prognostic significance of ventricular csf lactic acidosis

J Neurosurg 65:615-624, 1986

Prognostic significance of ventricular CSF lactic acidosis in severe head injury

ANTONIO A. F. DESALLES, M.D. , HERMES A. KONTOS, M.D. , PH.D. , DONALD P. BECKER, M.D. , MILDRED S. YANG, PH.D., JOHN D. WARD, M.D. , RICHARD MOULTON, M.D. , HANNS D. GRUEMER, M.D. , HARRY LUTZ, PH.D. , ANGELO L. MASET, M.D. , LARRY JENKINS, PH.D., ANTHONY MARMAROU, PH.D., AND PAUL MUIZELAAR, M.D. , PH.D.

Division of Neurological Surgery, Department of Surgery, Medical College of Virginia, Virginia Commonwealth University, Richmond, Virginia

v" Brain-tissue acidosis inferred by cerebrospinal fluid (CSF) lactic acidosis is considered to play an important role in the clinical course of severe head injury. Ventricular CSF lactate concentration was studied in 19 patients during the first 5 days after severe head injury. All patients were intubated, paralyzed, and artificially ventilated so that PaCO2 was kept at 33.2 _ 5.0 mm Hg and PaO2 at 122 - 18 mm Hg (mean ___ standard deviation). The mean Glasgow Coma Scale score on admission was 5.73 ___ 2.42. The first CSF sample was drawn within 18 hours after head injury. Over the first 4 days postinjury, patients with a poor outcome had significantly higher ventricular CSF lactate levels than did those with moderate disabilities or a good outcome. Patients showing favorable outcome had a significant decrease in ventricular CSF lactate levels 48 hours after injury. This decrease was not observed in patients with a poor outcome. Increased ventricular CSF lactate concentration was also reliably associated with increased intracranial pressure (ICP). Ventricular CSF lactate levels did not correlate with the magnitude of intraventricular bleeding. Arterial and jugular venous blood lactate levels, although high after head injury, were usually lower than the levels in the ventricular CSF and reached a normal range by the 3rd day following head trauma. At that time, the ventricular CSF lactate concentration was still above normal in patients with a poor outcome but had decreased to normal in patients with moderate disabilities or a good outcome. Ventricular CSF pH did not generally correlate with the ventricular CSF lactate concentration in patients under controlled ventilation; however, in a few patients close to death or with ventricular infection, a correlation was noted. Ventricular CSF lactate levels were not related to cerebral blood flow. In this study, profiles of ventricular CSF lactate concentration are defined in relation to the patients' clinical course and outcome. High ventricular CSF lactate concentration is present within 18 hours after severe head injury. Its decrease to normal in the following 48 hours is a reliable sign of clinical improvement; however, ventricular CSF lactate levels that are persistently high or that increase over time indicate the patient's deterioration. Serial assessment of ventricular CSF for acid-base status and metabolites in head-injured patients with a ventricular catheter already placed for ICP monitoring is useful in the evaluation of prognosis and clinical course.

KEY WORDS ~ head injury �9 cerebrospinal fluid �9 lactic acidosis �9 prognosis

B RAIN metabo l i c acidosis is cons idered to p lay

an i m p o r t a n t role in cell damage in var ious pathologica l states. 5~ W e bel ieve that , follow-

ing mechanica l b ra in injury, subcel lu lar e lements m a y become dysfunct ional , causing a focal and genera l ized brain- t issue acidosis tha t m a y be de t r imen ta l for po- tent ia l ly viable cells. M a n y studies have re la ted the degree o f bra in acidosis, inferred by the levels o f cere- brosp ina l fluid (CSF) lactate and pH, to the o u t c o m e after head injury. 9'11-13,17,20,31,32,37,40,54,59,67 Some o f these

studies c o m p a r e d the course o f C S F lactate concen-

t ra t ion in the acute phase after head in jury with the pa- t ient ' s cl inical course a n d f o u n d it to be re la ted to outcome. 9,~,~2,31,37,4~ Lac ta te concen t r a t i on has been shown to increase in b ra in t issue i tself af ter expe r imental bra in injury. 21"62'65 W e have d e m o n s t r a t e d in a smal l group o f pat ients tha t b ra in ex t race l lu la r cor t ical p H is abnorma l ly low close to areas o f con tus ion a n d / o r compress ion by a mass les ion (in prepara t ion) . More - over, exper imenta l b ra in - in ju ry s tudies in our laboratory have ind ica ted d e v e l o p m e n t o f mi ld bra in- t i ssue acidosis far f rom areas o f con tus ion in the absence

J. Neurosurg. / Volume 65/November, I986 615

of cerebral ischemia and/or hypoxia. 65 Rosner and Becker, 52 in a recent experimental study, demonstrated that treating brain acidosis reduced the incidence of high intracranial pressure (ICP) and decreased the mor- tality in animals under controlled ventilation over a 3- day period following brain injury. Wagner, et al.,62 have shown a very large increase in lactate concentration in the white matter within the territory of vasogenic edema of contused areas shortly after experimental head injury. Therefore, we have analyzed the ventricular CSF lactate concentration during the 5 days following severe head injury in patients under controlled ventilation, and have studied its correlation with development of increased ICP, cerebral blood flow (CBF), and the acid- base status. In addition, we have defined profiles of ventricular CSF lactate concentration in relation to the patients' clinical course and outcome.

Clinical Material and Methods

Patient Population The 19 patients in the study were selected from a

group of 36 severely head-injured patients who remained unresponsive at least 6 hours after admission to the hospital (mean Glasgow C o m a Scale (GCS) 6~ score of 5.97 _+ 2.12 (standard deviation)). The study group patients were aged 18 to 81 years (average 36 years) and had a mean GCS score of 5.73 ___ 2.42. These patients comprised the control group of a clinical trial for t reatment of brain-tissue acidosis and were selected on the basis of the level of hyperventilation. Only patients whose PaCO2 was maintained above 28 m m Hg were accepted for this study. Following diagnosis and treatment of any intracranial mass large enough to warrant surgical decompression, consent was obtained and patients were randomly assigned to one of three protocol groups. Group I was the control group treated with mild hyperventilation (PaCO2 maintained at 33 to 37 m m Hg); Group II received t romethamine (THAM)* plus hyperventilation (PaCO2 maintained at 22 to 26 m m Hg); Group III was treated with hyperventilation only (PaCO2 maintained at 22 to 26 m m Hg). 3 All patients were managed in a standard fashion as described elsewhere. 4'4~'63 This group of 19 patients with mild hyperventilation was selected for this study because high levels of hyperventilation may increase CSF

9 19 47 lactate concentration. ' ' These patients were kept on controlled ventilation with PaCO2 maintained at 33.2 _+ 5.0 m m Hg and PaO2 at 122 _+ 18 m m Hg. End- expiratory CO2 was monitored continuously with a capnometer, t which allowed us to keep the PaCO2 within this range.

