Serum procalcitonin levels in bacterial and abacterial meningitis

Stefan Schwarz, MD; Markus Bertram, MD; Stefan Schwab, MD; Konrad Andrassy, MD; Werner Hacke, MD

Objectives: To test the hypothesis that serum procalcitonin (PCT) levels are elevated in patients with bacterial meningitis and remain within normal limits in patients with abacterial meningitis. Design: Prospective case series. Setting: Tertiary care center. Patients: A total of 30 patients (13 men and 17 women) with a mean age of 52 yrs, having acute bacterial (n 16) or abacterial (n 14) meningitis. Interventions: Blood and cerebrospinal uid samples. Measurements and Main Results: Patients with abacterial meningitis were younger and had a shorter hospital stay. Of 16 patients with bacterial meningitis, 14 were in a septic condition at admission, but only 5 of 14 patients with abacterial meningitis were in a septic condition at admission. At discharge, 12 patients were without symptoms, 9 patients were moderately disabled, and 9 were severely disabled. No patient died. At admission, PCT, C-reactive protein, white blood cell and cerebrospinal uid leu-

kocyte counts, and cerebrospinal uid protein and lactate levels were higher and the serum/cerebrospinal uid glucose quotient was lower in patients with bacterial meningitis as compared with those with abacterial meningitis (p < .001). PCT was the variable with the highest specicity for bacterial infections (100%), but there were false-negative ndings in ve patients with bacterial meningitis (a sensitivity of 69%). Persistently elevated or increasing PCT levels after 2 days were associated with an unfavorable clinical course. Conclusions: Our results indicate that PCT is a useful additional variable for distinguishing bacterial from abacterial meningitis. In patients with abacterial meningitis, PCT levels do not increase even in cases of viral sepsis. Elevated PCT levels indicate a bacterial origin with high specicity, but false-negative results can occur. (Crit Care Med 2000; 28:1828 1832) KEY WORDS: sepsis; meningitis; procalcitonin; C-reactive protein; cerebrospinal uid; lactate

rocalcitonin (PCT) is a glycopeptide consisting of 116 amino acids produced under normal conditions in the C cells of the thyroid gland as the precursor molecule of calcitonin. In healthy subjects, PCT is undetectable or serum levels are very low (0.15 ng/mL) (1). In 1993, PCT was rst described as a prognostic indicator and marker of the extent and course of the systemic inammatory response to bacterial and fungal infections (2). Several subsequent studies demonstrated that PCT levels are closely related to severe invasive bacterial infections and decrease rapidly after appropriate antibiotic therapy (27). In contrast, localized bacterial infections, severe viral infections, and inammatory reactions of noninfectious origin do not or only slightly increase PCT levels (2, 8 10). The cellular source and mechanisms of PCT production in septic conditions are still unclear (211). A recent animal study suggested that PCT participates as a me-

From the Department of Neurology (Drs. Schwarz, Bertram, Schwab, Hacke), and the Department of Nephrology (Dr. Andrassy), University of Heidelberg, Heidelberg, Germany. Copyright 2000 by Lippincott Williams & Wilkins

diator in the systemic inammatory response in sepsis (12). The unique feature that PCT levels increase in bacterial and fungal infections, but remain unchanged even in severe viral infections and other inammatory diseases, makes PCT attractive as a potential diagnostic variable for the diagnosis of bacterial meningitis. Conventional clinical and laboratory variables such as cerebrospinal uid cytology, lactate and glucose, C-reactive protein (CRP), and leukocyte count are often not discriminative enough in the early phase of the disease (13). In patients with positive blood or cerebrospinal uid cultures and Gram staining of the cerebrospinal uid, the diagnosis of a bacterial origin is proven and further tests are not necessary. However, these tests have an overall sensitivity of 60% to 90% (14 17). In consequence, patients with meningitis in whom bacterial infection cannot be completely ruled out usually receive empirical antibiotic treatment until the microbiological ndings or the clinical course exclude a bacterial origin. PCT measurements have been reported to be a reliable tool for differentiation between viral and bacterial meningitis in children, but have not been systematically investigated in

adults (18). We measured PCT levels in adults with meningitis to test the hypothesis that PCT is elevated in patients with bacterial meningitis and remains within normal limits in patients with abacterial meningitis.

