Cerebrovascular reactivity during hypothermia and rewarming

Cerebrovascular reactivity during hypothermia and rewarming

British Journal of Anaesthesia 99 (2): 237–44 (2007) doi:10.1093/bja/aem118 Advance Access publication May 16, 2007 Cerebrovascular reactivity durin...

366KB Sizes 1 Downloads 5 Views

British Journal of Anaesthesia 99 (2): 237–44 (2007)

doi:10.1093/bja/aem118 Advance Access publication May 16, 2007

Cerebrovascular reactivity during hypothermia and rewarming A. Lavinio1 3, I. Timofeev1, J. Nortje2, J. Outtrim2, P. Smielewski1†, A. Gupta2, P. J. Hutchinson1, B. F. Matta2, J. D. Pickard1, D. Menon2 and M. Czosnyka1*† 1

Department of Clinical Neurosciences, Academic Neurosurgical Unit and 2Department of Anaesthesiology, Addenbrooke’s Hospital, Cambridge, UK. 3Institute of Anaesthesiology and Intensive Care Medicine, University of Brescia, Brescia, Italy

*Corresponding author: Department of Clinical Neurosciences, Academic Neurosurgical Unit, Box 167, Addenbrooke’s Hospital, Cambridge, UK. E-mail: [email protected]

Methods. This is a retrospective analysis of data acquired during a prospective, observational neuromonitoring and imaging data collection project. Brain temperature, intracranial pressure (ICP), and cerebrovascular pressure reactivity index (PRx) were continuously monitored. Results. Twenty-four TBI patients with refractory intracranial hypertension were cooled from 36.0 (0.9) to 34.2 (0.5)8C [mean (SD), P,0.0001] in 3.9 (3.7) h. Induction of hypothermia [average duration 40 (45) h] significantly reduced ICP from 23.1 (3.6) to 18.3 (4.8) mm Hg (P,0.05). Hypothermia did not impair cerebral vasoreactivity as average PRx changed non-significantly from 0.00 (0.21) to 20.01 (0.21). Slow rewarming up to 37.08C [rate of rewarming, 0.2 (0.2)8C h21] did not increase ICP [18.6 (6.2) mm Hg] or PRx [0.06 (0.18)]. However, in 17 (70.1%) out of 24 patients, rewarming exceeded the brain temperature threshold of 378C. In these patients, the average brain temperature was allowed to increase to 37.8 (0.3)8C (P,0.0001), ICP remained stable at 18.3 (8.0) mm Hg (P¼0.74), but average PRx increased to 0.32 (0.24) (P,0.0001), indicating significant derangement in cerebrovascular reactivity. After rewarming, PRx correlated independently with brain temperature (R¼0.53; P,0.05) and brain tissue O2 (R¼0.66; P,0.01). Conclusions. After moderate hypothermia, rewarming exceeding the 378C threshold is associated with a significant increase in average PRx, indicating temperature-dependent hyperaemic derangement of cerebrovascular reactivity. Br J Anaesth 2007; 99: 237–44 Keywords: autoregulation; brain, injury; hypothermia; monitoring, critical care Accepted for publication: March 16, 2007

Although there is evidence of effectiveness of moderate hypothermia in the treatment of refractory intracranial hypertension, the impact of hypothermia on outcome after head injury is still open to debate. Some studies point out that the benefit of controlling intracranial hypertension may be outweighed by subsequent complications such as pneumonia.1 – 3 During the EuroNeuro 2005 congress, the question of whether hypothermia could also affect cerebral blood flow (CBF) autoregulation was raised. At the time, the impact of hypothermia and subsequent rewarming on human cerebrovascular reactivity after head injury was unknown. On

the other hand, cerebrovascular reactivity had already been well established as a strong predictor of outcome in headinjured patients,4 5 and continuous monitoring technique was readily available.6 Although autoregulation of CBF is preserved under mild hypothermia during general anaesthesia in healthy brain,7 murine studies point out that hypothermia and fast †

Declaration of interest. ICMþ software (www.neurosurg.cam.ac.uk/ icmplus) is licensed by University of Cambridge, UK and P.S. and M.C. have a financial interest in the licensing fee.

