ANNALS OF CLINICAL AND LABORATORY SCIENCE, Vol. 17, No. 6 Copyright 1987, Institute for Clinical Science, Inc. U se o f Cerebrospinal Fluid Lactic Acid Concentration in the Diagnosis of Fungal M eningitis BARBARA A. BODY, P h.d., RUTH H. ONESO N, M.D., M.P.H. and DAVID A. H ERO LD, M.D., P hd. Department o f Pathology, University o f Virginia Medical Center, Charlottesville, VA 22908 ABSTRACT A patient with a several year history of normal pressure hydrocephalus was found to have an infection owing to Cryptococcus neoformans. Cryptococcal infection was not suspected until typical cells were observed in a W right s stained sm ear of cerebrospinal fluid (CSF). A review of past medical findings in this patient showed elevated CSF values for lactic acid and protein. This case prom pted us to review the use of lactic acid as an indicator of fungal meningitis and compare it to other m ore commonly used nonspecific indicators of fungal m eningitis, notably th e concentrations of glucose and protein, and the num ber of leukocytes in CSF. In our institution, all 10 culturally proven cases of fungal meningitis, for which the lactic acid concentration in the CSF was available, were found to have an elevated lactic acid concentration (range 3.2 to 13.3 mmol per L vs normal range 0.8 to 2.8 mmol per L). No other nonspecific indicator was elevated in all 10 patients. In view of the poor sensitivity of stained sm ear or w et preparations and cultures, when less than five ml of CSF are used for culture, an elevated lactic acid value in a patient with or without signs of m eningitis should raise th e suspicion of fungal infection. Introduction A patient with meningitis caused by a fungus may present with acute m eningeal sym ptom s. M ore commonly, how ever, the infection occurs chronically w ith a p ro lo n g e d h isto ry of w eeks, months, or even years.4,15 The diagnosis of fungal m en in g itis is co m p lic a ted because th e p resen tatio n m ay b e nonspecific, and the sensitivity of fungal cultu re is low w h en less th an five m l of 429 cerebrospinal (CSF) are cultured.14 Laboratory values for CSF specim ens that are acknowledged to be consistent with fungal m en in g itis in clu d e a m o d est increase in leukocytes, a low glucose value, and elevated p rotein c o n te n t.5 These param eters are relatively nonspecific and becom e even more difficult to in terp ret in patients w ith com plicated u n d e rly in g m ed ic al p ro b le m s th e patient population that is most likely to succum b to fungal m eningitis. M any 0091-7370/87/1100-0429 $00.90 Institute for Clinical Science, Inc.
4 3 0 BODY, ONESON, AND HEROLD p a tie n ts w ith m en in g itis cau sed by C r y p to c o c c u s n e o fo r m a n s h av e an underlying condition that may alter their im m une response w ith suppression of CSF leukocytosis. H yperglycem ia and rapidly fluctuating serum glucose values make the CSF glucose value difficult to interpret. For patients with hyperglycemia, Powers19 recom m ended that a CSF glucose value below 40 mg per dl or a ratio of cerebrospinal fluid glucose to serum glucose below 0.31 should be used in the detection of bacterial m eningitis. This ratio has not been defined for patients with hyperglycem ia and fungal m eningitis. An increase in CSF protein concentration appears to be of lim ited value. T he C SF protein content gradually increases with age so that the reference range in the 40 to 80 year age group is 20 to 60 mg per dl. As a consequence, a decision point of 100 mg per dl has been suggested by some investigators as the value to be used to raise suspicion of a septic process. In addition, a traum atic lum bar puncture could dramatically alter the value for the CSF protein because serum contains 250 times more protein than CSF. Although corrections based on the num ber of RBCs can be performed, these values are frequently in error.11 M arkedly increased CSF lactic acid values have been associated with acute bacterial m eningitis and other causes of b ra in hypoxia su ch as r e c e n t h e a d trau m a (w ithin 30 to 60 m inutes), seizure activity, or neurosurgical p roced u r e s.1,13,9 T h e re have b e e n a few reports of m oderately elevated lactic acid values in the CSF of patients with fungal m eningitis, tubercular m eningitis, and partially tre a te d bacterial m eningitis, but it has not