Get Permission Asuti, Jagadish, and Narayan: Estimation of CSF lactate as a diagnostic marker to differentiate pyogenic meningitis from nonpyogenic meningitis


Introduction

Inflammation of the membranes of meninges characterizes Meningitis - clinical syndrome that encloses the brain and spinal cord. These layers of meninges consist of the following: 1. Dura mater - A tough and outermost membrane 2. Arachnoid mater- A web-like middle membrane 3. Pia mater- the innermost layer which adheres closely to the brain. Subarachnoid space lies between arachnoid and pia mater and contains many of the blood vessels that feed the brain and spinal cord.1 Meningitis can be caused by various etiologies like tubercular, pyogenic and aseptic/viral. Meningitis is a major global health issue, particularly in developed countries. Among various types of meningitis, pyogenic/bacterial meningitis is one of the common infectious diseases of the CNS in India. Meningitis a critical disease associated with notable morbidity and mortality.2 Moreover, long-term sequelae such as palsies, hearing loss and personality changes influence approximately 40% of survivors.3 The worldwide distribution of Pneumococcal, Haemophilus, and meningococcal meningitis has been observed and mainly in young age groups. Incidence of meningitis has substantially reduced in developed countries after the introduction of vaccines against these agents. But incidence remains the same in the developing world. For identifying the cause of meningitis, clinical features, routine CSF parameters, and radiological findings are often inadequate. In developing countries, Gram’s stain and AFB stain of CSF are the most common rapid methods of detection of the organism, however, these methods lack sensitivity. Another common method used for diagnosis is the culture of CSF but it is time-consuming. Although PCR and nowadays Real-time-PCR are extremely sensitive and specific tests for diagnosis, however being expensive and less available these methods cannot be used for the routine purpose. Therefore, a robust cost-effective test should be available for differentiation among various types of meningitis. Because of all these limitations, the determination of the CSF lactate level may be a diagnostic and invaluable marker in differentiating pyogenic/bacterial from nonpyogenic meningitis.4 The detection of a high level of CSF lactate has shown promising results in the diagnosis of bacterial/pyogenic meningitis. Validation of CSF lactate in clinical studies would complement clinical diagnosis and enhance the accuracy of diagnosis.

We have undertaken this study with the main purpose of evaluating the efficacy of CSF lactate in differentiating pyogenic/ bacterial from nonpyogenic meningitis.

Materials and Methods

Patients and setting

A hospital-based cross-sectional study of all clinically suspected cases (40 cases) of meningitis of different age groups (18 years or more) admitted to M.S.Ramaiah hospitals from October 2011 to October 2013 was done. Conditions which can be responsible for the elevation of CSF lactate such as brain hypoxia, recent stroke, brain trauma, subarachnoid hemorrhage and seizures, Patients with fungal meningitis and HIV and Immunosuppressive therapy were excluded in the investigation. The sample size was estimated based on hypothesis testing for two means and evidence obtained from a study done by Ali Hassan Abro et al on CSF lactate level as a diagnostic tool to differentiate acute bacterial from viral meningitis, at Infectious disease dept., Rashid Hospital Dubai.5

Considering the SD in 1st group as 6.1, 2nd group as 0.6, mean difference as 3, effect size 0.8955, alpha error 5% and power of test at 80% for a 2 sided test, the estimated sample size was 33, which was rounded off to 40 subjects. Data was aggregated by pretested semi-structured questionnaire, clinical-diagnosis, and investigations. The diagnosis of meningitis was made on the basis of clinical symptoms and signs like headache, nausea, vomiting, fever, neck rigidity, presence of kernig’s sign and or brudzinski’s sign, altered sensorium, any focal neurological inadequacy, cranial nerves palsies, hemiparesis, seizures and or signs of cerebral dysfunction ranging from confusion, delirium, diminishing level of sensorium from lethargy to coma. Lumbar puncture was conducted on the patients instantly on admission. The assay used in the study was based on the enzymatic method by using Lactate Electrode using ABL 555 Blood gas analyzer (Radiometer Copenhagen, Denmark).