PATIENTS AND METHODS

We prospectively studied 30 consecutive patients (13 men, 17 women; mean age, 52 yrs; range, 16 87 yrs) who were hospitalized for acute meningitis at the Neurology Department of the University of Heidelberg Hospital from December, 1997 through September, 1998. The diagnosis of meningitis was based on compatible clinical features and cerebrospinal uid ndings. Patients who had received antibiotic treatment before admission were excluded from this study. As routine cerebrospinal uid variables, a leukocyte count was taken, a cytological study was performed, and protein, glucose, lactate and immunoglobulin levels were determined at admission, and the serum/cerebrospinal uid glucose relation was calculated. The microbiological work-up consisted of Gram staining and microscopy of the centrifuged cerebrospinal uid, detection of streptococcal antigens, and cerebrospinal uid cultures. According to the clinical and cerebrospinal uid ndings, serological and polymerase chain reaction tests from peripheral blood and cere-

1828

Crit Care Med 2000 Vol. 28, No. 6

brospinal uid for specic infectious agents were performed. Simultaneously with the lumbar puncture, blood samples were taken for routine laboratory variables including CRP, liver enzymes, serum urea and creatinine, white blood cell and platelet counts, lactate, and blood cultures. Before lumbar puncture, a brain computed tomographic (CT) scan was obtained. At admission and after 48 12 hrs, serum samples were taken for PCT measurements. These samples were frozen within 4 hrs and stored at 32C (25.6F) for later analysis. The following clinical data were collected from the initial examination: presence of neurologic abnormalities, body core temperature, systolic blood pressure, heart rate, and breathing rate. Sepsis was diagnosed according to the American College of Chest Physicians/ Society of Critical Care Medicine consensus classication (19) if two or more of the following signs were found: body temperature 38C (100.4F) or 36C (96.8F), heart rate 90 beats/min, breathing frequency 20 breaths/min, and leukocytosis (leukocyte count 12 109/L) or leukopenia (leukocyte count 4 109/L). Severe sepsis was diagnosed if sepsis was associated with organ dysfunction, signs of hypoperfusion, or hypotension. Septic shock was dened as sepsisinduced hypotension despite uid resuscitation, and signs of hypoperfusion. Immediately after the diagnosis of meningitis was established, all patients in whom a bacterial origin could not be ruled out were treated with antibiotic drugs, usually with a combination of a third-generation cephalosporin (cefotaxim) with ampicillin. The antibiotic regimen was later modied depending on the results of the microbiological studies. Corticosteroids were not given. Patients with presumed or proven herpes simplex meningoencephalitis were treated with acyclovir. Clinical outcome was classied at discharge according to the Glasgow Outcome Scale (GOS) into the following categories: dead (GOS 5), vegetative state (GOS 4), severe disability (GOS 3), moderate disability (GOS 2), and no decit (GOS 1) (20). At discharge, all microbiological, clinical and laboratory ndings were summarized, and the origin of the disease was categorized into bacterial, parasitic/fungal abacterial, or unknown. The diagnosis of bacterial meningitis was based in slight variation of the criteria of Durand et al. (17) on compatible clinical features and one of the following: a) positive cerebrospinal uid culture; b) negative cerebrospinal uid culture, but either a positive cerebrospinal uid antigen test or identication of bacteria on Gram staining of the cerebrospinal uid or a positive blood culture; or c) a cerebrospinal uid pleocytosis of 500 neutrophils/L and rapid improvement after