been used as frequently as other param eters.3,8,12 A recent case of meningitis at our institution owing to C. neoformans is reported. The patient had an increased lactic acid concentration in C S F th a t m ig h t hav e b e e n u se d to increase the suspicion for the inclusion of fungal infection in the differential diagnosis. To d eterm in e the potential diagnostic reliability of an increased CSF lactic acid concentration, charts of patients with fungal meningitis at our institution d u r in g th e la s t th r e e y e a rs w e re reviewed. Lactic acid levels in CSF were available for 10 of 11 patients w ith culturally proven meningitis, and all 10 had elevated lactic acid in their CSF. The data support the suggestion that a m oderately elevated lactic acid level might be used to increase the suspicion for the diagnosis of fungal m eningitis and prom ote requests for more specific diagnostic tests, such as cryptococcal antigen latex agglutination test and the culture of m ultiple specim ens of C SF of appropriately large volume (>5 ml). Case Report A 60-year-old insulin dependent diabetic woman was first admitted to the University of Virginia Hospital in Septem ber 1981 for evaluation of an eightmonth history of memory loss, confusion, and gait unsteadiness. Physical examination was remarkable for disorientation, to tim e and place, and a wide shuffling gait. Laboratory values for CSF are listed in table I (patient #1); cultures for bacteria and fungi w ere negative, and a cryptococcal antigen latex agglutination test was not requested. Computerized axial tomography scan revealed hydrocephaly, and a ventriculoperitoneal shunt was placed for presum ed norm al pressure hydrocephalus. H er sym ptom s rem itted. In October 1981, the patient experienced a left-sided Jacksonian seizure; therapy with diphenylhydantoin was started with resolution of seizures. In April 1982, she showed decreased memory, with increases in confusion and gait unsteadiness. H er shunt was placed and symptoms resolved. In January, 1983, the patient underwent another revision of her shunt to relieve recurrent symptoms. Then in August 1983, at the time of the third revision of her shunt, cells presumed to be Cryptococcus neoformans were detected during hematologic analysis of CSF. A latex agglutination test for cryptococcal antigen was not ordered on the original specimen and could not be done owing to insufficient quantity of specimen, but the culture of the specimen grew C. neoformans after two days. A specimen of CSF collected later was positive for cryptococcal antigen at a titer of 1:4.
Cerebrospinal Fluid Profiles of Patients with Fungal Meningitis 4 1/84 49 116 251 46 4.2 2 + + N A C. neoformans Kidney transplant, steroids Deceased 5 6/84 N A 60 133 45 4.8 45 - + N A B. derm atiliiis Prior CNS Blastomycosis Alive 9 2/85 42 49 N A - 3.5 42 - + N A C. albicans Lymphoma, bone marrow transplant Deceased 10 3/85 708 62 107 58 9.2 1600T. + NA C. albicans Astrocytoma, Alive 3/85 1728 18 90 20 10.6 442T - + NA neurosurgical procedure LACTIC ACID IN FUNGAL M EN IN G ITIS TABLE I Protein Lactic acid Patient Date 20-40 mg/dl C S F Glucose 60-80mg/dl Serum Glucose 0.8-2.8 65-llSmg/dl % mmoi/l W B C Smear Culture Serology Organism Underlying Condition Follow-up 1 9/81 543 81 192 42 N A 67 - - N A C. neoformans Insulin dependent diabetes Alive 6/82 61 106 299 35 3.2 3 - - N A 9/83 58 57 181 31 6.3 7 + + 1:4 2 9/82 66 243 458 53 3.6 1 - - 1: 1 C. neoformans AIDS, steroids Deceased 9/82 89 276 371 74 5.0 N A - - N A 3 11/82 284 24 174 14 N A 142 + + 1:2048 C. neoformans Alcoholism Deceased 12/82 98 98 233 42 5.0 225 + + 1: 512 12/82 48 117 294 40 4.3 176 + + N A 6 8/84 495 24 140 17 10.7 N A + + 1:4096 C. neoformans Myasthenia gravis, steroids Deceased 8/84 508 84 316 27 13.3 54 + + 1:4096 9/84 324 57 324 18 10.8 105 - - 1:4096 7 11/84 74 45 N A - N A 42 + + 1:8192 C. neoformans Non-Hodgkin's lymphoma Alive 12/84 260 24 125 19 8.3 96 + + 1:8192 12/84 45 40 95 42 2.7 6 + + 1:8192 8 1/85 95 47 128 37 4.7 54 - + 1: 256 C. neoformans Alcoholism, seizure disorder Deceased 1/85 119 162 153 41 4.0 68 - + 1: 16 1/85 69 62 192 32 2.9 34 - + 1: 512 2/85 60 139 240 58 2.9 82 + + 1: 128 NA=Not available as the test was either not performed or the result was not in chart. T=Traumatic lumbar pucnture indicated by presence of RB C in cerebrospinal fluid.