# The Board of Management and Trustees of the British Journal of Anaesthesia 2007. All rights reserved. For Permissions, please e-mail: [email protected]

Downloaded from http://bja.oxfordjournals.org/ at Seton Hall on April 6, 2015

Background. Experimental evidence from a murine model of traumatic brain injury (TBI) suggests that hypothermia followed by fast rewarming may damage cerebral microcirculation. The effects of hypothermia and subsequent rewarming on cerebral vasoreactivity in human TBI are unknown.

Lavinio et al.

rewarming may affect cerebral vasoreactivity after traumatic brain injury (TBI).8 We therefore addressed the hypothesis that hypothermia and subsequent rewarming may affect cerebral autoregulation after TBI in humans. In this report, we describe our observations regarding cerebral vasoreactivity in head-injured patients cooled for refractory intracranial hypertension.

Methods

Fig 1 Examples of (A) good pressure reactivity (negative PRx) and (B) disturbed pressure reactivity ( positive PRx). PRx was calculated as the moving correlation coefficient between slow waves of ICP and AP from a period of 4 min. Slow waves may be detected by low-pass filtering or simple moving averaging with a period of 5 s.

238

Downloaded from http://bja.oxfordjournals.org/ at Seton Hall on April 6, 2015

This is a retrospective analysis of data acquired as part of the ongoing prospective, observational neuromonitoring and imaging data collection project ‘Blood flow, swelling and metabolism after severe traumatic brain injury’. Data collection was approved by the Local Research Ethics Committee. All patients’ representatives agreed to the use of the collected data and signed the informed assent form. The setting was the Neurosciences Critical Care Unit at Addenbrooke’s Hospital in Cambridge, UK. The study included severely head-injured adults with acute intracranial hypertension [intracranial pressure (ICP) .20 mm Hg for a duration of 1 h] refractory to medical treatment. All patients were managed according to Advanced Trauma Life Support guidelines and to the protocol followed in our unit for severe head injury.9 This includes mechanical ventilation, sedation with propofol (3– 5 mg kg21 h21), fentanyl (1–2 mg kg21 h21), and paralysis with atracurium (0.5 mg kg21 h21). Before induction of hypothermia, infusion of propofol was replaced by infusion of midazolam (0.1–0.2 mg kg21 h21). Mean arterial pressure (AP) was manipulated with the use of fluid loading and inotropic drugs in order to achieve a cerebral perfusion pressure (CPP) .60 mm Hg. Patients had a cooling blanket (Blanketrol II, Cincinnati Sub Zero, OH, USA) applied until the target brain temperature ,34.58C was achieved. Moderate hypothermia (33–358C) was maintained with the blanket as clinically indicated in order to control ICP, and subsequently the patients were passively rewarmed. Mean AP was invasively monitored through a catheter in the radial artery; the pressure transducer was zeroed at heart level. ICP, pressure of brain tissue O2 (PtO2) and CO2 (PtCO2), pH, and brain temperature were continuously monitored with Codman/Neurotrend parenchymal probes (Johnson & Johnson Medical, Raynham, MA, USA) via a cranial access device (Technicam, Abbott, UK). Probes were positioned at a constant depth in the white matter, pericontusional in focal injuries or in the non-dominant frontal lobe in diffuse injuries. Probe positioning was verified by means of a head computed tomography (CT) scan. All data were digitalized and captured using a bedside computer, with a sampling rate of 30 Hz. Artifacts were removed manually. Daily investigations included C-reactive protein (CRP) and leucocyte count [white blood cells (WBC)]; CRP and WBC were averaged for each phase of the study that lasted .1 day.