A. A. F. DeSalles, et al.

Sample Collection and Analysis All 19 patients in this study had undergone place-

ment of a polyethylene catheter in the lateral ventricle for monitoring ICP, from which ventricular CSF was collected within 18 hours after injury and then at regular intervals of 12 hours. Arterial blood was obtained from a catheter inserted into the radial artery for continuous monitoring of arterial blood pressure, and jugular venous blood was drawn from a catheter inserted in the jugular vein with the tip placed into the jugular bulb. Blood samples were gathered at the same t ime as the collection of ventricular CSF samples.

Lactate concentrations in the ventricular CSF and blood were determined by an Automated Clinical AutoAnalyzer~ modification of the method by Mar- bach and Wei136 in which lactate is oxydized to pyruvate. Ventricular CSF and blood pyruvate concentrations were determined by the method of Marbach and Weil. Ventricular CSF hemoglobin concentration was determined according to the spectrophotometric technique of Shinowara) 5

Ventricular CSF and arterial blood acid-base parameters were measured in a blood gas analyzer.w The pH values were corrected for the patient 's temperature. Bicarbonate concentration was calculated from the pH and pCO2 values, using the Henderson-Hasselbach equation with a coefficient of dissociation (pK) of 6.10 for blood 22 and 6.13 for CSF.~ The solubility coefficient of CO2 was 0.031 for blood and 0.032 for CSF. Before the CSF samples were collected through the intraventricular catheter, the syringe was flushed with the patient's ventricular CSF and no air bubbles were allowed to enter. Arterial blood was collected in a heparinized syringe with the same care. No more than a 2-minute interval was allowed between collection of ventricular CSF and arterial blood. The normal range of lactate levels has two standard deviations of confidence.

For analysis of the biochemical data, the patients were separated into two groups based on their Glasgow Outcome Scale 2v score. The favorable category included those patients who had good outcome or were moderately disabled. Patients with poor outcome were severely disabled, vegetative, or dead. The follow-up period was from 6 months to 2 years. In addition, patients were separated into three groups according to the course profile of their ventricular CSF lactate values. In Profile 1, the values rapidly reached normal range within 48 hours after injury; in Profile 2, the values were abnormally elevated until Day 4 after injury, then declined slowly; in Profile 3, values presented a sustained increase above the level measured on Day 1.

In classifying the ICP, care was taken to exclude

* Hydroxymethyl aminomethane (tromethamine) is an in- tra- and extracellular buffer used in addition to hyperventilation to reverse brain acidosis.

t Capnometer, Model HP 47210A, manufactured by Hew- lett-Packard, Waltham, Massachusetts.

Automated Clinical AutoAnalyzer obtained from Du- Pont, Wilmington, Delaware.

w Blood gas analyzer, Coming 175 automatic pH/blood gas system, manufactured by Coming Medical, Medfield, Massa- chusetts.

616 J. Neurosurg. / Volume 65/November, 1986

CSF lactic acidosis in severe head injury

spuriously high ICP recordings attributable to suction- ing, coughing, gross head movements, or improper transducer calibration. Patients were considered to have a high ICP when their ICP reached high levels in spite of aggressive therapy previously described, 4,41,63 and were classified according to the highest sustained ICP level. "Sustained" was defined as a period of at least 2 minutes during the 5 days of study. Four groups were established: Group I included those whose ICP remained under 20 m m Hg; Group II comprised those who had ICP levels between 20 and 40 mm Hg; Group III was composed of those with ICP reaching levels between 40 and 60 m m Hg; and Group IV included those whose ICP reached levels above 60 mm Hg.

Eleven patients underwent CBF measurements within 2 hours of the time of ventricular CSF collection for lactate determination. Cerebral blood flow was measured by the xenon-133 inhalation technique as described by Obrist, et al. 45 Details of the procedures were reported elsewhere. 42

Statistical Analysis

Student's unpaired or paired t-tests were employed for parametric analysis. Variability is expressed as the standard deviation.

Results

The eight patients with favorable outcome had significantly elevated lactate concentrations in the ventricular CSI= and in the arterial and jugular venous blood on the day of head injury (p < 0.001) which returned to normal on the 2nd day. The 11 patients with poor outcome had elevated ventricular CSF lactate concentrations during the entire period of study (p < 0.001). The course of the arterial and jugular venous blood lactate concentrations was dissociated from that of the ventricular CSI= in this poor-outcome group. Lactate levels reached the normal range in blood on the 3rd day after injury, but were still elevated in the ventricular CSI= on the 4th day (p < 0.001). Comparing the two outcome groups, our results indicated that ventricular CSF lactate concentrations remained higher (p < 0.01) during the entire period of study in the group with poor outcome. In blood, this was the case only on the day of head injury (p < 0.05) (Figs. 1 and 2 and Table 1). The lactate/pyruvate (L/P) ratio was higher in the poor- outcome group during the entire period of study, but it reached statistical significance only on the day of injury (p < 0.05). In the good-outcome group, the L/P ratio remained in the normal range over the entire period of study (Table 1).