4 3 2 BODY, ONESON, AND HEROLD M ethods T he records of th e Mycology Laboratory of the University of Virginia Medical C enter w ere review ed to obtain a list of patients with cultures of CSF positive for clinically significant fungi. The charts of these patients w ere reviewed to obtain a brief m edical history and data from various laboratory tests. The CSF specimens were sent to the Chem istry Laboratory for determ ination of p ro te in, glucose, and lactic acid, which w ere perform ed as a battery of tests*; glucose determ inations on serum w ere also perform ed. *f Total cell analysis and differential of CSF w ere p e r form ed in th e Clinical M icroscopy section of the Hematology Laboratory. One ml of CSF was subjected to cytocentrifugation; the slide was fixed and stained w ith W right s stain. M icrobiologie analysis of CSF included Gram stain and routine bacteriologic cultures. Indir ink preparations, cultures for fungi, and the latex agglutination test for detection of cryptococcal antigen w ere perform ed only on request. Results T here w ere 11 patients with cultureproven fungal m eningitis; one or m ore CSF lactic acid values were available for 10 of these patients. In table I are listed p ertinent laboratory values and clinical history data for each patient. The organism s resp o n sib le for th ese infections in c lu d e d seven C. n eo fo rm a n s, tw o Candida albicans, and one Blastomyces dermatitidis. Initial India ink or stained sm ear preparations were negative for six patients, which represented three cases ow ing to C. n e o fo rm a n s, and th o se infections owing to B. derm atitidis and * DuPont ACA III analyzer, t SMAC analyzer. C. albicans. In m osteases, several specim ens of CSF w ere obtained. Specimens of C S F co llected after a fungus was found to b e grow ing from th e initial specim en w ere often positive by either sm ear or India ink preparation. It was u n c le a r w h e t h e r th is w as d u e to in c re a s e d su sp ic io n a n d p ro lo n g e d review of slides by laboratory technologists, the collection of larger amounts of C SF, or progression of infection. Lactic acid was the only param eter th a t was e le v a te d in th e C S F of all patients. T here was usually a m oderate protein elevation (60 to 100 mg per dl), w ith a few m arkedly e le v ated values (>100 mg per dl). A consistent leukocytosis was found in seven patients, but this was absent in three. Low values for glucose w ere rarely observed in the CSF of this group of patients, most of whom had elevated serum glucose levels owing to steroid therapy or diabetes. However, the ratios of the glucose level in CSF specim ens to glucose level in a randomly obtained serum specim ens was usually above 31 percent. Discussion D ecreased glucose and elevated pro tein and leukocyte count in cerebrospinal fluid are commonly used nonspecific indicators of m eningitis.5 In the case of our index p a tie n t, th ese p aram eters w ere not consistently elevated. A diagnosis of normal pressure hydrocephalus was consistent with many of the findings for this patient; however, Mangham et a l15 have d e scrib e d a serie s of four patients with cryptococcal infection who w ere also incorrectly diagnosed as having norm al p ressu re hydrocephalus. T hese authors stressed th e n eed for infection to be ruled out prior to placem ent of a shunt. The objective of this study was to review the levels of lactic acid in cases of fungal m eningitis and compare the usefulness of this variable to