Cerebral vasoreactivity was monitored using the cerebral pressure reactivity index (PRx), continuously calculated on bedside computers running ICMþ software for multimodal brain monitoring (‘Intensive Care Monitor’, University of Cambridge, UK; www.neurosurg.cam.ac.uk/icmplus). The PRx index is defined as the moving Pearson’s correlation coefficient (4 min window) between spontaneous slow waves (20 s to 3 min period) in averaged ICP and AP (5 s moving average).10 In simple terms, PRx implements continuous analysis of ICP and AP waves, and evaluates the capability of the intracranial arterioles to cope with spontaneous changes in AP. As an example, if autoregulation of CBF is intact and AP decreases, cerebral arterioles dilate in order to reduce vascular resistance and to maintain CBF constant. The arteriolar dilation leads to an increase in cerebral blood volume and, consequently, to an increase in ICP (Fig. 1A). Hence, in the case of preserved cerebral autoregulation, AP and ICP are negatively correlated and PRx is negative (i.e. when AP decreases, ICP increases and vice versa). On the contrary, when cerebrovascular reactivity is impaired, cerebral blood volume increases or decreases passively with changes in AP, and PRx is positive (i.e. when AP increases, so does ICP; Fig. 1B).

TBI, hypothermia, and cerebral vasoreactivity

To summarize, positive average PRx (.0.2) signifies disturbed pressure reactivity, whereas negative PRx (,0) implies good reactivity. PRx has previously been demonstrated to correlate strongly with dynamic autoregulation of CBF assessed using transcranial Doppler ultrasonography,11 and with static autoregulation assessed using positron emission tomography (PET)– CBF.12 To standardize between-patients’ analysis, six consecutive phases of hypothermia were defined (Fig. 2). All continuously monitored variables were averaged in every patient for each of the following phases: (1) baseline (B): before cooling (2) cooling phase (C): from the beginning of the cooling process until a brain temperature of 34.58C was achieved (3) initial hypothermia (IH): the first 3 h of hypothermia (4) hypothermia (H): the whole period of hypothermia, brain temperature always ,358C (5) rewarming (R): brain temperature increasing from 35 to 378C (6) post-rewarming phase (P): temperature .378C Univariate repeated measures ANOVA with Duncan post hoc test was used for initial analysis of differences in average Prx, CPP, ICP, PtO2, PtCO2, pH, and temperature between different phases of hypothermia. Data distribution between

groups was not significantly different, meeting assumptions for ANOVA (Levene’s test). Multivariate analysis of covariance (GLM2) was then used to evaluate the effect of covariates (continuous, normally distributed parameters) PtO2, PtCO2, and pH on relationship between PRx levels and temperature groups. Multiple linear regression analysis was used to identify correlations between PRx and covariates in the post-rewarming phase. To avoid the confounding effects of pooled data on regression relationships, we also investigated the correlation between PRx and brain temperature within all individual patients. We expressed discrete variables as counts ( percentage) or median [interquartile range (IQR)], and continuous variables as mean (SD). All tests were two-tailed, and P,0.05 was considered statistically significant.

Results Twenty-four patients required moderate hypothermia as part of their management. Patients characteristics, Marshall CT grade,13 injury severity score,14 surgical procedures, day of hypothermia initiation, and duration of hypothermia are given in Table 1. Average CPP was managed between 70 and 80 mm Hg throughout the cooling and rewarming process (Fig. 3). The baseline brain temperature was 36.0 (0.9)8C, with ICP

