BRIEF REPORT Recurrent Ventriculoperitoneal Shunt Infection Caused by Small-Colony Variants of Staphylococcus aureus Teresa Spanu, 1 Lucio Romano, 1 Tiziana D Inzeo, 1 Luca Masucci, 1 Alessio Albanese, 2 Fabio Papacci, 2 Enrico Marchese, 2 Maurizio Sanguinetti, 1 and Giovanni Fadda 1 1 Institute of Microbiology and 2 Department of Neurosurgery, Catholic University of the Sacred Heart, Rome, Italy Phenotypic variants of Staphylococcus aureus may be misidentified by routine microbiological methods, and they may also respond poorly to antibacterial treatment. Using molecular methods, we identified small-colony variants of methicillin-resistant S. aureus (which were misidentified by 3 widely used automated identification systems as methicillinsusceptible coagulase-negative staphylococci) as the cause of recurrent ventriculoperitoneal shunt related meningitis. Small-colony variants of Staphylococcus aureus (SCV-SA) are most commonly associated with respiratory tract infections in patients with cystic fibrosis and with foreign-body infections [1 5]. They are naturally occurring subpopulations characterized by an inability to synthesize thymidine or by deficient electron transport activity caused by their auxotrophism for hemin or menadione [1, 6, 7]. These alterations are responsible for the smallness of their colonies, which are nonpigmented and nonhemolytic on Columbia agar; their weak coagulase production; their failure to metabolize mannitol and other sugars; their altered exotoxin expression [6, 7]; and their unsatisfactory response to certain antimicrobials [1, 4, 6, 7]. SCV-SA can survive within mammalian cells for long periods without provoking host-cell death [1, 2], thanks to their diminished production of a-toxin, which, at wild-type levels, normally lyses eukaryotic cells; this feature helps explain the recurrence of certain S. aureus infections (occuring months or even years after their apparent eradication) [6, 7]. The frequency of SCV-SA infection may be underestimated [3, 6]. SCV-SA strains are often misidentified by the automated Received 20 January 2005; accepted 22 April 2005; electronically published 20 July 2005. Reprints or correspondence: Dr. Teresa Spanu, Istituto di Microbiologia, Università Cattolica del Sacro Cuore, Largo Francesco Vito 1, 00168 Rome, Italy (tspanu@rm.unicatt.it). Clinical Infectious Diseases 2005; 41:e48 52 2005 by the Infectious Diseases Society of America. All rights reserved. 1058-4838/2005/4105-00E2$15.00 systems routinely used in many microbiology laboratories [8]. Their slow growth also reduces the accuracy of conventional methods to determine methicillin resistance [9]. Awareness of these difficulties has led to the development of genetic identification tools for use when SCV-SA infection is clinically suspected. We describe a case of recurrent ventriculoperitoneal shunt related meningitis caused by an SCV-SA strain that was methicillin resistant (SCV-MRSA). The organism was isolated in pure culture of CSF and was characterized by molecular methods. Case report. A 69-year-old woman was hospitalized for hypertensive hydrocephalus, which had developed 2 months after a subarachnoid hemorrhage. A ventriculoperitoneal shunt was inserted. Seven days later, purulent material was observed in the surgical incision. Pus cultures grew typical S. aureus colonies on blood agar; CSF culture results were negative. Ciprofloxacin therapy (1 g/day) was administered for 10 days, and the patient was discharged from the hospital in good condition. Three months later, she was readmitted with headache, vomiting, fever (temperature, 39.5 C), and nuchal rigidity. A cloudy CSF sample obtained by shunt puncture had a leukocyte count of 1600 leukocytes/ml (80% polymorphonuclear cells); a protein level of 224 mg/dl; and a glucose level of 14 mg/dl (blood glucose level, 180 mg/dl). Cytocentrifuge Gram staining revealed the presence of gram-positive cocci resembling staphylococci. The shunt was replaced with an external CSF drain, and empiric therapy with vancomycin (2 g/day) was initiated to treat probable staphylococcal meningitis. Four days later, isolates cultured from the CSF samples and the shunt tip were definitively identified as SCV-MRSA. Defervescence occurred on day 9 after initiation of treatment; CSF findings were normal by day 18. After 4 weeks of vancomycin therapy, a new ventriculoperitoneal shunt was inserted. At that time, cultures of CSF and the shunt tip were negative. After 2 more weeks of treatment, the patient was discharged from the hospital in good condition. Twenty days later, she was readmitted with evidence of meningeal irritation, fever (temperature, 38.5 C), and abdominal pain. A CSF specimen obtained by shunt puncture had a protein level of 280 mg/dl, a glucose level of 25 mg/dl, and a leukocyte count of 1200 leukocytes/mm 3 (87% neutrophils). Cultures of a CSF specimen and the shunt tip, which had been promptly removed, again grew SCV-MRSA. After 8 weeks of treatment with vancomycin (2 g/day), ciprofloxacin (1 g/day), and rifampin (20 mg/kg/day), a new shunt was inserted, and treatment e48 CID 2005:41 (1 September)