antibacterial therapy despite negative cerebrospinal uid and blood cultures. Fungal infection was diagnosed by detection of fungal or bacterial antigens in the cerebrospinal uid or by microscopy. The diagnosis of abacterial meningitis was established in patients with negative blood and cerebrospinal uid cultures and antigen tests for bacterial infection and either positive viral culture in the cerebrospinal uid or positive polymerase chain reaction results or specic antiviral IgM antibodies in cerebrospinal uid and serum, or cerebrospinal uid pleocytosis with a predominance of mononuclear cells and recovery without antibiotic treatment. We chose the term abacterial to summarize patients with a proven viral infection and patients with abacterial meningitis in whom a viral agent could not be identied, although a viral infection is also highly probable in this group. Patients who did not t into one of these groups were classied as having meningitis of unknown origin. PCT serum levels were measured from the frozen samples with a specic immunoluminometric assay (LUMItest Procalcitonin, B.R.A.H.M.S. Diagnostica, Berlin, Germany). This assay, which is specic for the PCT molecule, uses a sandwich type luminescence immunoassay and a coated-tube technique. Two monoclonal antibodies are directed against two different sites of the PCT molecule, the C-terminal catacalcin and midregional calcitonin sequences. The anticatacalcin antibody is immobilized on the surface of the coated tube, and the anticalcitonin antibody is coupled to a luminescent acridine derivative. Serum samples (20-L) are incubated for 2 hrs. The samples are then placed in a luminometer and hydrogen peroxide and sodium peroxide solutions are automatically injected. These substances react with the acridine derivative bound to the anticalcitonin antibody that emits light as it transforms into acridone. The emitted light intensity is directly proportional to the PCT concentration, which is calculated from standard concentrations. The measurements take 2.5 hrs. To exclude any effect of the PCT results on the diagnosis, these measurements were performed after each patient had been discharged and after the origin had been determined. All PCT measurements were performed as double measurements by the same laboratory technician, who was blinded to the clinical diagnosis and other laboratory results. PCT levels 0.5 ng/mL were considered as abnormal. For statistical analysis, we used the nonparametric Mann-Whitney U test to detect differences between the groups. The association between origin of the disease and clinical outcome or presence of sepsis was tested with Fishers exact test. Because the diagnostic variables evaluated in this study could affect

future therapeutic management, a strict level of signicance () was chosen (p .01). This study was conducted according to local ethics committee standards, and informed consent for taking additional blood samples and data analysis was obtained from each patient.

RESULTS
The origin of meningitis was bacterial in 16 patients and abacterial in 14 patients. Parasitic or fungal infections were not observed. Clinical variables and the results from microbiological studies are depicted in Table 1. The causative infectious agent could be identied in 11 of 16 patients with bacterial infection. In 7 of 14 patients with abacterial meningitis, a viral origin was identied. In 6 of the 16 patients with bacterial infection, the bacterial origin was proven by microscopy shortly after admission. In the remaining ten patients with bacterial infection, a bacterial origin was conrmed by positive blood and/or cerebrospinal uid cultures; in one patient by the presence of serum and cerebrospinal uid Immunoglobulin M antibodies against Borrelia burgdorferi. In ve patients, bacterial infection was diagnosed on the grounds of typical cerebrospinal uid ndings and a clinical course unambiguous for bacterial meningitis. All ve patients with culturenegative bacterial meningitis had purulent cerebrospinal uid with 1,000 neutrophils/L and recovered rapidly with antibacterial therapy. The meningitis was hospital-acquired in only one patient (Klebsiella pneumoniae meningitis in a woman with severe Crohns disease). The staphylococcal meningitis in one patient was secondary to a stitch abscess after intramuscular injection of an analgesic drug for chronic arthritis. A total of 21 patients were initially treated with antibiotics. In ve patients, antibiotic therapy was later stopped after the diagnosis of a viral infection was established. Patients with abacterial meningitis were younger and had a shorter hospital stay (p .001). Five of the 14 patients with abacterial meningitis presented with sepsis, but none of these patients had severe sepsis or septic shock. In contrast, a septic condition was absent in two patients with bacterial meningitis only. One of these patients had culture-positive tuberculous meningitis; the other was diagnosed with acute Lyme meningitis. Outcome tended to be better in patients with abacterial meningitis (p .05). At hospi1829