CSF analysis

CSF lactate analysis was performed for all cases with abnormal CSF cytochemical pictures. All routine and other relevant investigations were recorded from case files. The types of meningitis were categorized on the following basis: Pyogenic meningitis: CSF exhibiting neutrophilic pleocytosis (10-10000cells/mm3), elevated protein (>45mg/dl), low sugars (<40 mg/dl) with or without bacteria displayed on Gram stain/culture. Nonpyogenic meningitis: a. Tubercular meningitis: CSF exhibiting lymphocytic pleocytosis (10-1000cells/mm3), elevated protein (>45mg/dl), low sugars (<2/3rd blood sugar) with high CSF ADA levels (at a higher level to the ref range of lab) with or without bacteria demonstrated on ZN staining/culture/PCR. b. Viral meningitis: CSF exhibiting lymphocytic pleocytosis (25-500cells/mm3), normal/slightly elevated protein (20-80mg/dl), normal sugars with absent/normal CSF ADA levels and negative for bacteria on microscopy/cultures/PCR. The normal reference range considered for study was 1.2 to 2.1 mmol/L as standardized by the University of Colorado hospital laboratories approved by the college of American pathologists (updated on June 1, 2011).

Statistical analysis

Statistical analyses of the data were conducted using the software IBM SPSS 17.0 version. The frequency distribution (mean), Independent t-test, S.D. with 95% confidence limit and correlation statistics were the statistical parameters utilized to analyze the data.

Results

Demographic data

Out of 40 cases of meningitis considered for investigation, 16 were pyogenic and 24 non-pyogenic (9 tubercular and 15 viral)Figure 1. In our study, almost equal sex distribution was undertaken with 52.5% being males and 47.5% females Figure 2. shows the type of the age group present in the study; most cases were of 18-29 (32.5%) and 40-49(25%) age groups.

Figure 1

Frequency of different groups of meningitis

https://s3-us-west-2.amazonaws.com/typeset-media-server/f2d52a44-3499-4126-a8fa-356298ee3752image1.png

Figure 2

Sex wise distribution of cases

https://s3-us-west-2.amazonaws.com/typeset-media-server/f2d52a44-3499-4126-a8fa-356298ee3752image2.png

Figure 3

Age wise distribution of cases

https://s3-us-west-2.amazonaws.com/typeset-media-server/f2d52a44-3499-4126-a8fa-356298ee3752image3.png

Clinical presentation

In the present investigation, fever (80%) was most common symptom followed by headache (77.5%), altered sensorium (60%) and vomiting (40%) Figure 4 . Signs of meningeal irritation were present in 72.5% of cases. Abducent nerve palsy was seen in 2 cases (5%).

Figure 4

Analysis of clinical symptoms and signs

https://s3-us-west-2.amazonaws.com/typeset-media-server/f2d52a44-3499-4126-a8fa-356298ee3752image4.png

The mean CSF total cell count was higher (with neutrophil predominance) in pyogenic meningitis group compared to nonpyogenic meningitis group in the current study.Table 1

Table 1
CSF parameters N Mean SD Min. Max. ‘F’ value ‘p’ value
TC(cells/mm3) Tubercular 9 103.00 89.01 10 264 3.0 0.058
Viral 15 133.47 119.94 27 450
Pyogenic 16 727.06 1181.10 28 3690
PMN Tubercular 9 29.11 31.970 0 90 22.9 <0.001
Viral 15 27.80 20.67 2 70
Pyogenic 16 78.00 18.18 25 98
Lymphocyte Tubercular 9 70.89 31.97 10 100 19.5 <0.001
Viral 15 68.87 23.20 30 98
Pyogenic 16 22.00 18.18 2 75

Analysis of CSF cytological parameters in different group of meningitis

Our study reveals that the mean CSF sugar was low in pyogenic and tubercular meningitis compared to viral meningitis.CSF sugar to blood sugar ratio was <0.4 in both pyogenic meningitis and tubercular meningitis and >0.6 in viral meningitis. Mean CSF protein was highest in pyogenic meningitis (267.78 mg/dl) followed by tubercular (184.99 mg/dl) and viral (96.01mg/dl). Mean CSF lactate was high in both pyogenic and tubercular meningitis compared to viral meningitis. However mean CSF lactate was very high (11.66 mmol /L) in pyogenic group compared to tubercular (4.0 mmol/L)Table 2 .

