Departments of *Emergency Medicine, tSurgery, and :f:Pediatrics, Henry Ford Hospital, Detroit, Michigan, U.S.A.. Summary: Animal studies have shown cerebral ...
Journal of Cerebral Blood Flow and Metabolism 11:479--484 © 1991 The International Society of Cerebral Blood Flow and Metabolism Published by Raven Press, Ltd., New York
Cerebral Lactate Uptake During Cardiopulmonary Resuscitation in Humans
*tEmanuel P. Rivers, *Norman A. Paradis, *Gerard B. Martin, :j:Mark E. Goetting, *Jack A. Rosenberg, *Howard A. Smithline, *Timothy J. Appleton, and *Richard M. Nowak Departments of *Emergency Medicine, tSurgery, and :f:Pediatrics, Henry Ford Hospital, Detroit, Michigan, U.S.A.
Summary: Animal studies have shown cerebral lactate uptake under conditions of anoxia and ischemia. Cerebral lactate uptake in humans during cardiopulmonary resus citation (CPR) has not been previously reported in the literature. Forty-five patients receiving CPR underwent simultaneous sampling through jugular venous bulb, right atrial, and central aortic catheterization. The mean net cerebral lactate uptake (central aortic minus jugular ve nous bulb) was 0.76 ± 1.86 and 0.80 ± 2.03 mM on initial measurement and 10 min later, respectively. Both mea surements were statistically significant (p 0.01) com pared to normal controls who have net cerebral output of lactate of 0.18 ± 0.1 mM Seventy-one percent of all patients had a cerebral uptake on initial sampling and this
gradient persisted upon sampling 10 min later in 68% of the remaining 40 patients who did not have a return of spontaneous circulation. Among mUltiple variables mea sured, patients who exhibited a cerebral lactate uptake were 13.2 years younger (p 0.004), received an addi tional 7.6 min of CPR (p 0.05), and had a mean arterial lactate concentration of 4.8 mM higher (p 0.005) than the nonuptake group. The pathophysiologic explanation of cerebral lactate uptake during CPR is multifactorial and includes utilization and/or diffusion. Key Words: Ce rebral lactate gradients-Cerebral lactate uptake Cerebral lactate utilization-CPR-Global ischemia Lactate-Lactic acid-Humans.
The brain is one of four organs that have the abil ity to take up lactate; the liver and kidney in vivo and the heart in vitro can (Yudkin and Cohen, 1975; Brandt et aI., 1984). Significant cerebral lactate up take does not occur under normal physiologic con ditions (Gibbs et aI., 1942). The cerebral uptake of lactate during ischemia was first observed in 1928 (McGinty, 1928), and evidence for "oxidation" of lactate by brain tissue was discovered in 1931 (Ash ford and Holmes, 1931). Subsequent studies have more conclusively implicated lactate as a cerebral substrate ( Wortis et aI. , 1941; Sacks, 1965;
Oldendorf, 1971-1972; Nemoto et aI., 1974; Zimmer and Lang, 1975). Because the observation of cere bral lactate uptake or utilization has been associ ated with pathologic states surrounding the brain, the present study investigated lactate gradients across the cerebral circulation in humans during CPR, a pathologic state of cerebral hypoxia and global ischemia.
=
=
=
=
-
.
METHODS With approval from the Henry Ford Hospital Institu tional Review Board (lRB), patients presenting to the De partment of Emergency Medicine in normothermic, non traumatic, cardiopulmonary arrest were studied. Cardio pulmonary arrest was defined by the absence of pulse and respiration and later verified by the absence of an aortic pressure tracing. Patients received only basic cardiac life support (BCLS) prior to arrival in the emergency depart ment. All patients were clinically managed by emergency department physicians using advanced cardiac life sup port (ACLS) guidelines. Catheter placement, data record ing, and blood sampling were performed by an on-call research team, who had no active involvement in the clin ical management of the patient. Right atrial and central
Received July 11, 1990; revised September 25, 1990; accepted October 31, 1990. Address correspondence and reprint requests to Dr. E. P. Riv ers at Department of Emergency Medicine, Henry Ford Hospi tal, 2799 West Grand Boulevard, Detroit, MI 48202, U. S.A. Abbreviations used: Ao, aortic; BBB, blood-brain barrier; BCLS, basic cardiac life support; CPR, cardiopulmonary resus citation; CPT, catheter placement time; DT, down time; IRB, institutional review board; JVB; jugular venous bulb; RA, right atrial; ROSC, return of spontaneous circulation; TTA, total time of arrest.
