Crum-Cianflone, Nancy F. MD, MPH*; Butera, Michael MD†
Although the first case of Candida meningitis was reported in 1933, Candida species have increasingly been described as important causes of central nervous system (CNS) infections.1 The rising occurrence of these infections mirrors the growing populace of immunocompromised hosts and patients with recent neurosurgical procedures; a review at 1 institution found that Candida is the fourth leading cause of meningitis among neurosurgical patients.2 Risk factors for Candida CNS infections include neurosurgical procedures, broad-spectrum antibiotics, indwelling catheters (eg, ventricular shunts), intravenous drug use, immunosuppressive therapy, and parenteral nutrition.2-4 Specific populations at risk consist of low birth weight infants and immunocompromised patients with acquired immune deficiency syndrome, cancer, or congenital disorders including severe combined immunodeficiency, chronic granulomatous disease, and myeloperoxidase deficiency.3,5-10 Most cases are caused by Candida albicans, followed by Candida parapsilosis, Candida tropicalis, and Candida krusei.11,12 We describe a rare case of meningitis and ventriculitis caused by Candida lusitaniae and provide a review of the literature summarizing the clinical manifestations, diagnosis, and treatment of this unusual infection.
An 84-year-old man underwent a ventriculoperitoneal (VP) shunt for the treatment of normal pressure hydrocephalus. The patient developed acute mental status changes and seizure activity on postoperative day 1; imaging showed an intracranial hemorrhage. He was treated with 5 days of intravenous Decadron to diminish cerebral edema. His condition stabilized until hospital day 5 when he developed fevers (103°F); a cerebrospinal fluid (CSF) specimen revealed 730 white cells/μL, with 93% neutrophils, 6000 red cells/μL, protein of 356 mg/dL, and glucose of 13 mg/dL. The CSF culture grew Enterobacter aerogenes sensitive to third-generation cephalosporins, carbapenems, and aminoglycosides. A computed tomography of the abdomen was negative for a viscus perforation or an abscess at the shunt tip. The shunt was externalized, and the peritoneal catheter tip grew Enterobacter. The patient was treated with intravenous cefotaxime and intrathecal gentamicin (2 mg in 3 mL of sterile dextrose for 5 days). Cerebrospinal fluid cultures became negative, and indices improved. The VP shunt was removed on day 18, and the VP shunt tip grew Enterobacter despite 12 days of antibiotic therapy with cefotaxime. A new internal shunt was simultaneously placed because the prior shunt tip abutted a blood vessel, and there was concern for recurrent intracranial bleeding.
The patient developed severe Clostridium difficile colitis, and the antibiotic regimen was switched to meropenem, intravenous metronidazole, and oral vancomycin; the latter 2 drugs were administered for the treatment of C. difficile. After improvement of the diarrhea, the patient was again treated with intravenous cefotaxime and remained in a stable, but bedridden, condition, with the ability to follow some commands. All follow-up CSF cultures obtained via the VP shunt were negative.
On hospital day 36, the patient developed decreased mental status, and the CSF fluid sampled via the shunt showed a protein of 144 mg/dL, glucose of 51 mg/dL, and a white blood cell (WBC) count of 55 cells/μL. The VP shunt was removed, and a culture of the old ventricular shunt tip grew C. lusitaniae. The patient was initially treated with liposomal amphotericin B (AmBisome) 5 mg/kg per day as well as intrathecal amphotericin B at 0.2 mg daily via a ventriculostomy for 4 days. The regimen was then changed to liposomal amphotericin B and 5-fluorocytosine (5FC); however, the patient developed a rash necessitating discontinuation of the 5FC. Sensitivities of the organism showed a minimum inhibitory concentration to amphotericin B of 0.3 μg/mL, fluconazole of 0.5 μg/mL, and caspofungin of 0.25 μg/mL. The liposomal amphotericin B at 5 mg/kg per day was continued for 12 days and then switched to intravenous fluconazole 400 mg twice daily. Cerebrospinal fluid indices 3 weeks after the diagnosis of Candida meningitis/ventriculitis showed a WBC count of 44 cells/μL, protein of 103 mg/dL, and glucose of 65 mg/dL; cultures were sterile. A new shunt was placed after 6 weeks of antifungal therapy, at which time the CSF had normalized, with a WBC count of 2 cells/μL, protein of 46 mg/dL, and glucose of 66 mg/dL, with repeat negative cultures. Fluconazole was discontinued after a total of 42 days of therapy (7 days after the new shunt was placed). Despite successful clearance of the C. lusitaniae infection, his mental status remained poor, and he was unresponsive to verbal stimuli. He eventually died because of his underlying medical problems and ventilator-associated pneumonia.