was continued for 2 more weeks. One month after discontinuation of antibiotics, CSF cultures yielded no growth, and laboratory analysis of CSF revealed no abnormalities. At the month 12 follow-up visit, the patient was healthy, with no evidence of infection relapse. Microbiological investigation. Our routine protocol for testing CSF specimens from patients with CSF shunts includes aerobic culture (at 35 C) on MacConkey agar; microaerobic culture (at 35 C in air with 5% CO 2 ) on Columbia and chocolate agars; anaerobic culture (at 35 C) on Schaedler agar; and aerobic and anaerobic cultures on brain-heart infusion broth supplemented with 5% NaCl. After 48 h of incubation, cultures of CSF and the shunt tip on Columbia agar yielded pinpoint, nonpigmented, nonhemolytic colonies, which grew normally on Schaedler agar. The isolates were identified as methicillinsusceptible Staphylococcus caprae by our routine automated identification system (Phoenix; Becton Dickinson Microbiology Systems). This result was contradicted by a positive (though weak) coagulase reaction (clotting period, 118 h), which, together with the culture characteristics noted above, was suggestive of SCV-SA. The Phoenix report of methicillin susceptibility was confirmed by the oxacillin screening test [10]. The Epsilometer test (AB Biodisk) yielded the following MICs: oxacillin, 0.5 mg/l; rifampin, 0.03 mg/l; ciprofloxacin, 0.03 mg/ L; quinupristin-dalfopristin, 0.06 mg/l; linezolid, 0.5 mg/l; teicoplanin, 0.5 mg/l; vancomycin, 0.5 mg/l; and gentamicin, 132 mg/l. In light of the conflicting data, we immediately initiated realtime PCR assays with the LightCycler instrument and the LightCycler Staphylococcus M GRADE and MRSA Detection Kits (Roche Diagnostics). For species-level identification, staphylococcal DNA was amplified with specific primers derived from the internal transcribed space region of the S. aureus genome. Methicillin resistance was detected with primers that amplified a 315-bp fragment of the meca gene. The results confirmed that both isolates were MRSA (figure 1). Additional characterization. The SCV-MRSA isolates displayed auxotrophism for hemin [3]. They were misidentified by 2 other automated systems (both biomérieux): the Vitek 2 (which reported methicillin-susceptible Staphylococcus hycus) and the API ID32 Staph (which reported methicillin-susceptible Staphylococcus scheifleri). Subsequent PCR assays revealed 3 other S. aureus specific genes: gap (figure 2A) [11], nuc [12], and coa [13], and methicillin resistance was confirmed with the PCR method described by Geha et al. [14] (figure 2B). PFGE, performed as described elsewhere [15], revealed the clonal identity of all isolates recovered from the patient, including the one originally isolated from the surgical wound (data not shown). Discussion. The SCV-MRSA strain responsible for our patient s shunt-related meningitis was incorrectly identified by 3 automated systems widely used for routine identification of staphylococcal isolates. Similar experiences have been reported by other investigators [8]. The results can be attributed to the short incubation times used by these systems, as well as the low levels of discrimination of their databases. The key to successful recovery and identification of SCV-SA is the use of an extended battery of culture and identification techniques [6, 16]. In the present case, SCV-SA was first suspected because of the observation of slow-growing, pinpoint, nonpigmented, and nonhemolytic colonies on Columbia agar (incubated for 48hat35 C) together with normal growth on Schaedler agar (incubated for 48 h at 35 C with CO 2 ). If the latter medium had not been included in our routine protocol, the true identity of the isolates might have remained obscure (although doubts about the Phoenix report of S. caprae were also raised by the finding of weak but clearly positive coagulase activity). The SCV-MRSA strain was also erroneously classified as methicillin-susceptible by the 3 automated systems, by the oxacillin screening test (recommended by the NCCLS) [10], and by the widely used Epsilometer test [9]. False-negative results for methicillin resistance can occur with culture-based methods because meca is heterogeneously expressed by clinical isolates of S. aureus (even those that are