessions of CPP manipulation were scheduled in 13 patients on Days 0, 1, and 2 after injury. Of 39 intended test sessions, data from 37 sessions were available for analysis; 1 patient missed an evaluation because of severe unstable ICHT and 1 patient because of a technical problem with data acquisition. During the test sessions, a median baseline MAP of 94 mm Hg (87–100 mm Hg) was decreased to 68 mm Hg (62–79 mm Hg) and increased to 126 mm Hg (119–132 mm Hg) in a median time of 15 min (11–18 min). This corresponded to a low CPP of 51 mm Hg (48–53 mm Hg) and a high CPP of 108 mm Hg (102–112 mm Hg).
Table 1 shows general demographic and clinical variables. Patients had predominantly isolated head injuries. Five patients showed signs of increased ICP on their admission computed tomography scans, without evidence of mass lesions. Table 2 shows the physiological and treatment-related variables of the patients on succeeding days. ICHT became more prevalent as the time after injury elapsed, although this effect did not reach statistical significance. This trend was accompanied by increases in CBFV, Sjo2, and Ptio2 over time. In most patients, CPP was maintained well above 70 mm Hg during the first days in the intensive care unit, but this required increasing infusion rates of catecholamines over time.
For all 37 study sessions, ICHT was absent on 26 occasions in 11 patients, corresponding to a pretest ICP of 12 mm Hg (9–16 mm Hg), and ICHT was present on 11 occasions in 6 patients, corresponding to a pretest ICP of 24 mm Hg (22–29 mm Hg). Figure 2 shows cerebral hemodynamic responses to the trial CPP manipulation for patients with normal ICP and ICHT. In the absence of ICHT, there were no clinically significant changes in ICP, sROR, or Ptio2, although the sign test indicated consistent but very small increases in ICP and Ptio2, as well as a small decrease in sROR during the phase of increased CPP. Conversely, in the presence of ICHT, CPP reduction resulted in a significant increase in ICP, a decrease in sROR, and a decrease in Ptio2, whereas CPP increase beyond its pretest level resulted in unchanged ICP but an increase in sROR (statistically not significant) and Ptio2. Sjo2 values less than 50% were not observed during testing either in the presence or absence of increased ICP. Response patterns similar to Figure 2 could be constructed by using MAP (instead of CPP) across the independent axis (figures not shown). Also, the aggregated response patterns shown in Figure 2 represented individual responses well in most patients.
Figure 3 shows the results of 28 CO2 reactivity tests in 11 of the 13 patients. CBFV responsiveness was 3.3% (2.8%–4.0%) and 3.2% (2.8%–3.7%) per millimeter of mercury of Paco2 change in the presence and absence of ICHT, respectively. Abnormal CO2 responsiveness was observed in only one patient during all three test sessions. This patient had a normal ICP and a good clinical outcome. Whereas CBFV decreased linearly with decreases in Paco2, Ptio2 did not decrease until the last (hypocapnic) stage of the CO2 reactivity test.
In this study we varied CPP on the first days after severe head injury over a range of approximately 57 mm Hg. A temporary reduction of CPP from a baseline of approximately 75–80 mm Hg in patients with ICHT further increased ICP, reduced the capacity to autoregulate, and decreased brain tissue oxygenation. Increasing CPP improved sROR and increased Ptio2. In contrast, when ICP was normal, only small changes were observed when CPP was either increased or reduced.
Patients with increased ICP in this series appeared to be operating on the lower shoulder of the cerebral autoregulation curve, implying that their individual curves were shifted toward higher CPP values compared with patients with normal ICP. Although the lower acceptable limit for CPP has often been set at 70 mm Hg (1,7), considerably higher CPP has also been recommended for these patients (2). In contrast, patients with normal ICP appeared to be operating within autoregulation limits. Consequently, the use of catecholamines to increase CPP to an arbitrary value in this group may not improve cerebral perfusion while still subjecting patients to an increased risk of adult respiratory distress syndrome and circulatory failure (3–6). In relation to this, in 2003 the Brain Trauma Foundation issued an update of their guidelines for CPP management, decreasing the lower limit of CPP that is considered acceptable from 70 to 60 mm Hg (15). Our findings may not be generalizable to all groups of head-injured patients. Impaired cerebral vasomotor reactivity has been reported in the presence of ICHT and is associated with poor outcome (12,16). Because all patients with increased ICP had preserved CO2 responsiveness, patients with the worst injuries may have been under-represented in the study sample. Furthermore, we assessed only the immediate effects of CPP manipulation. Consequently, only a therapeutic trial of long-term CPP increases in patients with ICHT can assess the effects on the cerebral resistance vessels, the blood-brain barrier, transcapillary fluid filtration, or, for that matter, patient survival and functional outcome.