REVIEW OF THE LITERATURE
We performed search of MEDLINE (1966-2006), OVID (1950-1965), EMBASE (1974-2006), and the Parasitology Database (1972-2006) using the terms Candida or lusitaniae or yeast infections or fungi or mycotic AND meningitis or brain or neuro or ventriculitis or central nervous system infection to identify cases of C. lusitaniae CNS infection. We also searched for Clavispora lusitaniae, which is the teleomorph of C. lusitaniae, and included cases among large case series of fungal or Candida meningitis reports. We limited our search to human data in the English language.
Including the current case, we found 7 cases in the literature, with most infections occurring in infants and children (Table 1)3,6,13-16; our case represented the oldest person to develop C. lusitaniae meningitis. The most common presenting signs were fevers and/or seizures. Like our patient, most published cases occurred in the setting of prior antibiotic therapy and a recent neurosurgical procedure including VP shunt placement; the likely route of these infections was direct inoculation of the organism into the CNS. The CSF indices were variable, with a normal or elevated WBC (mean, 196/μL; range, 8-8250/μL); the protein level was often increased (mean, 228; range, 61-585 mg/dL), but only 2 cases had a significantly depressed glucose level. None of the Gram stains was positive; all diagnoses were all established by CSF culture. Earlier cases were predominantly treated with amphotericin B and 5FC, whereas more recent cases were treated with fluconazole. Fifty-seven percent of cases survived.
Infection due to C. lusitaniae was first described in 1979 in a leukemic patient with fungemia17,18; the first case of meningitis was reported in 1986.6 Our case involved a patient with a recent neurosurgical procedure with the presence of a VP shunt; the portal of entry was likely direct inoculation via this device. Whether the intrathecal gentamicin in a dextrose solution could have become contaminated is a possibility, although this is unlikely, given the numerous negative repeat CSF cultures in proximity to the time of administration. Like other Candida meningitis cases occurring in neurosurgical patients, our patient had preceding bacterial meningitis and had received broad-spectrum antibiotics; both are described risk factors.2,11 In addition, our patient had prolonged admission to the intensive care unit and parenteral nutrition and received intravenous steroids-all known risk factors for Candida infections.
Candida CNS infections may occur in 3 primary settings: (1) among neurosurgical patients who had a recent craniotomy and/or have a VP shunt in place, as in our case; (2) as a manifestation of disseminated disease, especially among premature neonates; and (3) rarely as isolated chronic meningitis.10,19 Neurosurgical patients often present with meningitis and/or ventriculitis secondary to direct inoculation of the organism into the CNS, whereas those with disseminated disease usually have hematogenous spread of the organism into the parenchymal brain tissue (eg, microabscesses formation). Patients with isolated chronic meningitis, which likely also occurs by the hematogenous route, present similar to other cases of fungal (eg, cryptococcal) meningitis.19
Signs and symptoms of Candida meningitis in neurosurgical patients may include headache, fever, seizures, and altered mental status; in addition, abdominal symptoms may occur in the presence of a VP shunt.11,20 Patients with microabscesses may have subtle or no neurological symptoms; hence, a high index of suspicion is necessary. Central nervous system involvement should be considered early because permanent neurological dysfunction may occur if the condition is not diagnosed and treated in a timely fashion. Rarely, patients with CNS Candida infections develop macroabscesses, vasculitis, or mycotic aneurysms, which may manifest by the development of focal neurological signs and/or seizures.10
The diagnostic workup of a possible Candida CNS infection should include blood cultures, CNS imaging, and a lumbar puncture. Candidemia is frequently found among patients with disseminated disease and CNS microabscesses, whereas those with neurosurgical-associated Candida meningitis often have sterile blood cultures because the source of their infection was at the local site. Patients with cerebral microabscesses are the most difficult to definitively diagnose owing to the small size of the lesions (<3 mm) and to the fact that the CSF is often unremarkable; often, these cases are diagnosed postmortem.10 Macroabscesses and vascular complications are typically detected on computed tomography or magnetic resonance imaging scans. Candida meningitis is established by obtaining CSF fluid. The CSF characteristics are typically undistinguishable from that of bacterial meningitis2; in cases of Candida meningitis, the CSF shows pleocytosis (mean, 600 WBC/μL) with an equal predominance of lymphocytes and neutrophils and an elevated protein; some cases exhibit hypoglycorrhachia (glucose, <40 mg/dL). The rate of wet mount or Gram stain positivity is reportedly 40%21 but was 0% in our series. The diagnosis usually relied on CSF culture results, which may lead to a delay in the diagnosis of 2 to 7 days. Methods, to enhance Candida isolation from the CSF, have been previously described.10 Because C. lusitaniae shares several microbiological characteristics with C. tropicalis and C. parapsilosis, misidentification may occur.13 Differentiating characteristics include C. lusitaniae's ability to assimilate and ferment cellobiose and to assimilate rhamnose.