LACTIC ACID IN FUNGAL MENINGITIS 4 3 3 other, m ore w idely appreciated, nonspecific indicators of fungal m eningitis. It was found by the p resent authors that only lactic acid concentration was e le vated in all patients. A m oderate to large increase in p ro tein concentration and leukocytosis w ere usually, b u t not u niformly, present. A lthough it has b een known for 60 years that CSF lactic acid concentration is increased in m eningitis,10 its use has b een lim ited in th e past because the early m ethods that w ere used for the m easurem ent of lactic acid were not suitable for analysis of single specimens on a STAT basis.1,6,12,13 M ore recently, the d e v e lo p m en t of analytical enzym atic m ethods and th e in troduction of autom ated in stru m e n ts th at only req u ire small volumes of CSF for m easurem ent of lactic acid have overcom e this p roble m.13,16 T he d iag n o stic value of an increased lactic acid value as an indicator of bacterial m eningitis has been p re viously rep o rted, and th ere have also been several reports of increased CSF lactic acid in fungal m eningitis.1,8,12,13 The glucose concentration in cereb ro spinal flu id of th ese patients was not d e c re a s e d, as has b e e n c la ssic a lly described in bacterial meningitis, but, in fact, was elevated above normal values in m ost p a tie n ts. T he elevation was, in m ost cases, a re fle c tio n of e le v a te d serum glucose concentration owing to either diabetes or steroid therapy. These o b se rv a tio n s are sim ilar to th o se of Butler et al, who reported only 55 percent of patients with cryptococcal m eningitis had CSF glucose values of less than 40 m g p e r 100 m l.12 A lthough such u n d erly in g factors can be taken into account by determ ining the ratio of glucose in cerebrospinal fluid and serum, a normal or mildly elevated glucose value in the cerebrospinal fluid may not always provoke exam ination of this ratio. F u r therm ore, even this ratio can be an inaccurate reflection of C S F glucose concentration because of the complex rela tio n sh ip b e tw e en the serum and CSF concentrations of glucose. Glucose en ters th e C S F by at least two m echanism s. The active transport of glucose across th e cerebral endothelium by facilitated diffusion is red u ced w ith increasing seru m glucose co n cen tratio n s by this mechanism, a serum value of 300 mg p er dl would yield a CSF concentration of about 200 mg per dl. Passive diffusion of glucose into the CSF also occurs and is a function of both serum concentration and tim e. Following a change in serum glucose, there is a one to two hour delay in reaching equilibrium in the CSF.19 In chronically hyperglycem ic individuals, this delay is postulated to be longer.7 In contrast to CSF glucose, CSF lactic acid is in dependent of the serum concentrations of lactic acid and, therefore this value can be evaluated w ithout reference to other laboratory inform ation.18 Lactic acid is known to be elevated in association w ith o th er conditions th at cause brain hypoxia, including recen t head traum a (within 30 to 60 minutes), hem orrhage, seizure activity, stroke, and n e u ro su rg ic a l p ro c e d u re s. In many instances, knowledge of patient history w ithout reference to o th er laboratory inform ation will allow the physician to determ ine w hether or not the lactic acid is elevated owing to one of these factors. In addition, other nonspecific indicators of m eningitis will also be altered in most patients. T he latex agglutination test for cryptococcal infection is considered to be a sen sitiv e and specific test; how ever, M cginnis found it to be positive on the first C S F specim en from lum bar puncture in only 58 percent of cases,17 and it can not d e te c t o th er causes of fungal m eningitis. In addition, it is im portant to note that it is not cost effective to run this te s t in d iscrim in an tly on all C SF specim ens owing to high reag en t and labor costs. In contrast, th e lactic acid