Table 2
CSF parameters N Mean SD Min. Max. ‘F’ value ‘p’ value
Sugar(mg/dl) Tubercular 9 44.31 16.33 23.4 69.0 6.24 0.005
Viral 15 67.72 16.65 45.7 112.0
Pyogenic 16 41.31 28.22 8.4 123.0
RBS(mg/dl) Tubercular 9 131.8 33.79 84 180 1.66 0.203
Viral 15 114.6 30.72 86 201
Pyogenic 16 172.6 136.4 73 560
CSF sugar/blood sugar Ratio Tubercular 9 .3278 .0719 .20 .40 49.1 <0.001
Viral 15 .6066 .1144 .44 .86
Pyogenic 16 .2524 .1047 .08 .40
Protein(mg/dl) Tubercular 9 184.9 209.9 48 723 3.43 0.043
Viral 15 96.01 42.93 66 206
Pyogenic 16 266.7 236.1 68 895
Lactate(mmol/L) Tubercular 9 4.000 1.280 2.0 6.3 80.9 <0.001
Viral 15 1.993 .6923 1.0 3.7
Pyogenic 16 11.66 3.264 2.9 15.6

Analysis of CSF biochemical parameters in different group of meningitis

We found that the mean total leucocyte count (TLC) and mean absolute neutrophil count (ANC) was high in pyogenic group compared to nonpyogenic group. Mean serum procalcitonin (PCT) was high in pyogenic group and normal in nonpyogenic group Table 3 .

Table 3
Blood parameters N Mean SD Min. Max. ‘F’ value ‘p’ value
TLC(cells/mm3) Tubercular 9 8728.89 2216.4 5000 13060 10.33 <0.001
Viral 15 7816.00 2521.7 3790 11870
Pyogenic 16 13939.3 5504.9 6600 23670
ANC(cells/mm3) Tubercular 9 5933.89 2348.6 1950 10186 8.132 0.001
Viral 15 5671.20 2042.9 2609 9027
Pyogenic 16 10431.2 4986.0 3690 17924
ESR Tubercular 9 24.11 15.775 8 48 0.450 0.641
Viral 15 19.67 21.744 6 81
Pyogenic 16 27.75 28.539 5 109
Ser. PCT(ng/dl) Tubercular 9 .203 .1714 .02 .5 4.043 0.026
Viral 15 .240 .2828 .04 1.1
Pyogenic 16 4.813 7.8437 .20 27.2

Analysis of various blood parameters in different group of meningitis

Mean CSF lactate was high in both pyogenic meningitis and tubercular meningitis compared to viral meningitisTable 4,Table 5,Table 6 .

Table 4
Lactate Total 'F' 2 value ‘p’ value
<=2.1 >2.1
Tubercular 1 8 9 28.474 <0.001
Viral 13 2 15
Pyogenic 0 16 16
Total 14 26 40
35.0% 65.0% 100.0%

Analysis of CSF lactate(mmol/L) in different group of meningitis

Table 5
CSF Lactate Total P value
<=2.1 >2.1
Viral 13 2 15 <.001
Pyogenic 0 16 16
Total 13 18 31
41.9% 58.1% 100.0%

Comparison of CSF lactate between viral meningitis and pyogenic meningitis.

Table 6
CSF Lactate Total P value
<=2.1 >2.1
Tubercular 1 8 9 <.001
Viral 13 2 15
Total 14 10 24
58.3% 41.7% 100.0%

Comparison of CSF lactate between viral meningitis and tubercular meningitis

Using a higher cut off value of 5 mmol/L, CSF lactate was useful in differentiating pyogenic meningitis from tubercular meningitis with statistical significance Table 7 .

Table 7
CSF Lactate Total P value
<=5.0 >5.0
Tubercular 8 1 9 <.001
Pyogenic 1 15 16
Total 9 16 25

Comparison of CSF lactate between pyogenic meningitis and tubercular meningitis using a cut off value of 5 mmol/L.

The data obtained in the present study revealed that CSF lactate was directly correlating with serum procalcitonin level in pyogenic meningitis group Table 8 .

Table 8
Group CSF lactate Ser. PCT Total
<=0.5 >0.5
Pyogenic <2.1 0 0 0
>2.1 4 12 16
25% 75% 100.0%

CSF lactate in relation to serum procalcitonin (PCT) in pyogenic meningitis

Our study reveals that CSF lactate level was directly correlating with CSF sugar to blood sugar ratio in pyogenic meningitis group Table 9 .