239

Downloaded from http://bja.oxfordjournals.org/ at Seton Hall on April 6, 2015

Fig 2 Trends in brain temperature, ICP, CPP, and PRx in a 56-yr-old head-injured male monitored for 88 h. Phases in brain temperature include: [B] baseline (from 0 to 4.5 h), ICP .20 mm Hg, and increasing despite maximal medical treatment, PRx around 0.2; [C] cooling phase (from 4.5 to 9 h), brain temperature decreasing from 37 to 34.58C during active cooling; [IH] initial hypothermia (from 9 to 12 h), brain temperature 348C, ICP ,15 mm Hg and PRx ,0.2; [H] hypothermia (from 9 to 46 h), temperature below 358C, ICP still decreasing and PRx stable at 0.2; [R] rewarming (from 46 to 77 h), brain temperature increasing from 35 to 378C, ICP and PRx being normal; [P] post-rewarming (from 77 to 88 h), ICP being stable and CPP .80 mm Hg, but note the dramatic PRx impairment as brain temperature increases from 37 to 388C.

Lavinio et al.

Table 1 Patients characteristics, Marshall CT grade, injury severity score, surgical procedures, day of hypothermia initiation, and duration of hypothermia. *Patients who were not monitored in the post-rewarming phase. ISS, injury severity score. GCS, Glasgow coma scale at admission. CT grade, Marshall grade at admission (1, diffuse injury with no visible pathology; 2, diffuse injury with midline shift ,5 mm, small (,25 ml) focal lesion or both; 3, diffuse injury with brain swelling and compressed cisterns; 4, diffuse injury with compressed cisterns and midline shift .5 mm; 5, evacuated mass lesion; 6, non-evacuated mass lesion). SDH, subdural haematoma; EDH, epidural haematoma; cooled on day, day of initiation of hypothermia from admission; hypothermia, duration of hypothermia (brain temperature ,358C); GOS, Glasgow outcome scale. N/A, not available Patient no.

Sex

Age

ISS

GCS

Surgical procedure

1 2 3 4

M F M M

18 39 49 56

45 41 16 16

5

M

46

6

M

7 8

5 3 6 6

4 2 6 6

20

10

5

17

50

3

5

M M

22 25

34 16

3 4

2 3

9

M

41

20

14

6

10

M

27

38

3

6

11 12 13

M F M

32 30 56

25 38 25

3 3 6

3 3 6

14

M

49

21

8

5

15

M

37

9

11

6

16

M

66

25

3

5

17 18*

M M

43 50

34 27

5 6

2 5

19* 20*

F M

64 17

9 16

11 7

6 6

21*

M

21

16

7

5

22* 23* 24*

M F M

38 23 39

34 50 25

3 7 7

3 2 3

None None None Craniectomy, evacuation of SDH Craniectomy, evacuation of SDH Decompressive craniectomy None Decompressive craniectomy Decompressive craniectomy Decompressive craniectomy None None Craniotomy, evacuation of SDH Craniectomy, evacuation of EDH, lobectomy Craniectomy, evacuation of contusion Craniectomy, evacuation of SDH None Craniectomy, evacuation of SDH None Decompressive craniectomy Craniectomy, evacuation of contusion None None None

at 23.1 (3.6) mm Hg despite maximal medical treatment. Before cooling, mean PRx was 0.00 (0.21), demonstrating normal or only mildly disturbed autoregulation. Despite refractory intracranial hypertension, baseline cerebrovascular reactivity was impaired (PRx .0.2) only in 3 out of 24 patients. The duration of monitoring preceding the initiation of the cooling process was 8.2 (9.1) h. Patients were effectively cooled to the targeted brain temperature of 34.58C in 3.9 (3.7) h. Initial hypothermia, defined as the first 3 h of hypothermia, was characterized by a reduction in brain temperature to 34.2 (0.5)8C, and a

Cooled on day

Hypothermia (h)

ICU stay (days)