479
480
E. P. RIVERS ET AL.
aortic catheters are routinely placed in the medical man agement of all cardiac arrests in our emergency depart ment. All catheters including the jugular venous bulb (JVB) catheter (routinely used in the medical manage ment of brain-injured patients) were placed without inter ference with the resuscitation procedures. Cardiopulmo nary resuscitation (CPR) was performed on all patients by a pneumatic compression device (Thumper, Michigan In struments, Grand Rapids, MI, U.S.A.) with a rate of 80 beats/min and a chest compression excursion of 2 in. A ventilation rate of 16 breaths/min was provided between compressions at a 5: 1 ratio. All catheters were simultaneously placed. A double lumen 20 cm 7.5F catheter (Cook, Bloomington, IN, U.S.A.) was placed percutaneously in the right atrium via the subclavian vein. The proximal port was used for drug administration and the distal port for blood sampling. A 20 cm, 20 gauge JVB catheter (Cook, Bloomington, IN, U.S.A.) was percutaneously placed via a retrograde an terior jugular vein approach above the thoracic inlet (Grif fith and Greenbaum, 1973; Jakobsen and Enevoldsen, 1989). A 60 cm 5.8F catheter (Bunegin-Albin, Cook, IN, U.S.A.) was placed in the aortic arch, either percutane ously or by cut down via the femoral artery. All catheters were placed using the guidewire technique. All catheter positions were verified radiographically at the end of the resuscitation. Catheters were connected to transducers (Sorenson Transpac, Abbot Systems, Bloomington, IN, U.S.A.) through a heparinized fluid flush system. This was set up and calibrated in advance. Transducers were zeroed to the midaxillary line. The resulting signals were amplified (Hewlett Packard, Sunnyvale, CA, U.S.A.), and pressure tracings recorded (Hewlett Packard 7758 multichannel) throughout the resuscitation. Blood lactate, glucose, hemoglobin, hematocrit, and blood gases were drawn simultaneously from arterial (aorta, Ao), right atrial (RA), and JVB sites immediately after all catheters were in place (T-O) and again 10 min (T-10) later. All lactates and blood gases were placed in ice immediately after being drawn. Lactates were mea sured by enzymatic assay (measurement error of ± 0.1O mM) (Lloyd et aI., 1978; Weil et aI., 1986; Gerol et aI., 1987). Return of spontaneous circulation (ROSC) was defined as development of a spontaneous pulse wave form on the arterial tracing associated with a systolic blood pressure greater than 60 mm Hg. Patients who maintained this for more than 1 min were considered to have had ROSC. In those patients who had ROSC, T-O was considered as the time of rearrest in the emergency department, and if ROSC was obtained during sampling, the samples were excluded from the study. All values are means ± SD.