Several authors have proposed that the state of cerebral autoregulation should be integrated into management plans to target CPP for individual patients after head injury (1,16–18). Patterns of injury and the associated impairments of cerebral vasomotor responses are highly variable, and it is unlikely that one CPP threshold will be optimal for all patients at all times. In this study, we observed marked changes in individual cerebrovascular responses over time that were consistent with a transition from an early state of hypoperfusion to a state of possible hyperemia on Days 1 or 2 in some patients (19). Nonetheless, for the entire group, we were not able to establish any fixed temporal pattern for these observations, and this is consistent with findings by others (16,20). Therefore, without true assessment of cerebral hemodynamic responses, one cannot make assumptions about the state of autoregulation in a particular patient.
Many studies assessing pressure-flow autoregulation used only a single static MAP increase across a limited range and reported the cerebral vasomotor response in a dichotomized manner as absent or preserved (11,16,21). This study suggests that it may be important to investigate cerebral hemodynamic responses repeatedly and across a wider range to find a CPP at which the autoregulation capacity is maximized. Presumably, that CPP would then identify a pressure level at which brain tissue perfusion is adequate.
Some groups advocate using the ICP response to spontaneously occurring changes in MAP to make inferences about optimal CPP (17,22). Although it is attractive because of its simplicity, one problem with this approach is that the response of ICP to MAP variations may differ between patients with normal and increased ICP (20). The magnitude of the change in ICP then reflects the starting point on the ICP-volume curve, rather than the strength of the vasomotor response. This is illustrated by the exaggerated ICP changes in the group with ICHT in this study. Furthermore, when using the ICP-MAP correlation to assess cerebrovascular responses, it is implicitly assumed that perfusion pressure is the independent variable and ICP is the response variable. This assumption may be flawed when using observational data obtained during routine clinical practice, because MAP may respond to independent ICP changes (e.g., due to medullary vasomotor responses or the administration of vasopressors to maintain CPP), or both variables may simultaneously respond to a shared cause (e.g., CO2 retention and stress during endotracheal suctioning). Similar problems may also apply to studies that use more sophisticated monitoring, such as Ptio2, to assess cerebral autoregulation in an observational setting. Spontaneous fluctuations in CPP may be caused by variations in MAP, ICP, or both, and it is conceivable that unchanged cerebral hemodynamic conditions cannot be assumed in all such instances; this would invalidate inferences about optimal CPP.
Finally, when studying cerebral vasomotor responses to make inferences about a presumed optimal CPP, it is necessary to collect data on the pressure limits of autoregulation, rather than on the overall slope of the pressure-flow association. Studies that have attempted to detect the plateau and thresholds of the autoregulation curve have often pooled repeated observations from many different subjects within CPP ranges (18,23). Variations within and between subjects were thereby ignored. Recently, Lang et al. (18) used pharmacological manipulation of MAP and assessed cerebral autoregulation on 66 occasions in 14 head-injured patients with predominantly low ICP. After a pooled analysis of all observations, an autoregulatory plateau was recognized when CPP was between 70 and 90 mm Hg. However, the observations were made throughout two weeks after injury that had resulted in various combinations of epidural hematoma, subdural hematoma, and contusions. Pooling data relies on the assumption that the autoregulation curve is a population characteristic. Evidence against this assumption is provided by the differences between patients with normal and increased ICP in the present data and by other studies showing that the autoregulation curve can shift with changes in metabolism and intracranial pathophysiology (24).
We conclude that, in this study of severely head-injured patients with globally preserved cerebrovascular reactivity, pressure-flow autoregulation and tissue oxygenation were not critically dependent on CPP when ICP was within a normal range. In contrast, when ICHT was present, deliberate arterial hypertension reduced ICP, enhanced the capacity of the brain to autoregulate blood flow, and improved cerebral tissue oxygenation. More importantly, these data show that the technique of challenging the cerebral circulation by MAP manipulation over a wide range is feasible and can be used routinely to verify the response of ICP during CPP management in the first days after severe head injury. If more extensive neuromonitoring is available, a careful observation of individual responses to a daily trial manipulation of MAP can provide additional information that may be used to optimize CPP management.
We thank Dr. John Drummond for valuable suggestions and comments while reviewing the manuscript.