Mortality

GOS

2 2 2 5

65 29 14 34

18 16 2 14

Alive Alive Dead Dead

5 5 1 1

2

19

5

Alive

4

3

52

32

Alive

N/A

2 3

29 201

19 26

Dead Dead

1 1

10

124

48

Alive

4

3

25

28

Alive

5

4 9 2

7 26 25

19 32 17

Alive Alive Dead

4 5 1

5

82

36

Alive

N/A

2

14

22

Alive

N/A

2

30

9

Dead

1

5 3

10 19

29 16

Alive Alive

4 2

5 2

17 56

29 6

Alive Dead

3 1

2

10

12

Alive

4

2 2 5

60 13 79

6 19 20

Dead Alive Dead

1 4 1

prompt significant reduction in ICP to 19.3 (7.0) mm Hg (P,0.05). Considering the whole hypothermia phase [H—average duration 40 (45) h; IQR (14.9 – 52.7) h], the temperature was 34.3 (0.4)8C, and ICP was stable at 18.3 (4.8) mm Hg. Throughout the exposure to hypothermia, cerebrovascular autoregulation remained within normal values [PRx¼20.01 (0.21)]. Patients were then slowly rewarmed; brain temperature gradually increased from 35 to 378C for an average duration of 23.8 (21.6) h IQR (7.4 – 39.2) h. The average rate of rewarming (in 8C h21) was 0.2 (0.2)8C h21. During the

240

Downloaded from http://bja.oxfordjournals.org/ at Seton Hall on April 6, 2015

CT grade

TBI, hypothermia, and cerebral vasoreactivity

rewarming phase, average brain temperature was 35.9 (0.5)8C. ICP remained significantly lower than baseline at 18.6 (6.2) mm Hg (P,0.01), without any significant rebounds. During the rewarming phase, cerebrovascular autoregulation remained unaffected, as suggested by a non-significant change in mean PRx [0.06 (0.18); P¼0.40]. Average PRx was still within normal values (PRx,0.2) for brain temperatures up to 378C. While in 7 out of 24 patients we were not able to document brain temperatures increasing .378C, 17 patients were still monitored as brain temperature exceeded the 378C threshold. This period was previously defined as the post-rewarming phase [average monitoring time 32.4 (29.4) h; IQR (11.0– 52.2 h)]. During the post-rewarming phase, average brain temperature was allowed to increase to 37.8 (0.3)8C. ICP remained stable at 18.3 (8.0) mm Hg, but dramatic derangement in cerebrovascular reactivity was indicated by a significant increase in mean PRx to 0.32 (0.24) (P,0.0001). The degree of PRx worsening was not related to the rate of rewarming (P¼0.43). However, as the 378C threshold was exceeded, PRx became linearly related to the brain temperature (R¼0.53; n¼17, P,0.05; Fig. 4). The derangement in cerebral vasoreactivity observed during the post-rewarming period was investigated with multivariate analysis to assess the importance of the

monitored covariates (PtO2, PtCO2, pH, CPP, and ICP). All patients had invasive arterial blood gas monitoring, and the protocol implemented in our unit demanded normalization of PaCO2 before rewarming and maintenance of PaCO2,5.3 kPa during rewarming. Although high fluctuations in PaCO2 are unlikely, PtCO2 significantly increased from 5.5 (0.6) kPa (corresponding to a PaCO2 of approximately 4.8 kPa)15 during hypothermia to a PtCO2 of 6.8 (0.6) kPa (corresponding to a PaCO2 of 5.4 kPa)15 in the post-rewarming phase (P,0.001). However, average PtCO2

Fig 4 Scatter plot and regression line for PRx and average brain temperature after rewarming from moderate hypothermia (n¼17; P,0.05; Pearson’s R¼0.53).

241

Downloaded from http://bja.oxfordjournals.org/ at Seton Hall on April 6, 2015

Fig 3 Mean values and 95% CI for brain temperature, ICP, CPP, PRx during: [B] baseline, [C] cooling phase, [IH] initial hypothermia, [H] hypothermia, [R] rewarming, and [P] post-rewarming. CPP was stable throughout the cooling and rewarming process. ICP significantly decreased during IH and H, and there was no rewarming-related rebound. PRx significantly increased during P (mean PRx¼0.32, mean brain temperature¼37.88C). The asterisk indicates repeated measures analysis of variance, Duncan’s post hoc test (n¼17).