Table 9
Group Ratio Lactate Total
<=2.1 >2.1
Pyogenic <=0.4 16 16
100.0% 100.0%
Total 16 16
100.0% 100.0%

CSF lactate in relation to CSF glucose/blood glucose ratio pyogenic meningitis

Table 10
Group Outcome Total
Improved Expired
Tubercular 7 2 9
77.8% 22.2% 100.0%
Viral 14 1 15
93.3% 6.7% 100.0%
Pyogenic 13 3 16
81.3% 18.8% 100.0%
Total 34 6 40
85.0% 15.0% 100.0%

Analysis of prognosis in different groups of meningitis

Discussion

In the case of acute meningitis, the accurate initial diagnosis is the cornerstone in therapeutic decision making. Pyogenic meningitis is associated with significant morbidity and mortality and is a common and serious life-threatening disease.6 To differentiate between bacterial and viral meningitis, clinical signs have been manifested to have sensitivity and specificity within a range and are of less significance for differentiating among the types of meningitis6. In only 60-80% cases, direct CSF examination on an emergency basis provides evidence of bacterial meningitis.7 In our study, only 2 out of 16 cases of pyogenic meningitis had positive gram stain and only one case was culture-positive which clearly shows that the CSF cytochemical parameters play a key role in the diagnosis of the type of meningitis. In our study, there was almost equal sex distribution with 52.5% being males and 47.5% being females while in Ali Hassan Abro et al study, the mean age of the patients was 33.73 ± 11.7 years where Males (86.5%) outnumbered the females (13.43%) in the study.5   In Viallon et al study mean age group was 55 ± 20 years and 35 ± 18 years in bacterial meningitis and viral meningitis respectively. In our study,8 the most common symptom of meningitis was fever (80%), followed by headache (77.5%), altered sensorium (60%) and vomiting (40%). Signs of meningeal irritation were present in 72.5% of cases. This suggests that the classical clinical triad of fever, headache and neck stiffness is not consistently seen in all the cases. Neurologic deficits in the form of abducent nerve palsy were seen in 5% of cases. However, the incidence of neuro deficits may be underestimated in this study as major deficits like hemiplegia constituted exclusion criteria. In Khatua et al study, neck rigidity was noted in 54 % of cases and kernig’s sign in 40% of cases.9  In studies done by Van de Beek et al and Virmani et al cranial nerve palsy were seen in 33% and 27% cases respectively.10,11 The present study revealed that the mean CSF total cell count (TC) was higher in the pyogenic group compared to the non-pyogenic group i.e, tubercular and viral meningitis. Pyogenic meningitis showed neutrophil (PMN) predominance where a non-pyogenic group showed lymphocyte predominance. Baker RC et al and Viallon et al have found similar findings in their study with high CSF pleocytosis with neutrophil predominance in bacterial meningitis compared to viral meningitis.8,12    However, Negrini et al have reported that most of the patients with aseptic meningitis had neutrophil predominant pleocytosis in CSF.13 In our study CSF differential leucocyte count helped differentiate pyogenic from nonpyogenic meningitis with statistical significance. In the present findings, mean CSF sugar was low in pyogenic (41.31mg/dl) and tubercular meningitis (44.31mg/dl) compared to viral meningitis (67.72mg/dl). A study done by Ali Hassan Abro et al reported similar findings with mean CSF sugar 26.5 ± 21.6 and 67 ± 18.96 mg/dl in bacterial and viral meningitis respectively8. CSF sugar to blood sugar ratio was <0.4 in both pyogenic meningitis and tubercular meningitis and >0.6 in viral meningitis. Spanos et al and Viallon et al also noted similar findings in their study.8,14 Because CSF sugar varies with blood sugar, the ratio of CSF sugar to blood sugar is a better parameter.