the duration of ACLS from arrival to the emergency department until T-0 and this includes CPT. Mean CPR time was 32.6 ± 13 and 42.6 ± 12 min at T-O and T- 10, respectively. Total time of arrest (TTA) includes DT plus CPR time to T-O and T- 10. The mean TTA was 50.7 ± 18 and 60.3 ± 12 min at T-O and T- 10, respectively, the time from arrest to sam pling. Figure 1 provides a summary of study time characteristics. Prolonged times until sampling ex ist because the IRB required studying patients near the end of the resuscitation who were clinically deemed unsalvageable. Thus, catheter placement was delayed until this criteria was met as deter mined by the clinical team. The normal mean Ao - JVB lactate difference is -0. 18 mM; thus, in normal controls, JVB lactate levels are higher than Ao (Gibbs et aI., 1942). Of 45 patients at T-O, 32 (or 7 1%) had a mean net uptake of lactate. Of 40 patients at T-lO, 27 (or 68%) had a mean net uptake of lactate. At T-O and T- 10, the mean Ao - JVB lactate differences were 0.76 ± 1.86 and 0.80 ± 2.03 mM, respectively. This com pares to the normal controls, i.e., Ao - JVB -0. 18 mM. Both observations were statistically significant with a p value of 0, nonuptake Ao - JVB < 0). A multivariate regression analysis revealed that there were distinct differences in a number of measured variables between the uptake and non uptake groups. At T-O (Table 2), the uptake group was 13.6 years younger (p 0.004), had 8.5 min more of CPR (p 0.04), and had arterial lactates levels of 4.8 mM greater (p 0.005). When the same vari=
=
=
=
=
=
T-10 T-O sampling sampling
Arrival t o hospital
Cardiac arrest
Down time
Catheter placement I
RESULTS
Time
I
CPR time
Forty-five patients with a mean age of 66.6 ± 17.7 years entered the study. Mean down time (DT) or the time of arrest until arrival at the emergency de partment was 18 ± 1 1. 9 min. Patients received only BCLS before arrival at the hospital. Mean catheter placement time (CPT) was 14 ± 12 min. CPR time is
J
Cereb Blood Flow Metab. Vol.
11.
No. 3.
1991
Total time of arrest
o FIG. 1.
10
20
30
40
50
60
Study time (min): definitions and comparisons.
70
481
CEREBRAL LACTATE UPTAKE DISCUSSION
TABLE 1. Comparison of aortic (Ao), jugular venous
bulb (JVB), and right atrial (RA) lactate levels, and gradients (Ao - JVB) of study patients to normal controls Time: o min (n 45)
Time: 10 min (n 40)
Normal controlsa (n 50)
16.5 ± 0.8 15.7±0.7 15.6±0.7 0.76±1.90
17.0±0.8 16.2±0.7 16.4±0.8 0.80±2.00
1.1±0.3 1.9±0.3 1.9±0.1 -0.18±0.10
=
=
Ao JVB RA Ao - JVB
Studies have utilized the JVB for measurement of substrate uptake by the brain under various condi tions but none includes patients undergoing CPR. The validity of applying this technique during CPR relies on maintenance of jugular valvular compe tence, which prevents contamination from RA blood, thus allowing for accurate sampling of global cerebral venous drainage. The presence of cerebral blood flow during CPR is also a premise for inter preting this observation of lactate gradients. Human anatomists have identified jugular venous valves at the superior thoracic outlet. These valves have been shown to impede backflow of central ve nous blood into the jugular venous system (Fisher et aI., 1982). Jugular venous valvular competence is maintained with catheterization during CPR and maintains right atrial-jugular venous pressure