Lavinio et al.

Fig 5 Scatter plot and regression line for PRx and average brain O2 pressure after rewarming (n¼17; P,0.01; Pearson’s R¼0.66).

Fig 6 Pooled PRx values, grouped by brain temperature. In patients exposed to moderate hypothermia (n¼24), cerebrovascular reactivity worsened as temperature increased from 35 to 38.88C, becoming clearly impaired above the 378C threshold.

Discussion These findings describe the behaviour of cerebrovascular reactivity in head-injured patients exposed to moderate hypothermia for refractory intracranial hypertension. Fast induction of hypothermia led to a prompt reduction in average ICP, without compromising cerebrovascular reactivity. However, when the patients were rewarmed to brain temperatures exceeding 378C, a significant temperature-dependent impairment in cerebral vasoreactivity, indicating derangement in autoregulation of cerebral blood flow, occurred. These results should therefore increase our awareness regarding the importance of scrupulous management of systemic temperature after rewarming from moderate hypothermia. The significantly positive correlation between PRx and PtO2 in patients exposed to moderate hypothermia suggests that the cerebral vasoreactivity derangement that follows rewarming is of hyperaemic nature.16 Eleven out of 12 patients showing impaired vasoreactivity after rewarming (PRx.0.2) presented with average PtO2 readings .2 kPa. It was already demonstrated that brain tissue CO2 increased as brain temperature increased in patients treated with moderate hypothermia, even when arterial normocapnia was guaranteed.15 This finding was confirmed in the present series, and we hypothesized that a PtCO2-related dilation of cerebral vessels might have induced a derangement of cerebral vasoreactivity that was of hyperaemic nature. However, the multivariate model suggests that the temperature-dependent worsening in PRx remain significant even after controlling PtCO2 throughout the cooling and rewarming process. We infer that PtCO2 is not the only causal factor explaining the observed derangement in cerebral vasoreactivity. Moreover, we did not observe any significant correlation between average PtCO2 and PRx in the post-rewarming phase. We therefore conclude that the increased temperature has an independent detrimental impact on the reactivity of cerebral vessels, and that

242

Downloaded from http://bja.oxfordjournals.org/ at Seton Hall on April 6, 2015

did not correlate with PRx after excessive rewarming. On the other hand, average PtCO2 significantly correlated with PRx in the post-rewarming phase (R¼0.66; n¼17, P,0.01; Fig. 5). The model prediction for increase in PRx after rewarming remained significant (P,0.001) after controlling the values of all covariates, demonstrating an independent impact of brain temperature on PRx in patients exposed to moderate hypothermia. To avoid the confounding effects of pooled data on regression relationships, we also investigated the relationship between PRx and brain temperature within all individual patients, considering both the rewarming and post-rewarming period. In patients exposed to hypothermia, PRx and brain temperature were positively and significantly correlated in 16 of 24 patients [median (IQR) R¼0.85 (0.80– 0.88)]. Pooled data for brain temperature and cerebral vasoreactivity after the commencement of rewarming are shown in Figure 6.

Average CRP (normal range, 0–6 mg litre21) significantly increased from 77 (59) mg litre21 to 114 (54) mg litre21 from baseline to the hypothermia phase (paired t-test; P,0.05), but then remained stable throughout rewarming [116 (71) mg litre21] and post-rewarming phases 107 (82) mg litre21. There was no statistically significant difference in CRP or WBC count during hypothermia or rewarming between the 17 patients that were subsequently monitored in the post rewarming phase and the 7 patients who never exceeded the 378C threshold. In the subgroup of patients monitored in the postrewarming phase, mortality was 6 (35%) out of 17 patients, not statistically different from the mortality of the whole study population. Within this sub-group, patients who survived had a significantly lower average post-rewarming PRx [0.23 (0.19)], which was still better vasoreactivity than the 6 patients who died [PRx¼0.49 (0.23); P,0.05].

