- Email: [email protected]
Cerebral blood flow is proportional to cardiac index in patients with septic shock
Cerebral Blood Flow Is Proportional to Cardiac Index in Patients With Septic Shock Stephen
M. Smith,
Soundrie
Padayachee,
Kamran
B. Modaresi,
Mark N. Smithies,
and David J. Bihari
Purpose: In patients with septic shock, the cardiac indexoften increased. Maldistribution of blood flow and regional hypoperfusion has been implicated as a key factor in the pathogenesis of organ dysfunction in these patients. We have investigated the relationship between cerebral blood flow and cardiac index in patients with septic shock. Materials and Methods: We used Doppler ultrasound techniques to investigate limb and carotid blood flow in 15 patients with septic shock and 9 nonseptic controls. Results: In the nonseptic control patients, common femoral and brachial blood flow were proportional to cardiac index (r = 0.73 and 0.76; P = .036 and ,017,
respectively) reflecting a protective redistribution of flow to more vital organs. However, this relationship was absent in patients with septic shock (r = 0.23 and 0.21). Furthermore, in the septic patients but not the nonseptic controls, cerebral blood flow was correlated with the cardiac index (r = 0.66, P < .05 vs r = -0.36, NS in nonseptic controls). Carotid flow was independent of mean arterial pressure, Pacoz and Paoz in patients with septic shock. Conclusions: These data are consistent with a loss of autoregulation of cerebral blood flow and a change in the control of limb blood flow in humans with septic shock. Copyright o 1998 by W.B. Saunders Company
S
vasoparesis than normotensive septic patients so that any sepsis-induced loss of autoregulation of cerebral blood flow would be more pronounced in this group. We have determined the relationship between cardiac index and regional blood flow in three vascular beds using a noninvasive Doppler ultrasound technique. We demonstrate that the relationship between cardiac index and regional blood flow is different in patients with septic shock compared with other types of shock and specifically that carotid flow is proportional to cardiac index. This and the absence of a relationship between carotid flow and mean arterial pressure (MAP), Paco2 or Pao, is consistent with a loss of autoregulation of cerebral blood flow in patients with septic shock.
EPTIC SHOCK is a common medical condition with a poor prognosis.’ It remains a frequent cause of multiple organ failure (MOF) and death in patients in intensive care units.2s3Although the cardiac output is often increased, tissue hypoperfusion together with the resultant deficiency of oxygen and other nutrients has been proposed as an important contributing factor in the pathogenesis of MOE4x5 Cerebral dysfunction manifesting as an encephalopathy is observed in up to 70% of patients with the sepsis syndrome,6x7and this group of patients has a higher mortality rate than septic patients with normal cerebration. Patients with the sepsis syndrome and encephalopathy have been reported to have reduced cerebral blood flo~.~ Interestingly, similar patients have recently been shown to have intact autoregulation of their cerebral blood flow9 suggesting that sepsis-induced vasoparesis does not affect the cerebral vasculature. Such sparing of the cerebral vasculature is an intriguing finding given the profound effects of sepsis on other vascular beds,lO and it has important implications for our understanding of the pathogenesis of septic encephalopathy. Patients with septic shock have more severe
From the Departments of Intensive Care and Ultrasonic An&logy, Guy’s Hospital, St. Thomas St, London, England. Received December 20, 1997. Accepted March 12,1998. Address reprint requests to Stephen M. Smith, FRACP, PhD, Department of Molecular and Cellular Physiology, Beckman CenteK Stanford University Medical Centel; Stanford, CA 943055426. Copyright 0 1998 by WB. Saunders Company 0883-9441/98/1303-0002$8.00/0 104
MATERIALS
AND
METHODS
Subjects In this study, which was approved by Guy’s and Lewisham Hospital ethics committee, we investigated regional blood flow in 15 patients with septic shock and 9 control patients receiving treatment for hypovolemic or cardiogenic shock. All required invasive hemodynamic monitoring, and informed consent was obtained from the subject or their next of kin. Standard criteria for the diagnosis of septic shock were used” and the sources of infection were pneumonia (n = 7), in&a-abdominal sepsis (n = 7); and meningococcal septicemia (n = 1). In the control group, patients were studied following cardiac surgery (n = 4), acute myocardial infarction (n = 3). cystectomy (n = l), and transurethral resection of the prostate (n = 1).
Measurements The patients were resuscitated and studied in the supine position between 12 and 48 hours after the diagnosis of shock was made and once cardiac index and blood pressure were Journal
ofCriritica/
Care, Vol 13, No 3 (September).