Bleeding in the ventricular compartment was not the main cause of lactate increase in ventricular CSF; this was concluded because the highest lactate concentration was not associated with the highest hemoglobin concentration and because there was no significant difference in hemoglobin concentration in the ventricular CSI=

5.O A [ ] ARTERIAL 4.0 EAVORASLE OUTCOME ,JUGULAR 30 VENOUS

J 0

6.0 I OOR 5.0 i 4 . 0

3.0

~ 2.0

o DAY I DAY 2 DAY 5 DAY 4 NORMAl

FIG. 1. Lactate concentration course in arterial blood (hatched bars), ventricular cerebrospinal fluid (CSF, open bars), and jugular venous blood (black bars). The lines on the top of each bar represent standard deviation. A" Results in eight patients with favorable outcome. Lactate concentration was significantly elevated on the day of head injury (p < 0.001) and reached normal levels on Day 2 in arterial and jugular bulb blood and ventricular CSF. The ventricular CSF lactate concentration had a course dissociated from that of arterial and jugular bulb blood. B" Results in 11 patients with poor outcome. The ventricular CSF lactate concentration was elevated during the 4 days following head injury (p < 0.001). The arterial and jugular bulb blood lactate concentration had a course dissociated from that of the ventricular CSF. It reached the normal range in blood on Day 3 after injury, while in the ventricular CSF it was still elevated on Day 4 (p < 0.001). Comparing the two outcome groups, ventricular CSF lactate concentration remained higher in the group with poor outcome during the entire period of the study (p < 0.01). In blood, this was the case only on the day of head injury (p < 0.05).

~ 6.0

w ,~ 4 0 POOR OUTCOME I--

2.0 FAVORABLE OUTCOME n=8

DAYS FIG. 2. Ventricular cerebrospinal fluid lactate concen-

tration course in head-injured patients. Notice the significant elevation in the group with poor outcome, as compared with the group with favorable outcome (p < 0.01). Vertical lines = standard deviation; hatched area = normal range. Favorable outcome indicates moderate disability or good outcome; poor outcome indicates patients who became severely disabled or vegetative, or who died.

J. Neurosurg. / Volume 65/November, 1986 617

7 . 0 �9

n = 3 8 r = O. 257

5 0 ; . ~ o �9

3 . 0

, .o

i i i i i 0 50 I O0 150 200

HEMOGLOBIN ( m g / d l )

FIG. 3. Plot o f hemoglobin concentra t ion versus lactate concentrat ion. Each point represents concentrat ions determined in the same ventricular cerebrospinal fluid (CSF) sample. There was no correlation between the hemoglobin concentration and lactate concentra t ion in the ventricular CSF. Hatched area = normal range of ventricular CSF lactate concentrations.

between the two outcome groups (Fig. 3). No association was observed between the high lactate concentration and low pH in the ventricular CSF of head-injured patients under controlled ventilation (Fig. 4); however, in four patients with extremely high ventricular CSF lactate concentration due to infection or in those ap- proaching death, pH fell to values lower than the normal range. Ventricular CSF values of buffer base (bicarbonate) were significantly low in all patients in this study (p < 0.005). No significant difference in ventricular CSF pH, buffer base, or pCO2 was found between the two outcome groups (Table 2).

TABLE 1

Ventricular CSF lactate and lactate/pyruvate (L/P) ratio in severely head-injured patients*

ParameterS- Outcome Day 1 Day 2 Day 3 Day 4

lactate (mmol/liter)

p value

L/P ratio

p value

favorable mean SD

poor mean SD

favorable mean SD

poor mean SD

4.0 2.1 1.9 1.8 0.9 0.4 0.2 0.2

5.5 4.8 4.4 3.2 1.8 1.9 2.2 0.9

<0.05 <0.005 <0.05 <0.005

14.5 15.5 14.3 12.8 4.1 1.6 6.7 3.0

22.5 21.6 23.6 18.5 6.6 7.5 11.8 9.2 0.05 NS NS NS

* CSF = cerebrospinal fluid; SD = standard deviation; NS = not significant. Day 1 = day of injury.

I" Normal range for lactate: 0.8--2.2 mmol/liter; normal range for L/P ratio: 4.3--20.3.

A. A. F. DeSal les , et al.

7 . 5

n=40

=o 5 . 0 �9

E �9 �9 �9 �9 �9

-

2 . 5 �9

I

7;0 ,;5 ,;o pH

FIG. 4. Plot o f ventricular cerebrospinal fluid (CSF) pH versus ventricular CSF lactate concentrat ion. Each point represents pH and lactate concentrat ion obtained from samples collected at the same time. There was no correlation between these two factors in head-injured patients under controlled ventilation. Hatched area = normal range of ventricular CSF lactate concentrat ion; vertical broken line = normal ventricular CSF pH.

V e n t r i c u l i t i s o c c u r r e d in t w o p a t i e n t s in th i s series. Th i s c o m p l i c a t i o n c a u s e d a s t r ik ing i n c r e a s e i n ven t r i c - u lar C S F l ac ta t e c o n c e n t r a t i o n as well as a d e c r ea se in v e n t r i c u l a r C S F p H (Fig. 5).

H i g h levels o f v e n t r i c u l a r C S F lac ta te w e r e p r e s e n t by D a y 2 in p a t i e n t s w h o t e n d e d to d e v e l o p h igh I C P

TABLE 2

Course of pH, bicarbonate, and pC02 in ventricular CSF*

ParameterS- Outcome Day 1 Day 2 Day 3 Day 4

pH

p value

bicarbonate (mmol/liter)

p value

pCO2 (mm Hg)

p value

favorable mean SD

poor mean SD

favorable mean SD

poor mean SD

favorable mean SD

poor mean SD

7.32 7.37 7.36 7.34 0.03 0.03 0.03 0.02

7.33 7.31 7.37 7.30 0.05 0.03 0.01 0.06

NS NS NS NS

18.9 19.0 19.5 20.2 1.8 2.4 2.4 1.5

17.9 17.1 19.9 17.9 3.0 2.2 1.8 5.7

NS NS NS NS

38 35 37 40 5 1 5 3

34 37 36 35 5 6 3 8

NS NS NS NS

* CSF = cerebrospinal fluid; SD = standard deviation; NS = not significant. Day 1 = day of injury.

t Normal range for pH: 7.32-7.36; normal range for bicarbonate: 19.9-23.7 mmol/liter; normal range for pCO2:38--46 mm Hg.