Crit Care Med 2000 Vol. 28, No. 6

Table 1. Clinical variables and microbiological results Variable No. Age (mean and range) Sepsis Absent Sepsis Severe sepsis Septic shock GOS GOS 1 GOS 2 GOS 3 Length of hospital stay (mean and range) CSF microscopy Cultural/serological/PCR proven origin All Patients 30 52 (1687) 11 14 4 1 12 9 9 16 (350) n 18 Bacterial 16 61 (1687) 2 9 4 1 3 5 8 22 (758) 6 positive n 11 Borrelia burgdorferi (n 1) Haemophilus inuenzae (n 1) Klebsiella pneumoniae (n 1) Streptococcus pneumoniae (n 6) Staphylococcus aureus (n 1) Mycobacterium tuberculosis (n 1) Abacterial 14 42 (1948) 9 5 0 0 9 4 1 9 (318) n7 Epstein-Barr (n 1) Herpes simplex 1 (n 2) Herpes simplex 2 (n 1) Enterovirus (n 1) Central european encephalitis (n 2) p Value NA .001a .05

.05

.001a NA NA

NA, data not available; GOS, Glasgow Outcome Scale at discharge; CSF, cerebrospinal uid; PCR, polymerase chain reaction. GOS 45 were not noted; GOS 3, severe disability; GOS 2, moderate disability; GOS 1, no decit. a Statistically signicant.

tal discharge, 9 of the 14 patients with abacterial meningitis were without remaining symptoms, but only 3 of 16 patients with bacterial meningitis were without remaining symptoms. None of the patients died. The initial serum and cerebrospinal uid variables are shown in Figure 1. Patients with bacterial meningitis had signicantly higher serum PCT, CRP, and cerebrospinal uid lactate and protein levels and a higher blood and cerebrospinal uid leukocyte count, and a lower cerebrospinal uid/serum glucose quotient. The median initial serum PCT level was 1.75 ng/mL (range, 0.16 59.92 ng/ mL) in bacterial meningitis as compared with 0.24 ng/mL (range, 0.12 0.29 ng/ mL) in abacterial meningitis (p .0001). With a cutoff level for PCT of 0.5 ng/ mL, the specicity for diagnosis of bacterial meningitis was 100% (95% condence interval, 0.79 1.0) with a sensitivity of 69% (95% condence interval, 0.41 0.89). The positive predictive value for bacterial meningitis was 1.0; the negative predictive value was 0.74. Although no patient with abacterial meningitis, including ve patients with sepsis, had abnormal PCT levels, PCT remained within normal limits in ve patients with bacterial meningitis. Two of these patients have been mentioned: a 34-yr-old woman with proven tuberculous meningitis and a 69-yr-old man with Lyme meningitis. Neither patient had
1830

signs of a systemic inammatory response. The three remaining patients (a 44-yr-old woman with staphylococcal meningitis, a 74-yr-old woman with streptococcal meningitis, and an 87-yr old woman with meningitis attributable to Haemophilus inuenzae) in whom PCT measurements had failed to detect a bacterial infection had a septic condition; one presented with septic shock. In one of these ve patients with bacterial meningitis and initially normal PCT levels, PCT increased during the next 48 hrs from 0.32 ng/mL to 3.69 ng/mL, but in the other four patients, the PCT levels remained normal. After 48 12 hrs, median PCT values had decreased from a median of 1.75 to 1.05 in patients with bacterial meningitis. In three of these patients, PCT levels remained unchanged or even increased. All three patients had a complicated clinical course (hospital stays of 24, 28, and 58 days) and remained severely disabled (GOS 3). In all patients with abacterial meningitis, PCT levels remained within normal values at the follow-up analysis after 48 12 hrs. As compared with PCT, the other variables had a lower specicity, but some (CRP, white blood cell count, and cerebrospinal uid lactate) had a higher sensitivity for bacterial infections. The initial median CRP level was 174 mg/L (range, 7 400 mg/L) in bacterial and 7 mg/L (range, 2 65 mg/L) in abacterial menin-