TBI, hypothermia, and cerebral vasoreactivity

In conclusion, moderate hypothermia helped to control increased ICP, and did not impair cerebrovascular reactivity in head-injured patients. Severe impairment in autoregulation occurred when after the rewarming process brain temperature was allowed to drift above 378C. To prevent severe derangement of cerebral vasoreactivity, systemic temperature should be aggressively controlled after rewarming from moderate hypothermia.

Acknowledgements The authors are in debt to the whole team participating in the data collection: Mrs Pippa Al-Rawi, Mrs Dot Chatfield, Mr Peter Kirkpatrick, Dr Philippe Bijlenga, and all the nursing and research staff on the Neurosciences Critical Care Unit. This project was supported by the UK Government Technology Foresight Initiative, and the Medical Research Council (Grant No G9439390 ID 65883). J. Nortje was supported by a British Journal of Anaesthesia/Royal College of Anaesthetists Fellowship. P. J. Hutchinson is supported by an Academy of Medical Sciences Health Foundation, Senior Surgical Scientist Fellowship. M. Czosnyka is on unpaid leave from Warsaw University of Technology, Poland. I. Timofeev was supported by the grant from the Codman division, Johnson & Johnson Inc. for the duration of this study, and he is currently supported by the Evelyn Trust and BP-TNK Kapitza scholarship.

References 1 Ramani R. Hypothermiafor brain protection and resuscitation. Curr Opin Anaesthesiol 2006; 19: 487 – 91 2 McIntyre LA, Fergusson DA, Hebert PC, Moher D, Hutchison JS. Prolonged therapeutic hypothermia after traumatic brain injury in adults: a systematic review. JAMA 2003; 289: 2992 – 9 3 Henderson WR, Dhingra VK, Chittock DR, Fenwick JC, Ronco JJ. Hypothermia in the management of traumatic brain injury. A systematic review and meta-analysis. Intensive Care Med 2003; 29: 1637 – 44 4 Czosnyka M, Hutchinson PJ, Balestreri M, Hiler M, Smielewski P, Pickard JD. Monitoring and interpretation of intracranial pressure after head injury. Acta Neurochir 2006; 96: 114– 8 5 Czosnyka M, Balestreri M, Steiner L, et al. Age, intracranial pressure, autoregulation, and outcome after brain trauma. J Neurosurg 2005; 102: 450 – 4 6 Czosnyka M, Smielewski P, Piechnik S, Pickard JD. Clinical significance of cerebral autoregulation. Acta Neurochir 2002; 81: 117 – 9 7 Kincaid S, Rozet I, Benirschke S, Visco E, Lam A. Autoregulation and CO2 reactivity of cerebral blood flow is preserved under mild hypothermia during general anesthesia. Anesthesiology 2005; 103: A17 8 Suehiro E, Ueda Y, Wei EP, Kontos HA, Povlishock JT. Posttraumatic hypothermia followed by slow rewarming protects the cerebral microcirculation. J Neurotrauma 2003; 20: 381 – 90 9 Menon DK. Cerebral protection in severe brain injury: physiological determinants of outcome and their optimisation. Br Med Bull 1999; 55: 226 – 58 10 Smielewski P, Czosnyka M, Steiner L, Belestri M, Piechnik S, Pickard JD. ICM þ : software for on-line analysis of bedside monitoring data after severe head trauma. Acta Neurochir Suppl 2005; 95: 43 – 9 11 Czosnyka M, Smielewski P, Kirkpatrick P, Laing RJ, Menon D, Pickard JD. Continuous assessment of the cerebral vasomotor reactivity in head injury. Neurosurgery 1997; 41: 11 – 7