The significance of a single CSF positive culture among neurosurgical patients has been questioned, as this finding may be due to colonization rather than a true infection.22 In addition, culturing the organism from a catheter tip, rather from the spinal fluid itself, also may represent colonization versus a true infection. Data that suggest that an isolate is an invasive organism rather than a contaminant includes abnormal CSF indices, the lack of mixed bacterial flora within cultures, and the presence of risk factors and symptoms of meningitis.4,22 Cases must be accessed on an individual basis; we would suggest that great care should be taken in discarding a positive CSF culture result. Our case had both risk factors for the development of Candida meningitis and clinical and laboratory signs of meningitis at the time of fungal isolation.
Treatment of Candida meningitis and parenchymal infections are unclear, given the lack of controlled trials. Treatment guidelines suggest that the treatment of choice is a combination of intravenous amphotericin B (0.7-1.0 mg/kg daily) and 5FC (25 mg/kg 4 times daily).2,23,24 Other potential regimens include intravenous lipid formulations of amphotericin B, intrathecal amphotericin B, or intravenous fluconazole.7,25,26 Lipid formulations of amphotericin B have been used successfully in Candida meningitis among neonates.26 Intrathecal amphotericin B is usually reserved for severe cases because of its potential to cause chemical arachnoiditis or ventriculitis.9 Fluconazole has been used alone or in combination with 5FC and can be considered when amphotericin B is not an option.10,19 Therapy should be continued for a minimum of 4 weeks after the CSF is sterilized and neurological symptoms/signs have resolved or stabilized.24 For patients with CNS infections associated with a VP shunt, removal is recommended for eradication of the organism, which may adhere to the biofilm layer of the device.2
Therapy for C. lusitaniae meningitis may be problematic considering the possibility for primary or secondary resistance to amphotericin B. The first clinical case of C. lusitaniae in the literature described the organism's ability to develop resistance to amphotericin B.17 Since this time, studies have shown that C. lusitaniae, unlike most other Candida species, can generate amphotericin B-resistant lineages from previously sensitive colonies at a rate of 1 in 104 cells in vitro.27 Emergence of resistance has been attributed to alteration in the expression of ergosterol biosynthesis genes.28 Most experts recommend performing susceptibility testing on an initial Candida isolate causing a CNS infection. If the patient clinically deteriorates, or if cultures remain positive, repeat susceptibility testing on subsequent isolates and antifungal therapy modification because of the possibility of the development of resistance have been recommended. That being said, a recent review of C. lusitaniae infections showed that susceptibility testing results for amphotericin B did not predict clinical response or outcome.29 At this time, the treatment of choice for C. lusitaniae is unknown, but fluconazole may be the preferred agent because of its efficacy against most isolates and its favorable pharmacological profile including good CNS penetration.
The mortality rate of Candida meningitis without therapy approaches 100%; despite the availability of antifungal agents, the death rate remains substantial at approximately 10% to 30%.3,8,11,20,22 Mortality rates for C. lusitaniae CNS infections were 43% in our review. A poor outcome among patients with Candida meningitis has been associated with a delayed diagnosis, low CSF glucose levels (<35 mg/dL), and the development of intracranial hypertension or focal neurological symptoms.11
In summary, Candida species are an increasingly recognized cause of meningitis, especially among neurosurgical patients and immunocompromised hosts who have received prior broad-spectrum antibacterial therapy. Nonalbicans Candida infections of the CNS will likely rise because of the use of prophylactic antifungal agents in at-risk populations. Novel diagnostic strategies are needed to achieve earlier diagnosis of these infections, which are typically only identified by culture. Prompt recognition and treatment are advocated, given the high mortality rate associated with these emerging infections.
1. Smith LW, Sano ME. Moniliasis with meningeal involvement. J Infect Dis. 1933;53:187-196.
2. Nguyen MH, Yu VL. Meningitis caused by Candida species: an emerging problem in neurosurgical patients. Clin Infect Dis. 1995;21:323-327.
3. Huttova M, Kralinsky K, Horn J, et al. Prospective study of nosocomial fungal meningitis in children-report of 10 cases. Scand J Infect Dis. 1998;30:485-487.
4. Arisoy ES, Arisoy AE, Dunne WM Jr. Clinical significance of fungi isolated form cerebrospinal fluid in children. Pediatr Infect Dis J. 1994;13:128-133.