4 3 4 BODY, ONESON, AND HEROLD determ ination is not a labor intensive test, and reagent costs are m inim al thus m aking it a reasonable, albeit nonspecific, test to perform. Although our data are lim ited to ten p atien ts, they are consistent w ith the observations of others and support the conclusion that determ ination of lactic acid in the cerebrospinal fluid is useful as an indicator of fungal meningitis and may be advantageous when compared to other nonspecific indicators. A m oderately elevated lactic acid concentration in cerebrospinal fluid should be used as an indication that fungal m eningitis must b e c o n sid ered in th e differential diagnosis. If cultures are negative, this finding should prom ote the ordering of more specific diagnostic tests including the latex agglutination test for cryptococcal antigen and c u ltu re of ap p ro p riately large volumes of cerebrospinal fluid (>5 ml) for all fungi. Acknowledgments Thanks are extended to Drs. W. Michael Scheld and Michael R. Wills for their review of the manuscript and helpful comments. References 1. B r o o k, I., B r i c k n e l l, K. S., O v e r t u r f, G. D., and F i n e g o l d, S. M.: M easurement of lactic acid in the cerebrospinal fluid of patients with infections of the central nervous system. J. Infect. Dis. 157:384-390, 1981. 2. B u t l e r, W. T., A l l i n g, D. W., S p ic k a r d, A., and U t z, J. P. : Diagnostic and prognostic value of clinical and laboratory findings in cryptococcal m eningitis. New Engl. J. M ed. 270:59-67, 1964. 3. D o n a l d, P. R. a n d M a l a n, C.: Cerebrospinal fluid lactate and lactate dehydrogenase levels as diagnostic aids in tuberculous meningitis. South African Med. J. 67:19-20, 1985. 4. E l l n e r, J. J. and B e n n e t t, J. E. : Chronic meningitis. Medicine 55:341-369, 1976. 5. G a r d n e r, P. and P r o v in e, H. T.: Manual of A cute B acterial Infections. B oston, L ittle Brown, 1975, pp. 141-143. 6. G a s t r in, B., B r e i m, H., and R o m b o, L.: Rapid diagnosis of meningitis with use of selected clinical data and gas-liquid chromatographic determination of lactate concentration in cerebrospinal fluid. J. Infect. Dis. 139:529-533, 1979. 7. G j e d d e, A. and C r o n e, C.: Blood-brain glucose transfer in chronic hyperglycem ia. Science 214:456-457, 1981. 8. G o u ld, I. M., Irw in, W. J., and W adhw ani, R. R.: The use of cerebrospinal fluid lactate determ ination in the diagnosis of meningitis. Scand. J. Infect. Dis. 12:185-188, 1980. 9. H e r o l d, D. A., S avory, J., and B r u n s, D. E.: Lactic acid in cerebrospinal fluid: Evaluation and application of an automated enzymatic assay. Ann. Clin. Lab. Sci. 11:416-421, 1981. 10. K i l l ia n, J. A. and N i s h im u r a, K.: Lactic acid of normal and pathological spinal fluid samples. Proc. Soc. Exp. Biol. Med. 23:255-257, 1925. 11. K j e l d s b e r g, C. R. and K r i e g, A. F.: Cerebrospinal fluid and o th er body fluids. Clinical D iagnosis and M anagem ent by L aboratory Methods (Todd, Sanford, Davidsohn), Vol. 17. Henry, J. E., ed. Philadelphia, W. E. Saunders, 1984, pp. 459-4??. 12. K o m o r o w s k i, R. A., F a r m e r, S. G., H a n s o n, G. A., and H a u s e, L. L.: Cerebrospinal fluid lactic acid in diagnosis of meningitis. J. Clin. Microbiol. 8:89-92, 1978. 13. L a n n in g a n, R., M a c D o n a l d, M. A., M a r r ie, T. J., and V a n o r a, E.: Evaluation of cerebrospinal fluid lactic acid levels as an aid in the differential dignosis of bacterial and viral meningitis. J. Clin. Microbiol. 11:324-327, 1980. 14. L o u r ia, D. B., F e d e r, N., M i t c h e l l, W., and E m m o n s, C. W.: Influence of fungus strain and time in experimental histoplasmosis and volume of inoculum in cryptococcosis upon recovery of the fungi. J. Lab Clin. Med. 53:311-317, 1959. 15. M a n g h a m, D., G e r d i n g, D. N., P e t e r s o n, L. R., and S a r o s i, G. A.: Fungal meningitis manifesting as hydrocephalus. Arch. Intern. M ed. 143:728-731, 1983. 16. M a r b a c h, E. P. and W e i l, M. H.: Rapid enzymatic m easurem ent of blood lactate and pyruvate. Clin Chem. 16:314-325, 1967. 17. M c G i n n i s, M. R.: D e te c tio n o f fu n g i in c e r e b r o s p in a l flu id. Am. J. M e d. 75:129-138, 1983. 18. P o s n e r, J. and P l u m, F.: Independence of blood and cerebrospinal fluid lactate. Arch. Neuro. 16:492-496, 1967. 19. P o w e r s, W. J.: Cerebrospinal fluid to serum glucose ratios in diabetes mellitus and bacterial meningitis. Am. J. Med. 71:217-220, 1981.