618 J. Neurosurg. / Volume 65 /November , 1986


105 -

~ 7 . 5 -

~ 6 0 -

~ 4 5 -

~ 3 0 -

TIME ( HOURS AFTER INJURY)

FIG. 5. Time course of lactate values in the ventricular cerebrospinal fluid (CSF) in a severely head-injured patient with ventriculitis. Notice the abnormally sustained high lactate concentrations over the 4 days following injury and the sudden striking increase coupled with the onset of ventriculitis. Hatched area = normal range of ventricular CSF lactate concentrations.

(p < 0.005). This cor re la t ion s tar ted to appea r by the day o f in jury (Day 1, p < 0.01); however , i t became more evident on D a y 2 (Fig. 6). Al l pa t ients classified in ICP G r o u p IV presen ted uncont ro l l ab le ICP. Pa t ien ts in G r o u p III requi red intensive t r ea tmen t for lower ing ICP (ventr icular d ra inage and manni to l ) . One pa t i en t had only one episode in which ICP rose above 40 m m Hg; his ICP re sponded p r o m p t l y to t r e a tmen t a n d d id not increase again. His ini t ial vent r icu la r C S F lactate concen t ra t ion was 5.5 m m o l / l i t e r and r e m a i n e d above the n o r m a l range dur ing the ent i re pe r iod o f study. Pat ients in G r o u p II requi red less in tensive ICP treatment . One pa t ien t in this g roup presented on ly one peak o f ICP above 20 m m Hg; his ini t ia l vent r icu la r CSF lactate level was 3.9 m m o l / l i t e r and reached a no rma l level the next day. All o ther pat ients in G r o u p II presented sus ta ined per iods o f ICP above 20 m m Hg.

F o u r o f the five pa t ien ts whose ICP rose to over 60 m m Hg m a y have had episodes o f cerebral hypoper fu - sion as assessed by low cerebral per fus ion pressure (CPP). The pa t ien t who d id not present cr i t ical ly low CPP dur ing the pe r iod o f s tudy had serial C B F measurements o f 48, 32, and 41 ml /100 g m / m i n on Days 1, 2, and 3, respectively. N o hypoper fus ion was detected. In this pat ient , the vent r icu la r C S F lactate level increased f rom 3.9 m m o l / l i t e r on D a y 1 to 7.6 m m o l / l i ter on Day 3 with apparen t ly sufficient CBF. N o corre la t ion was observed between the vent r icu la r C S F lactate level and C B F measu red wi th in 18 hours af ter injury. The highest vent r icu la r C S F lacta te concent ra - t ions were not associa ted with cr i t ical ly low CBF's . N o large hemispher ic or regional b lood flow differences were detected in these pat ients . Therefore, we have c o m b i n e d the regional and hemispher ic m e a s u r e m e n t s into one m e a n value (Table 3).

The profiles o f the vent r icu la r C S F lactate course in the 19 pat ients in this series were as follows: eight pat ients exhib i ted Profi le 1, wi th a r ap id decrease o f

9 . 0

7 .5

6 . 0

4 . 5

3 0

1.5

n=19 �9 r �9 0 . 7 2 8 6 p< 0 . 0 0 5 �9

i i i i 0 - 2 0 2 0 - 4 0 4 0 - 6 0 > 6 0

M A X I M A L ICP L E V E L ( m m H g )

FIG. 6. Relationship between the maximal intracranial pressure (ICP) range reached during the 5-day period of the study with ventricular cerebrospinal fluid (CSF) lactate concentration found on Day 2 in 19 patients. There was a significant positive correlation: high levels of ventricular CSF lactate on Day 2 were present in patients who tended to develop high ICP (r = 0.7286; p < 0.005). This correlation started to appear on the day of injury (Day 1: r = 0.519; p < 0.01); however, it became more evident on Day 2. Hatched area = normal range of ventricular CSF concentration. Four of the patients who presented ICP over 60 mm Hg may have had episodes of cerebral hypoperfusion as assessed by low cerebral perfusion pressure.

lactate to the no rma l range wi th in 24 to 48 hours after injury, associated with favorable ou tcome; seven pa- t ients showed a Profile 2 course, wi th slow decrease o f lactate over the 5-day pe r iod af ter in jury , associa ted with a clinical course o f diff icul t m a n a g e m e n t a n d poo r outcome; and four pa t ien t s had a Profi le 3 course with

TABLE 3 Relationship between the ventricular CSF lactate level and

CBF on the day of injury*

CBF Case Lactate ( ml/100 PaC02 No. (mmol/liter) gm/min) (mm Hg)

1 3.4 55 44 2 2.0 35 25 3 4.7 36 34 4 4.9 20 35 5 3.9 13 30 6 7.0 21 32 7 3.9 48 38 8 8.0 26 32 9 7.6 24 35

10 4.1 28 29 11 2.8 15 28

* CSF = cerebrospinal fluid: CBF = cerebral blood flow, reported as the average of the two hemispheres. PaCO2 was determined at the time of CBF measurement.

J. Neurosurg. / Volume 65/November , 1986 619

- - 7 5 -

~ 6 0 - E - - 4 5 -

J 1 5


A B C

j *

__ MOOE RATE LY DISABLED VEGETATIVE ~ - ' - , ' ~ " ~ " 4 ~ " " 4 " ' ~ D E A O

! ! ! ! ! i ! ! I ! i ! !

24 48 ZZ 96 leO 24 48 72 96 a4 48 72 96

TIME (HOURS AFTER HEAD INJURY)

FIG. 7. Profiles of the ventricular cerebrospinal f luid (CSF) lactate concentration course in three head- injured patients wi th different outcomes. Hatched area = normal range of ventricular CSF lactate concentration. A: Profile of a 20-year-old man with an admission Glasgow Coma Scale (GCS) score of 6. His intracranial pressure (ICP) reached levels near 30 mm Hg. He is moderately disabled 8 months after trauma. B: Profile of a 44-year-old man wi th an admission GCS score of 5. His ICP started to rise on the day after injury and reached levels over 40 mm Hg during the period of the study. He survived in a vegetative condition. C: Profile of an 18-year-old man with an admission GCS score of 4. The ICP rose to over 30 mm Hg on the 3rd day after trauma and he died on the 5th day when his ICP became uncontrollable.

rapid increase of lactate after injury and a fulminant rise in ICP, followed by death (Fig. 7 and Table 4).