1998: PP 104-109
CEREBRAL
BLOOD
FLOW
IN SEPTIC
105
SHOCK
stable. Cardiac output was measured in triplicate by thermodilution. At this time, the MAP was noted and arterial blood drawn for measurement of arterial CO2 (Pacoz), 02 (Paoz) and lactate. Blood lactate was measured by a photometric method (Yellowsprings, CA). A color flow ultrasound scanner (Acuson 128x, Mountainview, CA) was used to obtain measurements of vessel diameter (d) by B-mode ultrasonography and the time averaged blood velocity (v) by Doppler scanning in the common carotid, brachial, and common femoral arteries. The flow (Q) in each vessel was calculated by multiplying the vessel cross-sectional area by the mean blood velocity (Q = ~v[d/2]~).‘~ The noninvasive nature of Doppler ultrasonography has resulted in its increasing use to measure blood flow in clinical investigations.13 But what of its accuracy? B-mode ultrasonography is used to measure changes as small as 3% to 10% in the diameters of brachial and femoral arteries of children and adults with a number of different conditions.‘J-‘6 Measurement of arterial flow by Doppler ultrasonography has a reproducibility of 18% in the brachial artery” where, because of its small size, errors will be more pronounced than in the femoral and carotid tieries.12 In the femoral artery and aorta blood flow estimated from Doppler ultrasound measurements has correlated well with flow measured by direct collection or flow meter (r = 0.90 to 0.96).‘8-2” The carotid artery is similar to the femoral artery in terms of its diameter and distance from the skin and so is amenable to the same type of measurement. An ~-MHZ linear array transducer was used for all measurements. Depth and gain settings were set to optimize the image of the artery. The vessel diameter was measured after checking the vessel had a circular cross-sectional area. Arterial blood velocity decreases from the center to the vessel wall, therefore, to obtain a true mean blood velocity the following measures were necessary. First, the Doppler sample volume gates were positioned on the B-mode image to include the entire internal vessel diameter and thus capture waves reflected from the whole of the internal cross-section of the artery. Second, a time-averaged velocity, rather than the peak velocity, was recorded.‘3,20 Third, the velocity measurements were made over 4.5 to 5 seconds to include up to 12 cardiac cycles. Fourth, the measurements were obtained consecutively and independently by two operators separated by 10 to 15 minutes and the mean values computed. The transducer was positioned to minimize the angle between the direction of insonation and direction of blood flow during blood velocity measurements. Arterial flow and cardiac output were divided by height and weight-derived body surface area and are presented as indexed values. To minimize artifact, measurements were not made from arteries containing cannulae, nor whenever possible, adjacent to cannulated veins. Carotid and femoral flows were not measured in 5 and 6 of the 24 patients, respectively. These included the first three patients entered into the study in who only brachial measurements were attempted. In three patients the arteries of interest were obscured by venous and arterial cannulae, massive tissue edema, or morbid obesity. Peripheral vascular disease caused turbulent flow in two other patients preventing meaningful Doppler measurements of blood velocity. Statistical analysis was performed using Student’s t test, Mann Whitney U test, Fishers exact test, and the Spearman rank correlation coefficient test as appropriate. Results are given as mean -f standard deviation.
RESULTS The clinical features of the patients and their vasoactive drug requirements are described in Table 1. Higher APACHE11 scores and adrenaline requirements were seen in patients with septic shock than in the group receiving treatment for nonseptic shock. Cardiac index was greater in the group with septic shock than in the control group (5.0 5 1.5 L/min/m* v 3.1 2 0.7 L/min/m2, P < .Ol). Plasma lactate levels were also greater in patients with septic shock (2.8 2 1.6 v 1.0 2 0.4 mmol/L, P < ,001). Indexed femoral, brachial, and carotid artery flows tended to be greater in patients with septic shock, but these differences did not reach statistical significance (Table 2). As a percentage of the cardiac index, the mean femoral, brachial, and carotid artery flows were the same for septic versus nonseptic patients (Table 2). The relationship between cardiac index and indexed regional flow is shown in Figure 1. In patients with septic shock, indexed femoral (Fig IA; r = 0.23) and brachial flows (Fig 1B; r = 0.21) did not correlate with cardiac index. However, there was a significant association between cardiac index and blood flow in the common carotid artery in these same patients (Fig 1C; r = 0.66, P < .05). In contrast, in the nonseptic group of patients, both indexed femoral (Fig 1D; r = 0.73, P < ,OS) and brachial flows (Fig 1E; r = 0.76, P < .02) were associated with cardiac index. However, there was no association between cardiac index and indexed carotid flow (Fig 1F; r = -0.37). The correlation between indexed carotid flow and cardiac index may imply a loss of autoregulaTable
1. Clinical
Features
of the 24 Patients Nonseptic Group (n = 9)
Age (years)
67 2 8
Range Gender (M/F)
55-81
712
APACHE II score Range No. of patients receiving inotropic
Values are expressed *p< .05. ,001.
Septic Shock (n = 15) 54 i 23 19-82 817 21.9 + 5.5* 9-32
support
(dose. pg/kg/min) Epinephrine Norepinephrine Dobutamine
tP<
16.7 2 6.6 5-27
Studied
2 (0.28 i 0.16) 0 3 (14.4 + 5.8) as mean
? SD.
14 (0.52 + 0.42)t 1 (0.681 9 (11.7 t_ 5.0)
106
SMITH
Table 2. Cardiac
Index
and Regional
Cardiac
index
Arterial Indexed
lactate (mmol/L) femoral flow (mL/min/m*)
(% of cardiac
(L/min/mz)
index)
Indexed brachial flow (% of cardiac index) Indexed
carotid
(% of cardiac
flow index)
(mL/min/mz)
(mL/min/mz)
Blood
Septic Shock
3.1 2 0.7 1.0 2 0.4 169 ? 89
5.0 t 1.5t 2.8 f 1.6$
(7.0 2 3.2)
243 ? 140 (5.1 5 1.4)
n=8 26 2 15
n = 10 49 t 24
(0.8 2 0.4) n=9 204 2 72 (5.4 -c 2.1) n=9
Hospital
mortality
“PC tP<
.05. .Ol.
s/J<
.OOl.