gitis. Using a upper normal limit of 8 mg/L, elevated CRP levels were present in 15 of 16 patients with bacterial meningitis (a sensitivity of 94%), but did not discriminate abacterial meningitis as precisely as PCT did (a specicity of 57%). In four of ve patients with abacterial meningitis presenting with sepsis, CRP was moderately elevated (up to 65 mg/L). However, highly elevated CRP levels (100 mg/L) were observed exclusively in patients with bacterial meningitis. The most sensitive cerebrospinal uid variable for bacterial infection was cerebrospinal uid lactate content (upper normal limit, 2.1 mmol/L). Median cerebrospinal uid lactate was 2.54 mmol/L (range, 1.7 4.8 mmol/L) in abacterial and 8.45 mmol/L (range, 1.720.3 mmol/L) in bacterial meningitis. Elevated cerebrospinal uid lactate levels were present in 15 of 16 patients with bacterial meningitis (a sensitivity of 94%), but the specicity was low (42%). White blood cell counts, serum lactate levels, cerebrospinal uid cells, cerebrospinal uid protein values, and the cerebrospinal uid/serum glucose quotient all showed relatively wide overlap (Fig. 1).

DISCUSSION
In this study, we showed that with a cutoff of 0.5 ng/mL, PCT levels are increased in patients with bacterial meningitis and remain within normal ranges in
Crit Care Med 2000 Vol. 28, No. 6

O
gitis.

ur results indicate that procalcitonin is a use-

ful additional variable to distinguish bacterial meningitis from abacterial menin-

Figure 1. Initial serum and cerebrospinal uid (CSF) variables in patients with bacterial (n 16) and abacterial (n 14) meningitis. Data are median, 25th-75th percentiles (boxes), 10th-90th percentiles (whiskers), and range (circles). The dotted lines represent the upper normal limits. All data except for CSF/serum glucose quotient are presented on logarithmic scales. CRP, C-reactive protein.

patients with abacterial meningitis. Compared with conventional infection variables such as CRP, PCT had the highest specicity for bacterial infections without false-positive results. These results are consistent with the observations of Gendrel et al. (18) in children with meningitis. During the last few years, a variety of cytokines and other markers of infection have been investigated to distinguish bacterial from viral meningitis. Most of these studies were performed in children. Cerebrospinal uid neopterin levels are higher in patients with bacterial meningitis, but do not distinguish bacterial from viral infection with sufcient accuracy (21). Interleukin (IL)-6 cerebrospiCrit Care Med 2000 Vol. 28, No. 6

nal uid measurements yielded conicting results. Some authors concluded that IL-6 cerebrospinal uid levels rise specifically in patients with bacterial meningitis (2224), but others found IL-6 not to be a useful variable in the differential diagnosis of meningitis (2527). Many other cerebrospinal uid variables have been proposed to distinguish bacterial from viral meningitis, among them cerebrospinal uid IL-8, tumor necrosis factor, transforming growth factor, soluble tumor necrosis factor receptor, or complement C3 and factor B cerebrospinal uid concentrations (28 32). None of these variables are used in clinical routine. In comparison with these variables, PCT has several theoretical advantages.

PCT levels rise rapidly after induction. In a human experimental study, PCT levels were raised 4 hrs after injection of bacterial endotoxin and remained elevated for 24 hrs because of its long half-life (11). Unlike many cytokines, PCT can be drawn with regular blood samples, remains stable for several hours even at room temperature, and can be measured along with other routine variables. The analysis is standardized, easy to perform, and takes 2.5 hrs (33). PCT is increasingly used as a standard laboratory variable to monitor systemic infections and is available in many clinical laboratories on a 24-hr basis. It should be taken into account that although PCT levels increase specically in bacterial infections, false-negative results may occur. PCT is an indicator of a systemic inammatory response to severe invasive bacterial infections but is not a marker of bacterial infection per se (34). Therefore, it is not surprising that PCT levels did not increase in two of our patients with bacterial meningitis conned to the central nervous system without a systemic inammatory response. In addition, these two patients suffered from uncommon bacterial infections (tuberculosis and Lyme disease) in which unusual inammatory pathways may be activated. To date, PCT has not been investigated systematically in patients with Lyme disease or tuberculosis. In a few human immunodeciency virus-positive patients with mycobacterial infections, PCT levels did not increase (35). In this study, three patients with sepsis attributable to proven purulent bacterial meningitis showed normal PCT levels. Similar cases of false-negative PCT measurements in bacterial sepsis have been described previously (7, 36). Because the pathophysiological mechanisms of PCT production
1831

are largely unclear, these cases remain unexplained.