243

Downloaded from http://bja.oxfordjournals.org/ at Seton Hall on April 6, 2015

temperature should be aggressively controlled in order to preserve cerebrovascular reactivity after rewarming from moderate hypothermia. A previous murine study proved that post-traumatic hypothermia followed by rapid rewarming induced impairment in pial vascular reactivity, and that slow rewarming maintained normal vasoreactivity.8 In the present series, all patients were slowly rewarmed, and the degree of PRx worsening was not related to the rate of rewarming. We are therefore unable to draw any conclusions about the impact of variable speed of rewarming on cerebral vasoreactivity. However, the novel finding is that when slow rewarming exceeded 378C after exposure to moderate hypothermia, PRx increased with temperature towards harmful values in the majority of the patients. Although an average PRx changing from 0.06 to 0.32 might seem to represent a rather unimpressive increase, we underline that prognosis and PRx are not linearly related, as mortality rate was demonstrated to increase steeply from 20 to 70% as average PRx increases above the 0.3 threshold.4 In patients exposed to brain temperatures ,34.58C for refractory intracranial hypertension, cerebral haemodynamics seem to become very sensitive to increments in brain temperature exceeding the 378C threshold. This phenomenon is usually undetected by plain ICP monitoring, which lacks explicit information about cerebral vasoreactivity. In this regard, a steady ICP after rewarming from moderate hypothermia might be falsely reassuring, and induce suboptimal control of systemic temperature. In the present series, brain temperature exceeded the 378C threshold in 70% of patients, and this was associated with a significant derangement of cerebral autoregulation in the majority of the cases. During hypothermia and rewarming 96 and 88% of 24 patients, respectively, had an average CRP above the 50 mg litre21 threshold. These findings are highly suggestive for sepsis, and might provide a causal interpretation for an excessive increase in brain temperature after rewarming in the majority of our patients.17 However, both CRP and WBC were not significantly correlated to post-rewarming PRx when they were added to brain temperature and PtCO2 in multiple regression analysis. This is a retrospective study, thus we cannot conclude whether the excessive increase in brain temperature after rewarming is a causal factor of cerebral vasoreactivity derangement. Those findings might be associated as proportional epiphenomena of sepsis, causing both pyrexia and the derangement of cerebral vasoreactivity. However, until a prospective study addressing those issues becomes available, we suggest that after rewarming from moderate hypothermia, an aggressive temperature management is achievable and should be guaranteed, as this might protect cerebral vasoreactivity in head-injured patients exposed to moderate hypothermia for refractory intracranial hypertension.

Lavinio et al.

12 Steiner LA, Coles JP, Johnston AJ, et al. Assessment of cerebrovascular autoregulation in head-injured patients: a validation study. Stroke 2003; 34: 2404– 9 13 Marshall LF, Marshall SB, Klauber MR, Van Berkum Clark M. A new classification of head injury based on computerized tomography. J Neurosurg 1991; 75: S14 –S20 14 Baker SP, O’Neill B, Haddon W jr, Long WB. The injury severity score: a method for describing patients with multiple injuries and evaluating emergency care. J Trauma 1974; 14: 187 – 96

15 Gupta AK, Al-Rawi PG, Hutchinson PJ, Kirkpatrick PJ. Effect of hypothermia on brain tissue oxygenation in patients with severe head injury. Br J Anaesth 2002; 88: 188– 92 16 Jaeger M, Schuhmann MU, Soehle M, Meixensberger J. Continuous assessment of cerebrovascular autoregulation after traumatic brain injury using brain tissue oxygen pressure reactivity. Crit Care Med 2006; 34: 1783– 8 17 Povoa P. C-reactive protein: a valuable marker of sepsis. Intensive Care Med 2002; 28: 235 – 43

Downloaded from http://bja.oxfordjournals.org/ at Seton Hall on April 6, 2015

244

Detroit Metal City BD Subtitle Indonesia | Paul Wesley | Saoirse Ronan