Discussion

These studies demonstrate that high ventricular CSF lactate concentrations that are sustained or increase further are associated with poor outcome. Moreover, the increase in lactate concentration after a period of decrease appears to be an important indicator of onset of secondary insult. The cause of increase in CSF lactate concentration following head injury is still unresolved. It is possible that the initial high lactate levels are a consequence of the mechanical injury itself. It may cause an impairment of mitochondrial function, im- peding the pyruvate molecule from being metabolized through the Krebs cycle, resulting in accumulation of lactic acid in the intracellular as well as in the extracellular space, and subsequently appearing in the

TABLE 4 Correlation of ventricular CSF lactate course profiles with

outcome in 19 head-injured patients

Lactate No. of Glasgow Outcome Scale Score Course Cases at 6 Mos*

Profile 1 8 5 good 3 moderately disabled

Profile 2 7 3 severely disabled 2 vegetative 2 deadt

Profile 3 4 4 dead:~ * Three of the patients followed for more than 6 months changed

their outcome category. One moderately disabled patient recovered completely. Of the two vegetative patients, one died and the other was severely disabled at l year.

]- Death after the acute phase of trauma due to clinical complica- tions.

:~ Death in the acute phase after trauma.

ventricular compartment. A further ventricular CSF lactate increase, found mainly in patients with a Profile 3 course, and the sustained high lactate levels found in patients with a Profile 2 course, suggest that the mitochondrial dysfunction caused by the mechanical injury may persist.

Recent studies in our laboratory support this hypothesis. They have demonstrated development of brain- tissue lactic acidosis in white and gray matter far from areas of contusion in experimental traumatic brain injury. 65 Data extracted from other experimental series with similar mechanical injury showed no reduction in CBF or development of hypoxia.15,34 Low brain cortical extracellular pH measured in patients during surgery correlated well with low ventricular CSF pH and high ventricular CSF lactate concentrations as well (unpub- lished data). However, undetected hypoxia at the scene of injury may also be considered as an initial causal factor of increased ventricular CSF lactate concentration in our patients. Also, although the CBF measured in these patients did not indicate association of high ventricular CSF lactate levels with critically low CBF, these data do not completely exclude the possibility of episodes of global or regional cerebral hylzoperfusion as one of the causes of ventricular CSF lactate increase. In a recent study, Yoshino, et al., 66 have presented data suggesting cerebral hypoperfusion in patients within 2 to 3 hours after severe head injury. It is possible that our CBF measurements were performed during the phase of reperfusion. Similar CBF findings following severe head injury were observed by Obrist, et al., 44 and Mendelow, et al. 39 The CBF measurements in these two studies were made at least 6 hours after the time of injury, as were ours.

Jenkins, et al., 26 have shown that traumatized cells are more vulnerable to ischemia. This finding supports the idea of subcellular dysfunction following traumatic brain injury; for example, traumatized cells are less

620 J. Neurosurg. / Volume 6 5 / N o v e m b e r , 1986


capable of overcoming the adverse environment present after injury. Whether this adverse milieu for cell healing is caused by the dysfunctional traumatized cells them- selves which may pour acid metabolites into the extracellular space, or is due to development of secondary insults to these already ill cells, is still a matter for further study. After trauma, focal edema may develop in the damaged cortex and white matterP 4,62 Edematous tissue may become hypoxic if an increase in intercap- illary distance impedes the oxygen exchange between the blood vessels and viable parenchyma. 25,3~ Thus, a continuous process of subcellular dysfunction aggra- vated by local hypoxia in the edematous areas may explain the high ventricular CSF lactate concentration in patients presenting Profile 2 and 3 lactate courses.

Caution must be exercised in evaluating brain metabolic state based on CSF lactate concentration following head injury. Blood is often present in the ventricular CSF in such cases, and blood in the CSF is known to increase lactate concentration.~8 In our study, there was no association between the highest lactate concentration and the highest hemoglobin concentration in the ventricular CSF, neither was there a significant difference in hemoglobin concentration between the two outcome groups. However, the lactate concentration was higher in patients with poor outcome during the entire period of study. These findings suggest that the high ventricular CSF lactate concentration in the poor- outcome group indeed reflects a more severe subcellular disturbance than that present in the good-outcome group. Lactate/pyruvate ratio, believed to better reflect the tissue redox state than does the lactate concentration itself, ~8 was also elevated in the poor-outcome group; however, the degree of elevation was not statistically significant, probably due to the small number of patients in this study and the variability of the pyruvate assay technique. The CSF lactate concentration can also be influenced by the serum lactate concentration. 51'64 While it is possible that serum lactate influenced the ventricular CSF lactate concentration in this study, it is unlikely that it represents the major factor. The lactate concentration was consistently higher in the ventricular CSF than in blood and had a course dissociated from that of blood in both outcome groups. This dissociation was clear in the poor-outcome group (Fig. 1). A high serum lactate concentration is a usual re- sponse to systemic injury and does not have the same prognostic significance~~ that CSF lactate concentration has in head injury.

Lactate can be produced by the choroid plexus in vivo 23 or in vitro. 29 However, this source is too small to be responsible for an increase of more than fourfold in the ventricular CSF lactate concentration observed in this study. Valenca, et al., 6~ indicated that clearance of lactate from the CSF is accomplished mainly by diffusion into the brain tissue, and that lactate is further metabolized to pyruvate by the cells of the central nervous system (CNS). Prockop 48 demonstrated that processes other than simple bulk flow through the

arachnoid villi may be operative in lactate clearance from the CSF; lactate should, in physiological circum- stances, be cleared from the CSF faster than the larger molecules that leave the CSF compartment only by bulk flow. We have demonstrated that the creatine- kinase-BB-isoenzyme (CKBB), a protein that may be removed from CSF mainly by bulk flow, ~4 was cleared from the ventricular CSF much faster than was the lactate molecule following severe head injury. 49 Follow- ing traumatic brain injury, the bulk CSF flow appears not to be impaired severely enough to disturb the CSF dynamics? 8 Thus, these observations when taken to- gether favor the hypothesis that mechanical brain injury causes a mitochondrial dysfunction that is responsible not only for an overproduction of lactate from brain metabolism, but also for an abnormal ability of these dysfunctioning cells to clear the excess lactate in the tissue, as well as that produced in the choroidal plexus.