Flow
Nonseptic Group
3/9 (33%)
(1.0 20.6) n = 15 257 2 106 (4.6 2 2.6) n = IO 12/15 (80%)”
tion but may also reflect systematic variation in other important determinants of cerebral blood flow such as MAP, Pace, or Pao,. We examined the effects of these three variables on carotid flow in patients with septic shock. There is no significant correlation between MAP, Paco2, or Paoz, and indexed carotid flow (data not shown). The correlation coefficients for these data are 0.32, 0.067, and 0.28, respectively. DISCUSSION In this preliminary investigation of the relationship between regional blood flow and cardiac index in patients with septic shock our main finding is that cerebral blood flow, as measured by indexed carotid flow, is proportional to the cardiac index in our patients with septic shock. This is in contrast to normal, in which the cerebral blood flow is autoregulated over a range of systemic blood pressures and cardiac indices.21-23 More surprisingly, it is also different from nonshocked patients with the sepsis syndrome in whom cerebral autoregulation remains intact.g This may simply reflect a difference in disease severity, for the presence of shock denotes a more severe form of vasoparesis that may underlie the inability to autoregulate. The much higher mortality rate of our patients with septic shock compared with the patients reported by Matta and Stow (80% v 0%) is also consistent with there being a difference in disease severity between the two studies. The association between cardiac index and cerebral blood flow has at least two possible explanations. First, the parallel increases in cerebral blood
ET AL
flow and cardiac output may be in response to parallel increases in cerebral and systemic oxygen consumption. We did not measure cerebral oxygen consumption, but other studies have demonstrated a reduction in the cerebral oxygen consumption of patients8 and of animals24 with severe sepsis. It would seem unlikely then that an increase in cerebral oxygen consumption underlies the increase in cardiac index observed in the patients with septic shock. The second more likely explanation is that sepsis-induced vasoparesis results in a loss of autoregulation so that carotid blood flow is not matched to cerebral oxygen consumption but rather dependent on the cardiac index. Additional support for this idea comes from animal studies in which a loss of cerebral autoregulation was observed following severe hypercapnia-induced vasodilation in dogs.22 Furthermore, a loss of autoregulation has been reported in patients with fulminant hepatic failure in which cardiovascular instability and low systemic vascular resistance are also part of the clinical picture.25 We observed no association between MAP, PacoZ, or Pao, and mean carotid blood flow arguing against the possibility that these modulators of cerebral blood flow were confounding variables. Moreover the lack of an effect of MAP, Paco2 or Pao2 on indexed carotid flow is additional evidence that there is a disturbance of the regulation of cerebral blood flow in patients with septic shock. The size and design of this study does not allow us to determine whether this information is helpful in the management of patients with septic shock. The pathophysiological changes underlying a loss of cerebral autoregulation in patients with septic shock are not understood. The blood supply to many organs is disturbed by sepsis,26and this has been explained in part by sepsis-induced reductions in arterial reactivity. 27 Such large vessel changes may result in an increase in the heterogeneity of blood flow to different microcirculatory beds so that supply does not match metabolic demand in places. Changes within the microcirculation itself also occur. Decreased perfusion pressure and blood cell deformability, increased endothelial adhesiveness, and damage all result in sludging of blood within the microcirculation.‘O This manifests as reduced capillary blood flow and tissue oxygenation from early in the course of the episode of sepsis.28,2pFurther studies are required to determine if it is these changes that result in a loss of
CEREBRAL
BLOOD
FLOW
IN SEPTIC
107
SHOCK
Septic Shock
Control
A
600
1
200-
0 1
I 3
I 5
3
1 5
E
B 120-j
120
1
80-
40-
01
0
3
7
5
1
C
F
soo-
Fig 1. Cardiac index (L/mitt/ m21 versus indexed regional blood flow (mL/min/m2) in patients with septic shock (8) and nonseptic control lo). Femoral (Al and brachial (Bt flows are not correlated with cardiac index unlike carotid flow (Cl in patients with septic shock. In the control group femoral (D) and brachial lE) flows are correlated with cardiac index unlike carotidflow (FI.
s s k P '2 2 u = d i I
a
5001
400-
400
300-
300
1
0
0
i
200-
200
loo-
100 1
3
5
7
Cardiac index
autoregulation of cerebral blood flow or whether the aberrations in autoregulation precede the microcirculatory changes associated with severe sepsis. It is tempting to hypothesize that the observed association between brachial and femoral blood flow and cardiac index in the nonseptic critically ill patients reflects a homeostatic mechanism, so that in the setting of a limited cardiac output, blood flow
OI 1
O@ OO OO
3
5
Cardiac index
to more vital organs is conserved by redistributing blood away from the limbs. This is in agreement with clinical observation and indirect assessments of limb blood flow, such as toe-ambient temperature gradient30 in intensive care patients. Furthermore, recent studies of healthy volunteers have shown that both forearm and splanchnic blood flow is decreased following reductions in cardiac output
108
SMITH
secondary to hypovolemia.‘3,31 We demonstrated that limb blood flow is not correlated with cardiac index in patients with septic shock, and this is consistent with the earlier findings of Ruokonen et a1.32The loss of the association between cardiac index and limb blood flow suggests that the proposed homeostatic mechanism is affected by septic shock. This is another potential mechanism by which vital organ blood flow can be compromised in patients with septic shock. The loss of autoregulation of blood flow to vital organs in humans with septic shock may reflect the disease process, per se, but it might also be related to the action of vasopressors on different vascular beds. Unfortunately, in our study it was not possible to use the nonseptic shock group as a control
ET AL
to assessthe effects of the vasopressors because the latter group was receiving significantly lower doses. Further studies are required to examine the effect of introducing a vasopressor32 or changing from one to another33 to answer this question. Additional studies based on a larger number of patients would also be helpful to confirm the observations made here. Nevertheless, our data are consistent with the hypothesis that autoregulation of cerebral blood flow is abnormal in resuscitated patients with septic shock. ACKNOWLEDGMENT We thank Drs. Chris Harrington, Torzillo for their helpful comments manuscript.