Mean CSF protein was found to be highest in pyogenic meningitis (267.78 mg/dl) followed by tubercular meningitis (184.99 mg/dl) and viral meningitis (96.01mg/dl). Similar findings noted in studies done by Viallon et al with mean CSF protein 4.9±4.6 g/L in bacterial meningitis versus 1±0.6 g/L in viral meningitis9. Ray P et al also noted similar findings in their study.15 Robert L et al in their study could not find CSF lactate a better marker compared to other biochemical parameters like sugar, protein, cell count and typing.16 In our study mean CSF lactate was high in both pyogenic and tubercular meningitis compared to viral meningitis with statistical significance (p <.001). However, mean CSF lactate was very high (11.66 mmol /L) in pyogenic meningitis compared to tubercular (4.0 mmol /L). In a study done by Ali Hassan Abro et al, mean CSF lactate was 14.96±6.13 mmol /L and 2.38±0.59 mmol /L in bacterial meningitis and viral meningitis respectively.  Smith et al and Genton B et al also have noted similar findings in their study.17,18 Viallon et al also noted similar findings with mean CSF lactate of 9±5 mmol /L in bacterial meningitis and 2.6±1.6 mmol /L in viral meningitis and this study was done in patients with negative direct examination of CSF for bacteria.8 In the present study, CSF lactate was high in all cases of pyogenic meningitis. The further evaluation suggests that there is no correlation between mean CSF lactate value and gram stain/culture results. Donald P et al have reported that CSF lactate does not hold marked advantages over conventional chemical analysis of CSF in differentiating Tubercular meningitis from aseptic meningitis.19 But in our study mean CSF lactate was able to differentiate Tubercular meningitis (4 mmol /L) from viral/aseptic meningitis (1.99 mmol /L) (p <0.001). Tang LM et al also noted higher CSF lactate in tubercular meningitis (3.9 mmol/L).20 In the present investigation using a higher cut off value of 5 mmol /L, CSF lactate helped in differentiating pyogenic meningitis from tubercular meningitis (p-value <0.001). In the present study, mean blood total leucocyte count (TLC) and mean absolute neutrophil count (ANC) were high in the pyogenic meningitis group compared to the non-pyogenic meningitis group. Ali Hassan Abro et al noted similar findings in their study. Also, in our study, mean serum procalcitonin (PCT) was high (4.83ng/ml) in the pyogenic meningitis group compared to the non-pyogenic meningitis group. In the study reported by Schwartz et al 10 among 16 patients with bacterial meningitis had serum PCT >0.5ng/ml. In our study 13 out of 16 patients with pyogenic meningitis had serum PCT >0.5. Jereb et al and viallon et al also noted similar findings in their study.8,21 The mean CSF lactate level (11.66 mmol /L) value obtained in our investigation directly correlated with mean serum procalcitonin level (4.83ng/ml) in the pyogenic meningitis group. All 16 bacterial meningitis patients had high CSF lactate levels and 13 out 16 patients had high serum PCT level which is supported by a similar correlation noted in Viallon et al study.8 Also, our study reveals the direct correlation of the mean CSF lactate level (11.66mmol/L) value with mean CSF sugar to blood sugar ratio (0.25) in the pyogenic meningitis group. A similar correlation was noted by Viallon et al in their study9. CSF lactate was found to be higher in those expired compared to those who survived in both pyogenic meningitis and tubercular meningitis group. Similar findings were noted by Ali Hassan Abro et al in pyogenic meningitis group in their study8. Thus the magnitude of CSF lactate level may help in predicting poor outcome and determine likely prognosis.

Conclusion

CSF lactate level is a rapid, relatively economical and modest procedure that can be of great value as a diagnostic marker in the early differentiation of pyogenic meningitis and tubercular meningitis from viral meningitis, helping in the early institution of appropriate treatment and decreasing mortality and complications.

In our study, the CSF lactate level was significantly higher in pyogenic meningitis and tubercular meningitis group compared to viral meningitis. With a cut off value of 2.1 mmol/L, CSF lactate value was useful in differentiating both pyogenic and tubercular meningitis from viral/aseptic meningitis. Also, with a cut off value of 5.0 mmol/L, CSF lactate was useful in differentiating pyogenic meningitis from tubercular meningitis. CSF lactate value directly correlates with serum procalcitonin level and CSF sugar/blood sugar ratio.

Acknowledgement

The authors thank the Professor and Head Department of General Medicine for their kind support. The authors are also grateful to authors/editors/ publishers of all those articles, journals and books from where the literature for this article has been reviewed and discussed. The authors also thank all the study subjects for their kind support.

Source of Funding

None.

Conflict of Interest

None.