No correlation was found between ventricular CSF lactate concentration and ventricular CSF pH. The latter is not identical to focal-tissue pH, but is closely related to blood pH, since CSF pH close to the choroid plexus varies almost instantaneously with changes in arterial pH. 35 We have demonstrated that ventricular CSF pH is higher than the extracellular cortical pH in areas of contusion (in preparation). Thus, only in cases with very extensive brain lesions, and/or in cases in which the brain lesions are located near the ventricular system, may the ventricular CSF pH be expected to reflect a degree of acidosis similar to that in the injured tissue. On the other hand, for a decrease in CSF HCO3- concentration to take place, a flux of H + in or flux of HCO3- out of the ventricular compartment to the sur- rounding brain tissue or blood must occur. 56 The level of buffer base was significantly low in all patients in this study, reflecting its expenditure to buffer the lactic acid produced. The H2CO3 formed from the interaction between lactic acid and HCO3- is promptly removed from the CSF compartment in the form of CO276 Therefore, the level of controlled ventilation applied in these patients appeared to be sufficient to compensate for the ventricular CSF metabolic acidosis. However, it was not sufficient to compensate for patients with extremely high lactate levels in the CSF, and significantly low ventricular CSF pH was found associated with very low ventricular CSF buffer base in a few patients who were close to brain death, or in cases with infection in the ventricular compartment.

Obviously, as discussed above, secondary insults such as hypoxia and/or ischemia can change the profile of the ventricular CSF lactate curve following head injury. Of particular interest is the change caused by CNS infection. The increase in lactate levels when bacterial infection is present in the CSF is well known. 5,43 In- creased lactate concentrations due to infection in the ventricular system were found in two of our patients, causing a striking rise in lactate concentration. There- fore, increase in the CSF lactate concentration in a patient with a clinically favorable course after severe

J. Neurosurg. / Volume 65/November , 1986 621

head injury must be further investigated to rule out the possibility of a bacterial infection in the CSF compartment.

The increased ventricular CSF lactate concentration on the 2nd day after injury was significantly associated with development of pathological ICP levels during the period of study (patients with lactate Profiles 2 and 3). Cerebral hyperemia has been associated with tissue lactic acidosis in several pathological states, 33 and increased ventricular CSF lactate concentration (reflecting tissue lactic acidosis) was shown to correlate with cerebral hyperemia in severely head-injured patients. 8,12 Since hyperemia after head injury is known to be a cause of increased ICP, 6"16"44 lactate concentration in ventricular CSF may be related to increased ICP.

Brain edema formation is also an important etiolog- ical factor associated with development of increased ICP. It is known that the lactate level is increased in brain tissue shortly after experimental head injury, 21'62'65 particularly in the white mat ter within the territory of vasogenic edema of contused areas. 6z Increased tissue osmolality 2'24 and brain-tissue water content 28,46,53 following experimental cerebral ischemia have been at- tributed in part to tissue lactic acidosis. Busse and Hof fmann 7 reported good correlation between CSF lactate concentration and the extension of brain edema as evaluated by computerized tomography in patients with cerebral ischemia. Moreover, they found that the increase in CSF lactate precedes the formation of brain edema. Also, lactic acidosis is associated with the development of cytotoxic edema. Accumulat ion of lactic acid in tissue represents failure of the cell energy pro- duction to a certain extent. The lack of energy for maintenance of cell ionic homeostasis allows a net accumulat ion of electrolytes and water into the cyto- plasm with consequent cell swelling. 3~

Since lactate concentration in tissue, as well as in CSF, may be related to development of hyperemia and edema after brain injury, it was not unexpected that there is an association between increased ICP and increased ventricular CSF lactate concentration in patients with severe head injury. This association with the high ventricular CSF lactate concentration was clearer on Day 2 than on Day 1 (the day of injury), confirming the importance of sustained tissue lactic acidosis in the clinical course of head-injured patients.

In conclusion, the sustained high concentration of ventricular lactate found in patients with a poor outcome in this study indicates that a persistent dysfunction in brain energy metabolism is present following severe head injury. This dysfunction is associated with a clinical course difficult to manage and development of high ICP. Also, the accumulat ion of lactic acid requires functioning cells. These cells may be in a borderline state of recovery or death. Therapy that supports these viable cells may improve their quality and rate of healing, improving the clinical course and outcome after severe head injury. 3 Finally, ventricular CSF lactate determination after severe head injury is a


convenient procedure to perform in patients who already have a ventricular catheter inserted for ICP monitoring, adding valuable information about the prognosis and clinical course, and also serving as a diagnostic tool in onset of secondary insults over the critical period after severe head injury.

References

1. Alexander SC, Gelfand R, Lambertsen CJ: The pK' of carbonic acid in cerebrospinal fluid. J Biol Chem 236: 592-596, 1961

2. Bandaranayake NM, Nemoto EM, Stezoski SW: Rat brain osmolality during barbiturate anesthesia and global brain ischemia. Stroke 9:249-254, 1978

3. Becker DP: Brain acidosis in head injury: a clinical trial, in Becker DP, Povlishock JT (eds): Central Nervous Sys- tem Trauma. Status Report 1985. National Institute of Neurological and Communicative Disorders and Stroke. Richmond, Va: William Byrd Press, 1985, pp 229-242

4. Becker DP, Miller JD, Young HF, et al: Diagnosis and treatment of head injury in adults, in Youmans JR (ed): Neurological Surgery, ed 2. Philadelphia: WB Saunders, 1982, Vol 4, pp 1938-2083

5. Berg B, G~irdsell P, Skhnsberg P: Cerebrospinal fluid lactate in the diagnosis of meningitis. Diagnostic value compared to standard biochemical methods. Scand 3 Infect Dis 14:111-115, 1982