Felicty Hawker, and Paul on earlier versions of the
REFERENCES 1. Parillo JE: Pathogenetic mechanisms of septic shock. N Engl J Med 328:1471-1477, 1993 2. Marshall JC, Cook DJ, Christou NC, et al: Multiple organ dysfunction score: A reliable descriptor of a complex clinical outcome. Crit Care Med 23:1638-1652,1995 3. Tran DD, Groenveld AB, van der Meulen J, et al: Age, chronic disease, sepsis and organ system failure, and mortality in a medical intensive care unit. Crit Care Med 18:474-479, 1990 4. Bihari D, Smithies M, Gimson, et al: The effects of vasodilation with prostacyclin on oxygen delivery and uptake in critically ill patients. N Engl J Med 317:397-403, 1987 5. Gutierrez G, Palizas F, Doligo G, et al: Gastric intramucosal pH as a therapeutic index of tissue oxygenation in critically ill patients. Lancet 339:195-199, 1992 6. Bolton CF, Young GB, Zochodne DW: The neurologic complications of sepsis. Ann Neural 33:94-100, 1993 7. Strung CL, Peduzzi PN, Satney CH, et al: for the Veterans Administration Systemic Sepsis Group: Impact of encephalopathy on mortality in sepsis. Crit Care Med 18:801-806, 1990 8. Bowton DL, Bertels NH, Prough DS, et al: Cerebral blood flow is reduced in patients with sepsis syndrome. Crit Care Med 17:399-403, 1989 9. Matta BF, Stow PJ: Sepsis-induced vasoparalysis does not involve the cerebral vasculature: Indirect evidence from autoregulation and carbon dioxide reactivity studies. Br J Anaesth 76:790-794, 1996 10. Hinshaw LB: Sepsis/septic shock: Participation of the microcirculation: An abbreviated review. Crit Care Med 24: 10721078,1996 11. Bone RC, Balk RA, Cerra FB, et al: Definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis. Chest 101:1644-1655, 1992 12. Gill RW: Pulsed Doppler with B-mode imaging for quantitative blood flow measurement. Ultrasound Med Biol 5:223-235, 1979 13. Dalton JM, Gore DC, Makhoul RG, et al: Decreased splanchnic perfusion measured by duplex ultrasound in humans undergoing small volume haemorrhage. Crit Care Med 23:491497,1995
14. Celermajer DS, Adams MR, Clarkson P, et al: Passive smoking and impaired endothelium-dependent arterial dilatation in healthy young adults. N Engl J Med 334:150-154, 1996 15. Celermajer DS, Sorensen KE, Gooch, et al: Non-invasive detection of endothelial dysfunction in children and adults at risk of atherosclerosis. Lancet 340:1111-1115, 1992 16. Sorensen KE, Celermajer DS, Georgakopoulos D, et al: Impairment of endothelium-dependent dilation is an early event in children with familial hypercholesterolemia and is related to the lipoprotein(a) level. J Clin Invest 93:50-55, 1994 17. Laurent S, Lacolley P, Brunel P, et al: Flow-dependent vasodilation of bra&al artery in essential hypertension. Am J Physiol258:H1004-HlOll, 1990 18. Chaveau M, Levy B, Dessanges JF, et al: Quantitative Doppler blood flow measurement method and in vivo calibration. Cardiovasc Res 19:700-706, 1985 19. Eik-Nes SH, Marsal K, Kristoffersen K: Methodology and basic problems related to blood flow studies in the human fetus. Ultrasound Med Biol 10:329-337, 1984 20. Zierler BK, Kirkman TR, Kraiss LW, et al: Accuracy of duplex scanning for measurement of arterial volume flow. J Vast Surg 16:520-526, 1992 21. Bouma GJ, Muizelaar JP: Relationship between cardiac output and cerebral blood flow in patients with intact and with impaired autoregulation. J Neurosurg 73:368-374, 1990 22. Harper AM: Autoregulation of cerebral blood flow: Influence of the arterial blood pressure on the blood flow through the cerebral cortex. J Neurol Neurosurg Psychiat 29:398-403, 1996 23. Lassen NA: Cerebral blood flow and oxygen consumption in man. Phys Rev 39:183-238, 1996 24. Tureen J: Effect of recombinant human tumor necrosis factor-alpha on cerebral oxygen uptake, cerebrospinal fluid lactate and cerebral blood flow in the rabbit: Role of nitric oxide. J Clin Invest 95:1086-1091, 1995 25. Larsen FS, Olsen KS, Ejlersen E, et al: Cerebral blood flow autoregulation and transcranial doppler sonography in patients with cirrhosis. Hepatology 22:730-736, 1995 26. Lang CH, Bagby GJ, Ferguson JL, et al: Cardiac output
CEREBRAL
BLOOD
FLOW
IN SEPTIC
SHOCK
and redistribution of organ blood flow in hypermetabolic sepsis. Am J Physiol246:R331-337, 1984 27. Martin CM, Yaghi A, Sibbald WJ, et al: Differential impairment of vascular reactivity of small pulmonary and systemic arteries in hyperdynamic sepsis. Am Rev Respir Dis 148:164-172, 1993 28. Piper RD, Pitt-Hyde M, Li F, et al: Microcirculatory changes in rat skeletal muscle in sepsis. Am J Respir Care Med 154:931-937, 1996 29. Sair M, Etherington PJ, Curzen NP, et al: Tissue oxygenation and perfusion in endotoxemia. Am J Physiol 271:H16201625, 1996
109
30. Vincent JL, Moraine JJ, van der Linden P: Toe temperature versus transcutaneous oxygen tension monitoring during acute circulatory failure. Intensive Care Med 14:64-68, 1988 3 1. Edouard AR, Degremont AC, Duranteau J, et al: Heterogeneous regional vascular responses to simulated transient hypovolemia in man. Intensive Care Med 20:414-420, 1994 32. Ruokonen E, Takala J, Kari A, et al: Regional blood flow and oxygen transport in septic shock. Crit Care Med 2 1: 1296. 1303,1993 33. Hanneman L, Reinhart K, Grenzer 0, et al: Comparison of dopamine and norepinephrine for oxygen delivery and uptake in septic shock. Crit Care Med 23:1962-1970, 1995