gra dients (Niemann et aI., 198 1; Fisher et aI., 1982; GUdipati et aI., 1986; Paradis et aI., 1989). These facts support insignificant mixing of central and jug ular venous blood during CPR. Jugular venous blood contains an average extra cranial contamination of less than 2% in spontane ous circulation (Shenkin et aI., 1948; Lassen and Lane, 196 1). During cardiopulmonary arrest, cen tralization of blood from peripheral vessels occurs and is enhanced by the administration of epineph rine (Michael et al., 1984; Schleien et aI., 1986). Cephalic muscle and skin blood flow (extracranial blood flow) decreases with the administration of epinephrine (Michael et aI., 1984; Schleien et aI.,
=
Lactate in mM. Two-sample t test, p < 0.01 for Ao - JVB time 0 vs. Ao JVB normal. Two-sample t test, p < 0.01 for Ao - JVB time 10 vs. Ao JVB normal. Normal controls obtained from Gibbs et al. (1942). a
abIes were compared between the uptake and non uptake group at T- 1O (Table 3), the uptake group was 2.2 years younger (p 0.69), had 4.4 min more of CPR (p 0.30), and had higher arterial lactate levels of 3.02 mM (p 0.07). None of these vari ables was statistically significant at T- 10. Eighty-five percent of patients who had a lactate uptake at T-O maintained this uptake at T- 1O. Five patients in the T-0 group had a ROSC before T- 1O samples were obtained. All ROSC patients had a lactate uptake at T-O accounting for the significant decrease in the number of lactate uptake patients at T-10. This affected statistical significance between uptake and nonuptake groups in variables measured at T- 1O. There were no long-term survivors. All pa tients with ROSC died within 72 h of emergency department arrival. =
=
=
TABLE 2. Patient variables at time 0 min
(n) Age (years) CPR (min) DT (min) TTA (min) Ao lactate (mM) JVB lactate (mM) RA lactate (mM) Ao - JVB lactate (mM) Ao - JVB glucose (mg/d!) Hemoglobin (g/dl) Ao O2 content (mild!) JVB O2 content (mlldl) Cerebral O2 extraction pH Ao pH JVB pH RA pC02 Ao (torr) pC02 JVB (torr) pC02 RA (torr)
(43) (43) (41) (41) (45) (45) (41) (45) (22) (45) (43) (42) (42) (43) (42) (41) (43) (42) (41)
All patients, mean±SD
66.3 ± 18.0 32.6±12.8 18.4±11.7 50.7±18.1 16.5±5.3 15.7±4.8 15.6±4.8 0.76±1.9 69.5±172.0 10.1±2.5 13.5±4.4 4.35±3.26 0.67±0.21 7.17±0.29 6.95±0.21 6.95±0.22 38.2±31.8 100.0±49.2 104.1±47.2
(n) (30) (30) (28) (28) (32) (32) (29) (32) (17) (32) (30) (29) (29) (30) (29) (28) (30) (29) (28)
Uptake, mean±SD
(n)
62.3±19.5 34.9±12.6 18.8±13.2 53.4 ± 18.8 17.9±5.2 16.3±5.0 16.6±4.8 1.56±1.5 82.1±190.0 10.2±2.6 14.1±4.0 4.85 ± 3.46 0.66 ± 0.21 7.17±0.29 6.95 ± 0.24 6.94±0.23 37.8±30.7 101.7 ± 51.6 108.5±49.7
(13) (13) (13) (13) (13) (13) (12) (13) (5) (13) (13) (13) (13) (13) (13) (13) (13) (13) (13)
Nonuptake, mean± SD
p Value
75.5±9.2 27.3±12.3 17.5±7.0 44.8±15.7 13.1±3.9 14.4±4.0 13.5±4.2 -1.23±1.02 26.8 ± 90.1 9.8±2.2 12.0±5.0 3.25±2.54 0.68±0.22 7.19±0.29 6.94±0.16 6.97±0.18 39.2±35.3 96.4±45.4 94.7±41.5
0.004* 0.04* 0.76 0.16 0.005* 0.22 0.08 0.001* 0.54 0.15 0.15 0.14 0.76 0.79 0.91 0.69 0.90 0.75 0.39
(W) (S) (S) (S) (S) (S) (S) (S) (S) (S) (S) (S) (S) (S) (S) (S) (S) (S) (S)
* Statistically significant at the p < 0.05 significance level. (S), Student's two-sample t test for the equal variance setting. (W), Welch's two-sample t test for the unequal variance setting.
J
Cereb Blood Flow Metab, Vol.