CONCLUSIONS
Our results indicate that PCT is a useful additional variable to distinguish bacterial meningitis from abacterial meningitis. Even in patients with septic abacterial meningitis, PCT levels do not increase. Although false-negative results can occur, elevated PCT levels indicate a bacterial origin with high specicity.
11.

12.

13.

REFERENCES
1. Becker K, Nylen E, Cohen R, et al: Calcitonin: Structure, molecular biology, and actions. In: Principles of Bone Biology. Bilenzikian J, Raisz L, Rodan G (Eds). New York, Academic Press, 1996, pp 471 494 2. Assicot M, Gendrel D, Carsin H, et al: High serum procalcitonin concentrations in patients with sepsis and infection. Lancet 1993; 341:515518 3. Smith MD, Suputtamongkol Y, Chaowagul W, et al: Elevated serum procalcitonin levels in patients with melioidosis. Clin Infect Dis 1995; 20:641 645 4. Gendrel D, Assicot M, Raymond J, et al: Procalcitonin as a marker for the early diagnosis of neonatal infection. J Pediatr 1996; 128: 570 573 5. Al-Nawas B, Shah PM: Procalcitonin in patients with and without immunosuppression and sepsis. Infection 1996; 24:434 436 6. Chiesa C, Panero A, Rossi N, et al: Reliability of procalcitonin concentrations for the diagnosis of sepsis in critically ill neonates. Clin Infect Dis 1998; 26:664 672 7. Monneret G, Labaune JM, Isaac C, et al: Procalcitonin and C-reactive protein levels in neonatal infections. Acta Paediatr 1997; 86: 209 212 8. Eberhard OK, Haubitz M, Brunkhorst FM, et al: Usefulness of procalcitonin for differentiation between activity of systemic autoimmune disease (systemic lupus erythematosus/systemic antineutrophil cytoplasmic antibody-associated vasculitis) and invasive bacterial infection. Arthritis Rheum 1997; 40:1250 1256 9. Rau B, Steinbach G, Gansauge F, et al: The potential role of procalcitonin and interleukin 8 in the prediction of infected necrosis in acute pancreatitis. Gut 1997; 41:832 840 10. de-Werra I, Jaccard C, Corradin SB, et al: Cytokines, nitrite/nitrate, soluble tumor ne-