M. Smith,
Soundrie
Padayachee,
Kamran
B. Modaresi,
Mark N. Smithies,
and David J. Bihari
Purpose: In patients with septic shock, the cardiac indexoften increased. Maldistribution of blood flow and regional hypoperfusion has been implicated as a key factor in the pathogenesis of organ dysfunction in these patients. We have investigated the relationship between cerebral blood flow and cardiac index in patients with septic shock. Materials and Methods: We used Doppler ultrasound techniques to investigate limb and carotid blood flow in 15 patients with septic shock and 9 nonseptic controls. Results: In the nonseptic control patients, common femoral and brachial blood flow were proportional to cardiac index (r = 0.73 and 0.76; P = .036 and ,017,
respectively) reflecting a protective redistribution of flow to more vital organs. However, this relationship was absent in patients with septic shock (r = 0.23 and 0.21). Furthermore, in the septic patients but not the nonseptic controls, cerebral blood flow was correlated with the cardiac index (r = 0.66, P < .05 vs r = -0.36, NS in nonseptic controls). Carotid flow was independent of mean arterial pressure, Pacoz and Paoz in patients with septic shock. Conclusions: These data are consistent with a loss of autoregulation of cerebral blood flow and a change in the control of limb blood flow in humans with septic shock. Copyright o 1998 by W.B. Saunders Company
S
vasoparesis than normotensive septic patients so that any sepsis-induced loss of autoregulation of cerebral blood flow would be more pronounced in this group. We have determined the relationship between cardiac index and regional blood flow in three vascular beds using a noninvasive Doppler ultrasound technique. We demonstrate that the relationship between cardiac index and regional blood flow is different in patients with septic shock compared with other types of shock and specifically that carotid flow is proportional to cardiac index. This and the absence of a relationship between carotid flow and mean arterial pressure (MAP), Paco2 or Pao, is consistent with a loss of autoregulation of cerebral blood flow in patients with septic shock.
EPTIC SHOCK is a common medical condition with a poor prognosis.’ It remains a frequent cause of multiple organ failure (MOF) and death in patients in intensive care units.2s3Although the cardiac output is often increased, tissue hypoperfusion together with the resultant deficiency of oxygen and other nutrients has been proposed as an important contributing factor in the pathogenesis of MOE4x5 Cerebral dysfunction manifesting as an encephalopathy is observed in up to 70% of patients with the sepsis syndrome,6x7and this group of patients has a higher mortality rate than septic patients with normal cerebration. Patients with the sepsis syndrome and encephalopathy have been reported to have reduced cerebral blood flo~.~ Interestingly, similar patients have recently been shown to have intact autoregulation of their cerebral blood flow9 suggesting that sepsis-induced vasoparesis does not affect the cerebral vasculature. Such sparing of the cerebral vasculature is an intriguing finding given the profound effects of sepsis on other vascular beds,lO and it has important implications for our understanding of the pathogenesis of septic encephalopathy. Patients with septic shock have more severe
From the Departments of Intensive Care and Ultrasonic An&logy, Guy’s Hospital, St. Thomas St, London, England. Received December 20, 1997. Accepted March 12,1998. Address reprint requests to Stephen M. Smith, FRACP, PhD, Department of Molecular and Cellular Physiology, Beckman CenteK Stanford University Medical Centel; Stanford, CA 943055426. Copyright 0 1998 by WB. Saunders Company 0883-9441/98/1303-0002$8.00/0 104
MATERIALS
AND
METHODS
Subjects In this study, which was approved by Guy’s and Lewisham Hospital ethics committee, we investigated regional blood flow in 15 patients with septic shock and 9 control patients receiving treatment for hypovolemic or cardiogenic shock. All required invasive hemodynamic monitoring, and informed consent was obtained from the subject or their next of kin. Standard criteria for the diagnosis of septic shock were used” and the sources of infection were pneumonia (n = 7), in&a-abdominal sepsis (n = 7); and meningococcal septicemia (n = 1). In the control group, patients were studied following cardiac surgery (n = 4), acute myocardial infarction (n = 3). cystectomy (n = l), and transurethral resection of the prostate (n = 1).
Measurements The patients were resuscitated and studied in the supine position between 12 and 48 hours after the diagnosis of shock was made and once cardiac index and blood pressure were Journal
ofCriritica/
Care, Vol 13, No 3 (September).