6. Bruce DA, Raphaely RC, Goldberg AI, et al: Pathophys- iology, treatment and outcome following severe head injury in children. Childs Brain 5:174-191, 1979

7. Busse O, Hoffmann O: CSF lactate and CT findings in middle cerebral artery infarction. A comparative study. Stroke 14:960-963, 1983

8. Cold G: Cerebral blood flow in the acute phase after head injury. Part 2: Correlation to intraventricular pressure (IVP), cerebral perfusion pressure (CPP), PaCO2, ventricular fluid lactate, lactate/pyruvate ratio and pH. Acta Anaesth Stand 25:332-335, 1981

9. Cold G, Enevoldsen E, Malmros R: Ventricular fluid lactate, pyruvate, bicarbonate and pH in unconscious brain-injured patients subjected to controlled ventilation. Acta Neurol Stand 52:187-195, 1975

10. Collins JA, Simmons RL, James PM, et al: The acid-base status of seriously wounded combat casualties: I. Before treatment. Ann Surg 171:595-608, 1970

11. Crockard HA, Taylor AR: Serial CSF lactate/pyruvate values as a guide to prognosis in head injury coma. Eur Neurol 8:151-157, 1972

12. Enevoldsen EM, Cold G, Jensen FT, et al: Dynamic changes in regional CBF, intraventricular pressure, CSF pH, and lactate levels during the acute phase of head injury. J Neurosurg 44:191-214, 1976

13. Enevoldsen EM, Jensen FT: Cerebral fluid lactate and pH in patients with acute severe head injury. Clin Neurol Neurosurg 80:213-225, 1977

14. Davson H: Physiology of the Cerebrospinal Fluid. Lon- don: J & A Churchill, 1967, p 165

15. DeWitt DS, Jenkins LW, Wei EP, et al: Effects of fluid- percussion brain injury on regional cerebral blood flow and pial arteriolar diameter. J Neurosurg 64:787-794, 1986

16. Fieschi C, Battistini N, Beduschi A, et al: Regional cerebral blood flow and intraventricular pressure in acute head injuries. J Neurol Neurosurg Psychiatry 37: 1378-1388, 1974



17. Gordon E: The acid-base balance and oxygen tension of the cerebrospinal fluid, and their implications for the treatment of patients with brain lesions. Acta Anaesthe- siol Scand Suppl 39:1-36, 1971

18. Granholm L: The effect of blood in the CSF on the CSF lactate, pyruvate and bicarbonate concentrations. Scand J Clin Lab Invest 23:361-366, 1969

19. Granholm L, Siesj6 BK: The effects of hypercapnia and hypocapnia upon the cerebrospinal fluid lactate and pyruvate concentrations and upon the lactate, pyruvate, ATP, ADP, phosphocreatine and creatine concentrations of cat brain tissue. Acta Physiol Scand 75:257-266, 1969

20. Gulati SC, Sood SC, Bali IM, et al: Cerebral metabolism following brain injury. I. Acid-base and pO2 changes. Acta Neurochir 53:39-46, 1980

21. Gurdjian ES, Webster JE, Stone WE: Experimental head injury with special reference to certain chemical factors in acute trauma. Surg Gynecol Obstet 78:618-626, 1944

22. Hamilton LH: Respiratory and blood gas analysis. Prog Clin Pathol 2:284-340, 1969

23. Hochwald GM, Magee J, Ferguson V: Cerebrospinal fluid glucose: turnover and metabolism. J Neurochem 44: 1832-1837, 1985

24. Hossmann KA: Pathophysiology of vasogenic and cytotoxic brain edema, in Hartmann A, Brock M (eds): Treat- ment of Cerebral Edema. Berlin/Heidelberg/New York: Springer-Verlag, 1982, pp 1-10

25. Hossmann KA, Takagi S: Osmolality of brain in cerebral ischemia. Exp Neurol 51:124-131, 1976

26. Jenkins L, Marmarou A, Lewelt W, et al: Increased vulnerability of the traumatized brain to early ischemia, in Baethmann A (ed): Mechanisms of Secondary Brain Damage. New York: Plenum Press, 1986, pp 273-286

27. Jennett B, Bond M: Assessment of outcome after severe brain damage. A practical scale. Lancet 1:480-484, 1975

28. Katzman R, Clasen R, Klatzo I, et al: Brain edema in stroke. Stroke 8:512-540, 1977

29. Kemeny A, Kutas F, Caspar J, et al: In vitro study of respiration and glycolysis in the choroid plexus under various experimental conditions. Biochemistry 26: 680-685, 1961

30. Kimelberg HK, Bourke RS, Stieg PE, et al: Swelling of astroglia after injury to the central nervous system: mechanisms and consequences, in Grossman RG, Gildenberg PL (eds): Head Injury: Basic and Clinical Aspects. New York: Raven Press, 1982, pp 31-44

31. King LR, McLaurin RL, Knowles HC Jr: Acid-base balance and arterial and CSF lactate levels following human head injury. J Neurosurg 40:617-625, 1974

32. Kurze T, Tranquada RE, Benedict K: Spinal fluid lactic acid levels in acute cerebral injury, in Caveness WF, Walker AE (eds): Head Injury. Philadelphia: JB Lippin- cott, 1966, pp 254-259

33. Lassen NA: Control of cerebral circulation in health and disease. Circ Res 34:749-760, 1974

34. Lewelt W, Jenkins LW, Miller JD: Effects of experimental fluid-percussion injury of the brain on cerebrovascular reactivity to hypoxia and to hypercapnia. J Neurosurg 56:332-338, 1982

35. Loeschcke HH, Sugioka K: pH of cerebrospinal fluid in the cisterna magna on the surface of the choroid plexus of the 4th ventricle and its effect on ventilation in experimental disturbances of acid base balance. Transients and steady states. Pflugers Arch 312:161 - 188, 1969

36. Marbach EP, Weil MH: Rapid enzymatic measurement of blood lactate and pyruvate. Use and significance of metaphosphoric acid as a common precipitant. Clin Chem 13:314-325, 1967

37. Marchau M, Calliauw L, Polderman J: La detrrmination du lactate et du pyruvate dans le LCR comme moyen de pronostic en cas de traumatismes cranio-crrrbraux graves. Neurochirurgie 26:299-304, 1980