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

crosis factor receptors, and procalcitonin concentrations: Comparisons in patients with septic shock, cardiogenic shock, and bacterial pneumonia. Crit Care Med 1997; 25:607 613 Dandona P, Nix D, Wilson MF, et al: Procalcitonin increase after endotoxin injection in normal subjects. J Clin Endocrinol Metab 1994; 79:16051608 Nylen ES, Whang KT, Snider R Jr, et al: Mortality is increased by procalcitonin and decreased by an antiserum reactive to procalcitonin in experimental sepsis. Crit Care Med 1998; 26:10011006 Lindquist L, Linne T, Hansson L, et al: Value of cerebrospinal uid analysis in the differential diagnosis of meningitis: A study of 710 patients with suspected central nervous system infection. Eur J Clin Microbiol Infect Dis 1988; 7:374 380 Roos K, Scheld W: The management of fulminant meningitis in the intensive care unit. Crit Care Med 1988; 4:375392 Roos K, Tunkel A, Scheld W: Acute bacterial meningitis in children and adults. In: Infections of the Central Nervous System. Scheld W, Whitley R, Durack D (Eds). New York, Raven Press, 1991, pp 335 409 Sigurdardottir B, Bjo rnsson O, Jonsdottir K, et al: Acute bacterial meningitis in adults. Arch Intern Med 1997; 157:425 430 Durand M, Calderwood S, Weber D, et al: Acute bacterial meningitis in adults. N Engl J Med 1993; 328:2128 Gendrel D, Raymond J, Assicot M, et al: Measurement of procalcitonin levels in children with bacterial or viral meningitis. Clin Infect Dis 1997; 24:1240 1242 The ACCP/SCCM American College of Chest Physicians/Society of Critical Care Medicine Consensus Conference: Denitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis. Crit Care Med 1992; 20:864 874 Jennett B, Bond M: Assessment of outcome after severe brain damage. A practical scale. Lancet 1975; i:480 484 Hagberg L, Dotevall L, Norkrans G, et al: Cerebrospinal uid neopterin concentrations in central nervous system infection. J Infect Dis 1993; 168:12851288 Dulkerian SJ, Kilpatrick L, Costarino A Jr, et al: Cytokine elevations in infants with bacterial and aseptic meningitis. J Pediatr 1995; 126:872 876 Matsuzono Y, Narita M, Akutsu Y, et al: Interleukin-6 in cerebrospinal uid of patients with central nervous system infections. Acta Paediatr 1995; 84:879 883

24. Hashim IA, Walsh A, Hart CA, et al: Cerebrospinal uid interleukin-6 and its diagnostic value in the investigation of meningitis. Ann Clin Biochem 1995; 32:289 296 25. Torre D, Zeroli C, Ferraro G, et al: Cerebrospinal uid levels of IL-6 in patients with acute infections of the central nervous system. Scand J Infect Dis 1992; 24:787791 26. Lopez-Cortes LF, Cruz-Ruiz M, GomezMateos J, et al: Interleukin 6 in cerebrospinal uid of patients with meningitis is not a useful diagnostic marker in the differential diagnosis of meningitis. Ann Clin Biochem 1997; 34:165169 27. Chavanet P, Bonnotte B, Guiguet M, et al: High concentrations of intrathecal interleukin-6 in human bacterial and nonbacterial meningitis. J Infect Dis 1992; 166:428 431 28. Seki T, Joh K, Oh-ishi T: Augmented production of interleukin-8 in cerebrospinal uid in bacterial meningitis. Immunology 1993; 80: 333335 29. Ossege LM, Sindern E, Voss B, et al: Expression of tumor necrosis factor-alpha and transforming growth factor-beta 1 in cerebrospinal uid cells in meningitis. J Neurol Sci 1996; 144:113 30. Leist TP, Frei K, Kam-Hansen S, et al: Tumor necrosis factor alpha in cerebrospinal uid during bacterial, but not viral, meningitis. Evaluation in murine model infections and in patients. J Exp Med 1988; 167:17431748 31. Ichiyama T, Hayashi T, Furukawa S: Cerebrospinal uid concentrations of soluble tumor necrosis factor receptor in bacterial and aseptic meningitis. Neurology 1996; 46: 837 838 32. Stahel P, Nadal D, Pster H, et al: Complement C3 and factor B cerebrospinal uid concentrations in bacterial and aseptic meningitis. Lancet 1997; 349:1886 1887 33. Meisner M, Tschaikowsky K, Schnabel S, et al: Procalcitonin-inuence of temperature, storage, anticoagulation and arterial or venous asservation of blood samples on procalcitonin concentrations. Eur J Clin Chem Clin Biochem 1997; 35:597 601 34. Karzai W, Oberhoffer M, Meier-Hellmann A, et al: Procalcitonin-a new indicator of the systemic response to severe infections. Infection 1997; 25:329 334 35. Gerard Y, Hober D, Assicot M, et al: Procalcitonin as a marker of bacterial sepsis in patients infected with HIV-1. J Infect 1997; 35:41 46 36. Huber W, Schweigart U, Bottermann P: Failure of PCT to indicate severe fungal infection in two immunodecient patients. Infection 1997; 25:377378

1832

Crit Care Med 2000 Vol. 28, No. 6