1998: PP 104-109
CEREBRAL
BLOOD
FLOW
IN SEPTIC
105
SHOCK
stable. Cardiac output was measured in triplicate by thermodilution. At this time, the MAP was noted and arterial blood drawn for measurement of arterial CO2 (Pacoz), 02 (Paoz) and lactate. Blood lactate was measured by a photometric method (Yellowsprings, CA). A color flow ultrasound scanner (Acuson 128x, Mountainview, CA) was used to obtain measurements of vessel diameter (d) by B-mode ultrasonography and the time averaged blood velocity (v) by Doppler scanning in the common carotid, brachial, and common femoral arteries. The flow (Q) in each vessel was calculated by multiplying the vessel cross-sectional area by the mean blood velocity (Q = ~v[d/2]~).‘~ The noninvasive nature of Doppler ultrasonography has resulted in its increasing use to measure blood flow in clinical investigations.13 But what of its accuracy? B-mode ultrasonography is used to measure changes as small as 3% to 10% in the diameters of brachial and femoral arteries of children and adults with a number of different conditions.‘J-‘6 Measurement of arterial flow by Doppler ultrasonography has a reproducibility of 18% in the brachial artery” where, because of its small size, errors will be more pronounced than in the femoral and carotid tieries.12 In the femoral artery and aorta blood flow estimated from Doppler ultrasound measurements has correlated well with flow measured by direct collection or flow meter (r = 0.90 to 0.96).‘8-2” The carotid artery is similar to the femoral artery in terms of its diameter and distance from the skin and so is amenable to the same type of measurement. An ~-MHZ linear array transducer was used for all measurements. Depth and gain settings were set to optimize the image of the artery. The vessel diameter was measured after checking the vessel had a circular cross-sectional area. Arterial blood velocity decreases from the center to the vessel wall, therefore, to obtain a true mean blood velocity the following measures were necessary. First, the Doppler sample volume gates were positioned on the B-mode image to include the entire internal vessel diameter and thus capture waves reflected from the whole of the internal cross-section of the artery. Second, a time-averaged velocity, rather than the peak velocity, was recorded.‘3,20 Third, the velocity measurements were made over 4.5 to 5 seconds to include up to 12 cardiac cycles. Fourth, the measurements were obtained consecutively and independently by two operators separated by 10 to 15 minutes and the mean values computed. The transducer was positioned to minimize the angle between the direction of insonation and direction of blood flow during blood velocity measurements. Arterial flow and cardiac output were divided by height and weight-derived body surface area and are presented as indexed values. To minimize artifact, measurements were not made from arteries containing cannulae, nor whenever possible, adjacent to cannulated veins. Carotid and femoral flows were not measured in 5 and 6 of the 24 patients, respectively. These included the first three patients entered into the study in who only brachial measurements were attempted. In three patients the arteries of interest were obscured by venous and arterial cannulae, massive tissue edema, or morbid obesity. Peripheral vascular disease caused turbulent flow in two other patients preventing meaningful Doppler measurements of blood velocity. Statistical analysis was performed using Student’s t test, Mann Whitney U test, Fishers exact test, and the Spearman rank correlation coefficient test as appropriate. Results are given as mean -f standard deviation.
RESULTS The clinical features of the patients and their vasoactive drug requirements are described in Table 1. Higher APACHE11 scores and adrenaline requirements were seen in patients with septic shock than in the group receiving treatment for nonseptic shock. Cardiac index was greater in the group with septic shock than in the control group (5.0 5 1.5 L/min/m* v 3.1 2 0.7 L/min/m2, P < .Ol). Plasma lactate levels were also greater in patients with septic shock (2.8 2 1.6 v 1.0 2 0.4 mmol/L, P < ,001). Indexed femoral, brachial, and carotid artery flows tended to be greater in patients with septic shock, but these differences did not reach statistical significance (Table 2). As a percentage of the cardiac index, the mean femoral, brachial, and carotid artery flows were the same for septic versus nonseptic patients (Table 2). The relationship between cardiac index and indexed regional flow is shown in Figure 1. In patients with septic shock, indexed femoral (Fig IA; r = 0.23) and brachial flows (Fig 1B; r = 0.21) did not correlate with cardiac index. However, there was a significant association between cardiac index and blood flow in the common carotid artery in these same patients (Fig 1C; r = 0.66, P < .05). In contrast, in the nonseptic group of patients, both indexed femoral (Fig 1D; r = 0.73, P < ,OS) and brachial flows (Fig 1E; r = 0.76, P < .02) were associated with cardiac index. However, there was no association between cardiac index and indexed carotid flow (Fig 1F; r = -0.37). The correlation between indexed carotid flow and cardiac index may imply a loss of autoregulaTable
1. Clinical
Features
of the 24 Patients Nonseptic Group (n = 9)
Age (years)
67 2 8
Range Gender (M/F)
55-81
712
APACHE II score Range No. of patients receiving inotropic
Values are expressed *p< .05. ,001.
Septic Shock (n = 15) 54 i 23 19-82 817 21.9 + 5.5* 9-32
support
(dose. pg/kg/min) Epinephrine Norepinephrine Dobutamine
tP<
16.7 2 6.6 5-27
Studied
2 (0.28 i 0.16) 0 3 (14.4 + 5.8) as mean
? SD.
14 (0.52 + 0.42)t 1 (0.681 9 (11.7 t_ 5.0)
106
SMITH
Table 2. Cardiac
Index
and Regional
Cardiac
index
Arterial Indexed
lactate (mmol/L) femoral flow (mL/min/m*)
(% of cardiac
(L/min/mz)
index)
Indexed brachial flow (% of cardiac index) Indexed
carotid
(% of cardiac
flow index)
(mL/min/mz)
(mL/min/mz)
Blood
Septic Shock
3.1 2 0.7 1.0 2 0.4 169 ? 89
5.0 t 1.5t 2.8 f 1.6$
(7.0 2 3.2)
243 ? 140 (5.1 5 1.4)
n=8 26 2 15
n = 10 49 t 24
(0.8 2 0.4) n=9 204 2 72 (5.4 -c 2.1) n=9
Hospital
mortality
“PC tP<
.05. .Ol.
s/J<
.OOl.