11,
No.3,
1991
E. P. RIVERS ET AL.
482
TABLE 3. Patient variables at time 10 min
(n) Age (years) CPR (min) DT (min) TTA (min) Ao lactate (mM) JVB lactate (mM) RA lactate (mM) Ao - JVB lactate (mM) Ao - JVB glucose (mg/d!) Hemoglobin (g/dl) Ao O2 content (mild!) JVB O2 content (ml/dl) CNS O2 extraction pH Ao pH JVB pH RA pC02 Ao (torr) pC02 JVB (torr) pC02 RA (torr)
All patients. mean ±SD
(39) (39) (37) (37) (40) (40) (37) (40) (23) (40) (37) (38) (37) (37) (38) (38) (37) (38) (38)
67.3 42. 6 18. 2 60.3 17. 0 16. 2 16. 4 0. 80 74. 9 9. 2 12. 3 3. 54 0.68 7. 22 6.93 6.91 37. 6 94. 3 99. 4
±18. 2 ±11. 6 ±12. 3 ±17. 4 ±4. 8 ±4. 3 ±4.8 ±2.03 ±184. 0 ±2.5 ±4. 5 ±2. 43 ±0.21 ±0.27 ±0. 14 ±0.13 ±37. 5 ±50. 2 ±48. 9
(n) (26) (26) (24) (24) (27) (27) (25) (27) (17) (27) (26) (26) (26) (26) (26) (26) (26) (26) (26)
Uptake. mean ±SD 65. 7 44. 0 20. 7 64. 2 18. 0 16. 1 16. 3 1. 86 100.8 9. 6 12. 2 3.33 0.70 7. 20 6. 93 6.91 42. 6 103.3 106.4
±19. 3 ±10. 6 ±13. 2 ±16.3 ±4.3 ±4. 1 ±4.6 ±1. 43 ±214.0 ±2.3 ±4. 5 ±2.25 ±0. 20 ±0. 29 ±0.14 ±0.11 ±40. 6 ±53. 1 ±51. 4
(n ) (13) (13) (13) (13) (13) (13) (12) (13) (6) (13) (II) (12) (II) (11) (12) (12) (II) (12) (12)
Nonuptake, mean ±SD 69. 5 39. 4 13. 7 53. 1 15. 0 16. 4 16.7 - 1. 42 18. 8 8.4 12. 6 3. 99 0.61 7. 26 6. 94 6. 89 25. 7 74. 7 84.1
±16. 1 ±13.2 ±9. 4 ±17. 6 ±5. 3 ±4. 9 ±5. 4 ±1. 07 ±84. 1 ±2. 8 ±4. 8 ±2. 82 ±0. 23 ±0.24 ±0.16 ±0.18 ±27. 4 ±38. 1 ±40.6
p Value 0. 54 0. 24 0. 10 0. 06 0. 06 0. 86 0. 81 0.001* 0.25 0.15 0. 82 0. 44 0.25 0. 56 0.90 0.65 0.22 0. 10 0.19
(W) (S) (S) (S) (S) (S) (S) (S) (W) (S) (S) (S) (S) (S) (S) (S) (S) (S) (S)
* Statistically significant at the p < 0.05 significance level. (S), Student's two-sample t test for the equal variance setting. (W), WeIch's two-sample t test for the unequal variance setting.
1986). The low-flow state and the use of catechola mines in CPR enhances shunting of blood from sur face cranial vessels to more central venous and ar terial cerebral vessels (Michael et aI. , 1984; Schleien et aI., 1986). Thus, under conditions of CPR, extracerebral contamination should be less than prearrest states. Cerebral blood flow during CPR has been shown to range from 20 to 50% of normal (Voorhees et aI., 1980; Niemann, 1984). The venous valves at the thoracic outlet maintain a pressure gradient during compression systole and therefore enable cerebral blood flow (Voorhees et aI., 1980; Niemann et aI., 198 1; Niemann, 1984). In the presence of cerebral blood flow, comparing central aortic and JVB sam pling during CPR is a valid representation of uptake across the cerebral circulation. Lactate is a product of anaerobic metabolism and indicates tissue oxygen deprivation (Mizok, 1987). It is normally produced by the brain at a rate of 0.42 mg/IOO g of brain/min (Scheinberg et aI., 1964). Studies measuring Ao - JVB lactate in normal in dividuals have noted a negative gradient (i.e., JVB greater than Ao, a net release of lactate from the cerebral circulation) (Gibbs et aI. , 1942; Lying Tunell et aI., 1980). Under conditions of CPR, evidence exists for anaerobic metabolism. The inadequacy of cerebral oxygen delivery is substantiated by elevated cere bral oxygen extraction ratios of 0.66 and 0.67 (nor mal is