5. Fernandez M, Moylett EH, Noyola DE, et al. Candidal meningitis in neonates: a 10-year review. Clin Infect Dis. 2000;31:458-463.
6. Fleischmann J, Church JA, Lehrer RI. Case report: Candida meningitis and chronic granulomatous disease. Am J Med Sci. 1986;291:334-341.
7. Marr B, Gross S, Cunningham C, et al. Candidal sepsis and meningitis in a very-low-birth-weight infant successfully treated with fluconazole and flucytosine. Clin Infect Dis. 1994;19:795-796.
8. Casado JL, Quereda C, Oliva J, et al. Candidal meningitis in HIV-infected patients: analysis of 14 cases. Clin Infect Dis. 1997;25:673-676.
9. McCullers JA, Vargas SL, Flynn PM, et al. Candidal meningitis in children with cancer. Clin Infect Dis. 2000;31:451-457.
10. Sanchez-Portocarrero J, Perez-Cecilia E, Corral O, et al. The central nervous system and infection by Candida species. Diagn Microbiol Infect Dis. 2000;37:169-179.
11. Bayer AS, Edwards JE, Seidel JS, et al. Candida meningitis. Medicine. 1976;55:477-486.
12. Dorko E, Pilipcinec E, Tkacikova L. Candida species isolated from cerebrospinal fluid. Folia Microbiol (Praha). 2002;47:179-181.
13. Sanchez PJ, Cooper BH. Candida lusitaniae sepsis and meningitis in a neonate. Pediatr Infect Dis J. 1987;6:758-759.
14. Leggiadro RJ, Collins T. Postneurosurgical Candida lusitaniae meningitis. Pediatr Infect Dis J. 1988;7:366-367.
15. Sarma PSA, Durairaj P, Padhye AA. Candida lusitaniae causing fatal meningitis. Postgrad Med J. 1993;69:878-880.
16. Kremery V Jr, Mateicka F, Grausova S, et al. Invasive infections due to Clavispora lusitaniae. FEMS Immunol Med Microbiol. 1999;23: 75-78.
17. Pappagianis D, Collins MS, Hector R, et al. Development of resistance to amphotericin B in Candida lusitaniae infecting a human. Antimicrob Agents Chemother. 1979;16:123-126.
18. Holzschu DL, Presley HL, Miranda M, et al. Identification of Candida lusitaniae as an opportunistic yeast in humans. J Clin Microbiol. 1979; 10:202-205.
19. Voice RA, Bradley SF, Sangeorzan JA, et al. Chronic candidal meningitis: an uncommon manifestation of candidiasis. Clin Infect Dis. 1994;19:60-66.
20. Sanchez-Portocarrero J, Martin-Rabadan P, Daldana CJ, et al. Candida cerebrospinal fluid infection. Report of two new cases and review of the literature. Diagn Microbiol Infect Dis. 1994;20:33-40.
21. Lipton SA, Hickey VF, Morris JH, et al. Candidal infection in the central nervous system. Am J Med. 1984;76:101-109.
22. Geers TA, Gordon SM. Clinical significance of Candida species isolated from cerebrospinal fluid following neurosurgery. Clin Infect Dis. 1999;28:1139-1147.
23. Smego RA, Perfect JR, Durack DT. Combined therapy with amphotericin B and fluorocytosine for Candida meningitis. Rev Infect Dis. 1984;6:791-801.
24. Pappas PG, Rex JH, Sobel JD, et al. Infectious Diseases Society of America. Guidelines for treatment of candidiasis. Clin Infect Dis. 2004;38:161-189.
25. Gurses N, Kalyci AG. Fluconazole monotherapy for candidal meningitis in a premature infant. Clin Infect Dis. 1996;23:645-646.
26. Scarcella A, Pasquariello MB, Giugliano B, et al. Liposomal amphotericin B treatment for neonatal fungal infections. Pediatr Infect Dis J. 1998;17:146-148.
27. Yoon SA, Vazquez JA, Steffan PE, et al. High-frequency, in vitro reversible switching of Candida lusitaniae clinical isolates from amphotericin B susceptibility to resistance. Antimicrob Agents Chemother. 1999;43:836-845.
28. Young LY, Hull CM, Heitman J. Disruption of ergosterol biosynthesis confers resistance to amphotericin B in Candida lusitaniae. Antimicrob Agents Chemother. 2003;47:2717-2724.
29. Hawkins JL, Baddour LM. Candida lusitaniae infections in the era of fluconazole availability. Clin Infect Dis. 2003;36:e14-e18.
© 2007 Lippincott Williams & Wilkins, Inc.