Flow
Nonseptic Group
3/9 (33%)
(1.0 20.6) n = 15 257 2 106 (4.6 2 2.6) n = IO 12/15 (80%)”
tion but may also reflect systematic variation in other important determinants of cerebral blood flow such as MAP, Pace, or Pao,. We examined the effects of these three variables on carotid flow in patients with septic shock. There is no significant correlation between MAP, Paco2, or Paoz, and indexed carotid flow (data not shown). The correlation coefficients for these data are 0.32, 0.067, and 0.28, respectively. DISCUSSION In this preliminary investigation of the relationship between regional blood flow and cardiac index in patients with septic shock our main finding is that cerebral blood flow, as measured by indexed carotid flow, is proportional to the cardiac index in our patients with septic shock. This is in contrast to normal, in which the cerebral blood flow is autoregulated over a range of systemic blood pressures and cardiac indices.21-23 More surprisingly, it is also different from nonshocked patients with the sepsis syndrome in whom cerebral autoregulation remains intact.g This may simply reflect a difference in disease severity, for the presence of shock denotes a more severe form of vasoparesis that may underlie the inability to autoregulate. The much higher mortality rate of our patients with septic shock compared with the patients reported by Matta and Stow (80% v 0%) is also consistent with there being a difference in disease severity between the two studies. The association between cardiac index and cerebral blood flow has at least two possible explanations. First, the parallel increases in cerebral blood
ET AL
flow and cardiac output may be in response to parallel increases in cerebral and systemic oxygen consumption. We did not measure cerebral oxygen consumption, but other studies have demonstrated a reduction in the cerebral oxygen consumption of patients8 and of animals24 with severe sepsis. It would seem unlikely then that an increase in cerebral oxygen consumption underlies the increase in cardiac index observed in the patients with septic shock. The second more likely explanation is that sepsis-induced vasoparesis results in a loss of autoregulation so that carotid blood flow is not matched to cerebral oxygen consumption but rather dependent on the cardiac index. Additional support for this idea comes from animal studies in which a loss of cerebral autoregulation was observed following severe hypercapnia-induced vasodilation in dogs.22 Furthermore, a loss of autoregulation has been reported in patients with fulminant hepatic failure in which cardiovascular instability and low systemic vascular resistance are also part of the clinical picture.25 We observed no association between MAP, PacoZ, or Pao, and mean carotid blood flow arguing against the possibility that these modulators of cerebral blood flow were confounding variables. Moreover the lack of an effect of MAP, Paco2 or Pao2 on indexed carotid flow is additional evidence that there is a disturbance of the regulation of cerebral blood flow in patients with septic shock. The size and design of this study does not allow us to determine whether this information is helpful in the management of patients with septic shock. The pathophysiological changes underlying a loss of cerebral autoregulation in patients with septic shock are not understood. The blood supply to many organs is disturbed by sepsis,26and this has been explained in part by sepsis-induced reductions in arterial reactivity. 27 Such large vessel changes may result in an increase in the heterogeneity of blood flow to different microcirculatory beds so that supply does not match metabolic demand in places. Changes within the microcirculation itself also occur. Decreased perfusion pressure and blood cell deformability, increased endothelial adhesiveness, and damage all result in sludging of blood within the microcirculation.‘O This manifests as reduced capillary blood flow and tissue oxygenation from early in the course of the episode of sepsis.28,2pFurther studies are required to determine if it is these changes that result in a loss of
CEREBRAL
BLOOD
FLOW
IN SEPTIC
107
SHOCK
Septic Shock
Control
A
600
1
200-
0 1
I 3
I 5
3
1 5
E
B 120-j
120
1
80-
40-
01
0
3
7
5
1
C
F
soo-
Fig 1. Cardiac index (L/mitt/ m21 versus indexed regional blood flow (mL/min/m2) in patients with septic shock (8) and nonseptic control lo). Femoral (Al and brachial (Bt flows are not correlated with cardiac index unlike carotid flow (Cl in patients with septic shock. In the control group femoral (D) and brachial lE) flows are correlated with cardiac index unlike carotidflow (FI.
s s k P '2 2 u = d i I
a
5001
400-
400
300-
300
1
0
0
i
200-
200
loo-
100 1
3
5
7
Cardiac index
autoregulation of cerebral blood flow or whether the aberrations in autoregulation precede the microcirculatory changes associated with severe sepsis. It is tempting to hypothesize that the observed association between brachial and femoral blood flow and cardiac index in the nonseptic critically ill patients reflects a homeostatic mechanism, so that in the setting of a limited cardiac output, blood flow
OI 1
O@ OO OO
3
5
Cardiac index
to more vital organs is conserved by redistributing blood away from the limbs. This is in agreement with clinical observation and indirect assessments of limb blood flow, such as toe-ambient temperature gradient30 in intensive care patients. Furthermore, recent studies of healthy volunteers have shown that both forearm and splanchnic blood flow is decreased following reductions in cardiac output
108
SMITH
secondary to hypovolemia.‘3,31 We demonstrated that limb blood flow is not correlated with cardiac index in patients with septic shock, and this is consistent with the earlier findings of Ruokonen et a1.32The loss of the association between cardiac index and limb blood flow suggests that the proposed homeostatic mechanism is affected by septic shock. This is another potential mechanism by which vital organ blood flow can be compromised in patients with septic shock. The loss of autoregulation of blood flow to vital organs in humans with septic shock may reflect the disease process, per se, but it might also be related to the action of vasopressors on different vascular beds. Unfortunately, in our study it was not possible to use the nonseptic shock group as a control
ET AL
to assessthe effects of the vasopressors because the latter group was receiving significantly lower doses. Further studies are required to examine the effect of introducing a vasopressor32 or changing from one to another33 to answer this question. Additional studies based on a larger number of patients would also be helpful to confirm the observations made here. Nevertheless, our data are consistent with the hypothesis that autoregulation of cerebral blood flow is abnormal in resuscitated patients with septic shock. ACKNOWLEDGMENT We thank Drs. Chris Harrington, Torzillo for their helpful comments manuscript.