References

1 

L Ginsberg D Kidd Chronic and recurrent meningitisPract Neurol. Dec20088634861

2 

Roos KL Bacterial MeningitisCurr Treat Ootions Neurol19991147156

3 

A R Tunkel B J Hartman S L Kaplan B A Kaufman K L Roos R J Whitley Practise guidelines for the management of bacterial meningitisClin Infect Dis20043912671284

4 

A Viallon S Guyomarch F Zeni I Piteaud F Lucht M Levy Serum procalcitonin and cerebrospinal fluid lactate levels in emergency room differential diagnosis of acute meningitis with a negative Gram stainCrit Care 200316412

5 

8.Ali HA Ahmed SA Ali H Ustadi AM Hasab AAH Cerebrospinal fluid analysis, acute bacterial versus viral meningitisPak J Med Sci200824645650

6 

M Holub O Beran O Dzypova J Hnykoya Z Lacinova J Prihodova Cortisol levels in cerebrospinal fluid correlate with severity and bacterial origin of meningitisCrit Care2007114141

7 

A Harrell F E Durack D T . Differential diagnosis of acute meningitis: an analysis of the predictive value of initial observationsJAMA198926227002707

8 

Nicolas Desseigne Olivier Marjollet Albert Birynczyk Mathieu Belin Stephane Guyomarch Meningitis in adults with a negative direct cerebrospinal fluid examination: value of cytochemical markers for differential diagnosisCritical Care20111529

9 

Khatua SP. Bacterial meningitis in children: Analysis of 231 caseJ Indian Med Ass196137332332

10 

D De Beek J De Gans L Spanjaard M Weisfelt J B Reitsma M Vermeulen Clinical features and prognostic factors in adults with bacterial meningitisN Engl J Med2004352950950Erratum)

11 

V Virmani G Rangan G Shrinivas A study of the cerebrospinal fluid in atypical presentations of tuberculous meningitisJ Neurol Sci197526458792

12 

R C Baker A M Lenane The predictive value of cerebrospinal fluid differential cytology in meningitisPediatr Infect Dis J19898329359

13 

B Negrini K J Kelleher E R Wald Cerebrospinal fluid findings in aseptic versus bacterial meningitisPediatrics2000105316325

14 

A Spanos F E Harrell D T Durack Differential diagnosis of acute meningitis: an analysis of the predictive value of initial observationsJAMA198926227002707

15 

P Ray G Badarou-Acossi A Viallon D Boutoille M Arthaud D Trystram B Riou Accuracy of the cerebrospinal fluid results to differentiate bacterial from non bacterial meningitis, in case of negative gramstained smearAm J Emerg Med200725179184

16 

L Robert A M Margaret J M Thomas E V Haldane Evaluation of Cerebrospinal fluid lactate levels as an aid in differential diagnosis of bacterial and viral meningitisJ Clin Mic1980324331

17 

S M Smith R H Eng J M Campos H Chmel D-lactic acid measurements in the diagnosis of bacterial infectionJ Clin Microbiol198927385393

18 

B Genton J P Berger Cerebrospinal fluid lactate in 78 cases of adult meningitisIntensive Care Med199016196200

19 

P Donald C Malan Cerebrospinal fluid lactate and lactate dehydrogenase activity in the rapid diagnosis of bacterial meningitisS Afr Med J1986693942

20 

Tang LM Serial lactate determinations in tuberculous meningitisScand J Infect Dis19882018184

21 

M Jereb I Muzlovic S Hojker F Strle Predictive value of serum and cerebrospinal fluid procalcitonin levels for the diagnosis of bacterial meningitisInfection200129209212



jats-html.xsl


This is an Open Access (OA) journal, and articles are distributed under the terms of the Creative Commons Attribution-NonCommercial-ShareAlike 4.0 License, which allows others to remix, tweak, and build upon the work non-commercially, as long as appropriate credit is given and the new creations are licensed under the identical terms.

  • Article highlights
  • Article tables
  • Article images

View Article

PDF File   Full Text Article


Copyright permission

Get article permission for commercial use

Downlaod

PDF File   XML File  


Digital Object Identifier (DOI)

Article DOI

https://doi.org/10.18231/j.ijn.2019.015


Article Metrics






Article Access statistics

Viewed: 2157

PDF Downloaded: 603