1. Robertson CS. Management of cerebral perfusion pressure after traumatic brain injury. Anesthesiology 2001;95:1513–7.
2. Rosner MJ, Rosner SD, Johnson AH. Cerebral perfusion pressure: management protocol and clinical results. J Neurosurg 1995;83:949–62.
3. Robertson CS, Valadka AB, Hannay HJ, et al. Prevention of secondary ischemic insults after severe head injury. Crit Care Med 1999;27:2086–95.
4. Contant CF, Valadka AB, Gopinath SP, et al. Adult respiratory distress syndrome: a complication of induced hypertension after severe head injury. J Neurosurg 2001;95:560–8.
5. Cremer OL, Moons KG, Bouman EA, et al. Long-term propofol infusion and cardiac failure in adult head-injured patients. Lancet 2001;357:117–8.
6. Kassell NF, Peerless SJ, Durward QJ, et al. Treatment of ischemic deficits from vasospasm with intravascular volume expansion and induced arterial hypertension. Neurosurgery 1982;11:337–43.
7. The Brain Trauma Foundation. The American Association of Neurological Surgeons: The Joint Section on Neurotrauma and Critical Care—guidelines for cerebral perfusion pressure. J Neurotrauma 2000;17:507–11.
8. Maas AI, Dearden M, Teasdale GM, et al. EBIC-guidelines for management of severe head injury in adults: European Brain Injury Consortium. Acta Neurochir (Wien) 1997;139:286–94.
9. Eker C, Asgeirsson B, Grände PO, et al. Improved outcome after severe head injury with a new therapy based on principles for brain volume regulation and preserved microcirculation. Crit Care Med 1998;26:1881–6.
10. Kongstad L, Grande PO. Arterial hypertension increases intracranial pressure in cat after opening of the blood-brain barrier. J Trauma 2001;51:490–6.
11. Panerai RB. Assessment of cerebral pressure autoregulation in humans: a review of measurement methods. Physiol Meas 1998;19:305–38.
12. Klingelhofer J, Sander D. Doppler CO2
test as an indicator of cerebral vasoreactivity and prognosis in severe intracranial hemorrhages. Stroke 1992;23:962–6.
13. van Santbrink H, Maas AIR, Avezaat CJJ. Continuous monitoring of partial pressure of brain tissue oxygen in patients with severe head injury. Neurosurgery 1996;38:21–31.
14. Johnston AJ, Steiner LA, Chatfield DA, et al. Effects of propofol on cerebral oxygenation and metabolism after head injury. Br J Anaesth 2003;91:781–6.
15. Brain Trauma Foundation. Update notice: guidelines for the management of severe traumatic brain injury: cerebral perfusion pressure—5-1-2003. Available at: http://www2.braintrauma.org/guidelines./
Accessed: June 17, 2004.
16. Lee JH, Kelly DF, Oertel M, et al. Carbon dioxide reactivity, pressure autoregulation, and metabolic suppression reactivity after head injury: a transcranial Doppler study. J Neurosurg 2001;95:222–32.
17. Steiner LA, Czosnyka M, Piechnik SK, et al. Continuous monitoring of cerebrovascular pressure reactivity allows determination of optimal cerebral perfusion pressure in patients with traumatic brain injury. Crit Care Med 2002;30:733–8.
18. Lang EW, Czosnyka M, Mehdorn HM. Tissue oxygen reactivity and cerebral autoregulation after severe traumatic brain injury. Crit Care Med 2003;31:267–71.
19. Martin NA, Patwardhan RV, Alexander MJ, et al. Characterization of cerebral hemodynamic phases following severe head trauma: hypoperfusion, hyperaemia, and vasospasm. J Neurosurg 1997;87:9–19.
20. Ter Minassian A, Dube L, Guilleux AM, et al. Changes in intracranial pressure and cerebral autoregulation in patients with severe traumatic brain injury. Crit Care Med 2002;30:1616–22.
21. Mascia L, Andrews PJ, McKeating EG, et al. Cerebral blood flow and metabolism in severe brain injury: the role of pressure autoregulation during cerebral perfusion pressure management. Intensive Care Med 2000;26:202–5.
22. Lang EW, Lagopoulos J, Griffith J, et al. Cerebral vasomotor reactivity testing in head injury: the link between pressure and flow. J Neurol Neurosurg Psychiatry 2003;74:1053–9.
23. Czosnyka M, Smielewski P, Piechnik S, et al. Cerebral autoregulation following head injury. J Neurosurg 2001;95:756–63.
© 2004 International Anesthesia Research Society
24. Paulson OB, Strandgaard S, Edvinsson L. Cerebral autoregulation. Cerebrovasc Brain Metab Rev 1990;2:161–92.