Felicty Hawker, and Paul on earlier versions of the
REFERENCES 1. Parillo JE: Pathogenetic mechanisms of septic shock. N Engl J Med 328:1471-1477, 1993 2. Marshall JC, Cook DJ, Christou NC, et al: Multiple organ dysfunction score: A reliable descriptor of a complex clinical outcome. Crit Care Med 23:1638-1652,1995 3. Tran DD, Groenveld AB, van der Meulen J, et al: Age, chronic disease, sepsis and organ system failure, and mortality in a medical intensive care unit. Crit Care Med 18:474-479, 1990 4. Bihari D, Smithies M, Gimson, et al: The effects of vasodilation with prostacyclin on oxygen delivery and uptake in critically ill patients. N Engl J Med 317:397-403, 1987 5. Gutierrez G, Palizas F, Doligo G, et al: Gastric intramucosal pH as a therapeutic index of tissue oxygenation in critically ill patients. Lancet 339:195-199, 1992 6. Bolton CF, Young GB, Zochodne DW: The neurologic complications of sepsis. Ann Neural 33:94-100, 1993 7. Strung CL, Peduzzi PN, Satney CH, et al: for the Veterans Administration Systemic Sepsis Group: Impact of encephalopathy on mortality in sepsis. Crit Care Med 18:801-806, 1990 8. Bowton DL, Bertels NH, Prough DS, et al: Cerebral blood flow is reduced in patients with sepsis syndrome. Crit Care Med 17:399-403, 1989 9. Matta BF, Stow PJ: Sepsis-induced vasoparalysis does not involve the cerebral vasculature: Indirect evidence from autoregulation and carbon dioxide reactivity studies. Br J Anaesth 76:790-794, 1996 10. Hinshaw LB: Sepsis/septic shock: Participation of the microcirculation: An abbreviated review. Crit Care Med 24: 10721078,1996 11. Bone RC, Balk RA, Cerra FB, et al: Definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis. Chest 101:1644-1655, 1992 12. Gill RW: Pulsed Doppler with B-mode imaging for quantitative blood flow measurement. Ultrasound Med Biol 5:223-235, 1979 13. Dalton JM, Gore DC, Makhoul RG, et al: Decreased splanchnic perfusion measured by duplex ultrasound in humans undergoing small volume haemorrhage. Crit Care Med 23:491497,1995
14. Celermajer DS, Adams MR, Clarkson P, et al: Passive smoking and impaired endothelium-dependent arterial dilatation in healthy young adults. N Engl J Med 334:150-154, 1996 15. Celermajer DS, Sorensen KE, Gooch, et al: Non-invasive detection of endothelial dysfunction in children and adults at risk of atherosclerosis. Lancet 340:1111-1115, 1992 16. Sorensen KE, Celermajer DS, Georgakopoulos D, et al: Impairment of endothelium-dependent dilation is an early event in children with familial hypercholesterolemia and is related to the lipoprotein(a) level. J Clin Invest 93:50-55, 1994 17. Laurent S, Lacolley P, Brunel P, et al: Flow-dependent vasodilation of bra&al artery in essential hypertension. Am J Physiol258:H1004-HlOll, 1990 18. Chaveau M, Levy B, Dessanges JF, et al: Quantitative Doppler blood flow measurement method and in vivo calibration. Cardiovasc Res 19:700-706, 1985 19. Eik-Nes SH, Marsal K, Kristoffersen K: Methodology and basic problems related to blood flow studies in the human fetus. Ultrasound Med Biol 10:329-337, 1984 20. Zierler BK, Kirkman TR, Kraiss LW, et al: Accuracy of duplex scanning for measurement of arterial volume flow. J Vast Surg 16:520-526, 1992 21. Bouma GJ, Muizelaar JP: Relationship between cardiac output and cerebral blood flow in patients with intact and with impaired autoregulation. J Neurosurg 73:368-374, 1990 22. Harper AM: Autoregulation of cerebral blood flow: Influence of the arterial blood pressure on the blood flow through the cerebral cortex. J Neurol Neurosurg Psychiat 29:398-403, 1996 23. Lassen NA: Cerebral blood flow and oxygen consumption in man. Phys Rev 39:183-238, 1996 24. Tureen J: Effect of recombinant human tumor necrosis factor-alpha on cerebral oxygen uptake, cerebrospinal fluid lactate and cerebral blood flow in the rabbit: Role of nitric oxide. J Clin Invest 95:1086-1091, 1995 25. Larsen FS, Olsen KS, Ejlersen E, et al: Cerebral blood flow autoregulation and transcranial doppler sonography in patients with cirrhosis. Hepatology 22:730-736, 1995 26. Lang CH, Bagby GJ, Ferguson JL, et al: Cardiac output
CEREBRAL
BLOOD
FLOW
IN SEPTIC
SHOCK
and redistribution of organ blood flow in hypermetabolic sepsis. Am J Physiol246:R331-337, 1984 27. Martin CM, Yaghi A, Sibbald WJ, et al: Differential impairment of vascular reactivity of small pulmonary and systemic arteries in hyperdynamic sepsis. Am Rev Respir Dis 148:164-172, 1993 28. Piper RD, Pitt-Hyde M, Li F, et al: Microcirculatory changes in rat skeletal muscle in sepsis. Am J Respir Care Med 154:931-937, 1996 29. Sair M, Etherington PJ, Curzen NP, et al: Tissue oxygenation and perfusion in endotoxemia. Am J Physiol 271:H16201625, 1996
109
30. Vincent JL, Moraine JJ, van der Linden P: Toe temperature versus transcutaneous oxygen tension monitoring during acute circulatory failure. Intensive Care Med 14:64-68, 1988 3 1. Edouard AR, Degremont AC, Duranteau J, et al: Heterogeneous regional vascular responses to simulated transient hypovolemia in man. Intensive Care Med 20:414-420, 1994 32. Ruokonen E, Takala J, Kari A, et al: Regional blood flow and oxygen transport in septic shock. Crit Care Med 2 1: 1296. 1303,1993 33. Hanneman L, Reinhart K, Grenzer 0, et al: Comparison of dopamine and norepinephrine for oxygen delivery and uptake in septic shock. Crit Care Med 23:1962-1970, 1995