38. Maser AL, Ward JD, Lutz HA, et al: Isolation of factors leading to raised ICP in head injured patients: a prelimi- nary report, in Miller JD, Teasdale GM, Rowan JO, et al (eds): Intracranial Pressure VI. Berlin/Heidelberg/New York: Springer-Verlag (In press, 1986)

39. Mendelow AD, Teasdale GM, Russell T, et al: Effect of mannitol on cerebral blood flow and cerebral perfusion pressure in human head injury. J Neurosurg 63:43-48, 1985

40. Metzel E, Zimmermann WE: Changes of oxygen pressure, acid-base balance, metabolites and electrolytes in cerebrospinal fluid and blood after cerebral injury. Acta Neuro- chit 25:177-188, 1971

41. Miller JD, Butterworth JF, Gudeman SK, et al: Further experience in the management of severe head injury. J Neurosurg 54:289-299, 1981

42. Muizelaar JP, Lutz HA III, Becker DP: Effect of mannitol on ICP and CBF and correlation with pressure autoregu- lation in severely head-injured patients. J Neurosurg 61: 700-706, 1984

43. Nishimura K: The lactic acid content of blood and spinal fluid. Proc Soc Exp Biol Med 22:322-324, 1924

44. Obrist WD, Langfitt TW, Jaggi JL, et al: Cerebral blood flow and metabolism in comatose patients with acute head injury. Relationship to intracranial hypertension. J Neurosurg 61:241-253, 1984

45. Obrist WD, Thompson HK Jr, Wang HS, et al: Regional cerebral blood flow estimated by 133xenon inhalation. Stroke 6:245-256, 1975

46. Plum F: What causes infarction in ischemic brain? The Robert Wartenberg Lecture. Neurology 33:222-233, 1983

47. Plum F, Posner JB: Blood and cerebrospinal fluid lactate during hyperventilation. Am J Physiol 212:864-870, 1967

48. Prockop LD: Cerebrospinal fluid lactic acid. Clearance and effect on facilitated diffusion of a glucose analogue. Neurology 18:189-196, 1968

49. Rabow L, DeSalles AAF, Becker DP, et al: CSF brain creatine kinase levels and lactic acidosis in severe head injury. J Neurosarg 65:625-629, 1986

50. Rehncrona S, Rosrn I, Siesj6 BK: Excessive cellular acidosis: an important mechanism of neuronal damage in the brain? Acta Physiol Stand 110:435-437, 1980

51. Ronquist G, Terrnt A: Cerebrospinal fluid markers of disturbed brain cell metabolism. Prog Neurobiol 18: 167-180, 1982

52. Rosner MJ, Becker DP: Experimental brain injury: suc- cessful therapy with the weak base, tromethamine. With an overview of CNS acidosis. 3 Neurosurg 60:961-971, 1984

53. Schuier FJ, Hossmann KA: Experimental brain infarcts in cats. II. Ischemic brain edema. Stroke 11:593-601, 1980

54. Seitz HD, Ocker K: The prognostic and therapeutic importance of changes in the CSF during the ztcute stage of brain injury. Aeta Neurochir 38:211-231, 1977

55. Shinowara GY: Spectrophotometric studies on blood serum and plasma. The physical determination of hemoglobin and bilirubin. Am J Clin Pathol 24:696-710, 1954

56. Siesj6 BK: Brain Energy Metabolism. New York: John Wiley & Sons, 1978, pp 288-323

57. Siesj6 BK: Cerebral circulation and metabolism. J Neu- rosurg 60:883-908, 1984

J. Neurosurg. / Volume 65/November, 1986 623


58. Siesj6 BK, Wieloch T: Brain injury: neurochemical aspects, in Becker DP, Povlishock JT (eds): Central Nervous System Trauma. Status Report 1985. National Institute of Neurological and Communicative Disorders and Stroke. Richmond, Va: William Byrd Press, 1985, pp 513-532

59. Sood SC, Gulati SC, Kumar M, et al: Cerebral metabolism following brain injury. II. Lactic acid changes. Aeta Nen- rochir 53:47-51, 1980

60. Teasdale G, Jennett B: Assessment of coma and impaired consciousness. A practical scale. Lancet 2:81-84, 1974

61. Valenca LM, Shannon DC, Kazemi H: Clearance of lactate from the cerebrospinal fluid. Neurology 21: 615-620, 1971

62. Wagner KR, Tornheim PA, Eichhold MK: Acute changes in regional cerebral metabolite values following experimental blunt head trauma. J Neurosurg 63:88-96, 1985

63. Ward JD, Becker DP: Head trauma, in Shoemaker WC, Thompson WL, Holbrook PR (eds): Textbook of Critical Care. Philadelphia: WB Saunders, 1984, pp 926-928

64. Weyne J, Leusen I: Lactate in CSF in relation to brain and blood, in Cserr HF, Fenstermacher JD, Fencl V (eds): Fluid Environment of the Brain. New York: Academic Press, 1975, pp 255-276

65. Yang MS, DeWitt DS, Becker DP, et al: Regional brain metabolite levels following mild experimental head injury in the cat. J Neurosurg 63:617-621, 1985

66. Yoshino E, Yamaki T, Higuchi T, et al: Acute brain edema in fatal head injury: analysis by dynamic CT scanning. J Nenrosurg 63:830-839, 1985

67. Zupping R: Cerebral acid-base and gas metabolism in brain injury. J Neurosnrg 33:498-505, 1970

Manuscript received October 1, 1985. Accepted in final form May 27, 1986. This work was supported by National Institutes of Health

Grant NS-12587. Dr. DeSalles was supported in part by the Conselho Na-

cional de Desenvolvimento Cientifico e Tecnolrgico, Brazil. This paper was presented at the Annual Meeting of the

American Association of Neurological Surgeons in Atlanta, Georgia, April 21-25, 1985.

Address reprint requests to: Antonio A. F. DeSalles, M.D., Massachusetts General Hospital, 15 Parkman Street - ACC 324, Boston, Massachusetts 02114.


prognostic significance of ventricular csf lactic acidosis

Documents