metabolically normal). Moreover, the generation time for SCV-SA (6 9 times longer than that of normal S. aureus) can diminish the sensitivity of conventional methicillin-susceptibility testing [9]. The E-test and all 3 automated systems concordantly classified the isolates as susceptible to all agents tested, with the exception of gentamicin. We cannot exclude the possibility that 1 of those results were false negative, since they were not confirmed by genetic analysis, but the final treatment regimen (which seems to have been successful) included 3 of the drugs that displayed good in vitro activity against the strain. Our PFGE data suggests that the SCV-MRSA strain arose from the normal-phenotype parent strain isolated from the surgical wound. In vivo, the emergence of SCVs is strongly associated with antibiotic use, although isolation of SCVs has also been reported after prolonged antibiotic-free intervals [1, 4]. Our patient s wound infection had been treated with ciprofloxacin, and during her numerous hospitalizations, she received various antimicrobials, including gentamicin, for recurrent urinary tract infections. Although SCV-SA has been implicated in several antibioticrefractory, recurrent, and/or persistent infections [2 5], including a case of brain abscess [17], there have been no reports of CSF shunt related infection with SCVs. Infections are a major problem for patients with CSF shunts. When MRSA is the pathogen responsible, the standard treatment includes shunt removal and intravenous vancomycin therapy [18]. However, vancomycin s penetration of CSF is weak and sometimes unpredictable, and intrathecal administration and continuous intravenous administration have been proposed to overcome this CID 2005:41 (1 September) e49
Figure 1. Melting-curve analysis. PCR products were identified by melting-curve analysis, and the melting temperature of each of the products was compared with that of the positive control. All of the samples showed positive results for the Staphylococcus aureus specific gene (melting temperature, 62 C) (A), and the mec gene (melting temperature, 66.4 C) (B). The samples were as follows: 4 CSF specimens (blue, yellow-green, red, and black lines), 4 bacterial isolates (pink, olive green, blue, and gray lines), and positive control (violet line). The negative control is also indicated (brown line).
Figure 2. Fluorogram of the agarose gel electrophoresis for the staphylococcal gap gene (A) and the meca gene (B) of the small-colony variant (SCV) Staphylococcus aureus strain isolated from the patient s CSF sample. Lane SCV, SCV S. aureus isolate; lane M50, 50-bp ladder; lane M100, 100-bp ladder. limitation [18]. The duration of therapy is generally based on the clinical (and, ideally, microbiological) response [18, 19]. Closer follow-up and longer treatment may be necessary before shunt reinsertion is attempted. In fact, S. aureus CSF-shunt infection is a recognized risk factor for subsequent shunt infection, even when the strain is phenotypically normal [19]. As soon as Gram-stain data suggested the possibility of staphylococcal meningitis, empiric treatment with vancomycin was started because of the high frequency of infection with methicillin-resistant staphylococci in our hospital. Four days later, when the isolates were definitively identified as SCV-MRSA, the decision was made to continue vancomycin therapy, in light of the patient s clinical improvement and decreasing fever (which abated completely after 9 days of treatment) and the in vitro evidence of the isolates susceptibility to the drug. Nonetheless, cell wall active antimicrobials such as vancomycin often display reduced efficacy against these slow-growing organisms, particularly when they are bound to the surface of a foreign body [20]. Seifert et al. [5] suggest that SCVs play major roles in intravascular device related infections, and they emphasize the fact that complete eradication is highly unlikely unless these devices are removed. In spite of shunt removal, a prolonged 6-week course of vancomycin therapy, and excellent clinical and microbiological responses, our patient s infection recurred!1 month after her discharge. The optimal therapy (and duration) for SCV-MRSA infection has yet to be defined. Rifampin plus trimethoprim-sulfamethoxazole was the most effective regimen against intraendothelial cell SCVs in a tissue-culture system [6]. Rifampin plus vancomycin is a treatment option for MRSA CSF infections; others include trimethoprim-sulfamethoxazole and linezolid [18]. Linezolid and new drugs such as quinupristin-dalfopristin or daptomycin are active against normal MRSA [18], but their effectiveness against SCVs has not been demonstrated. Our patient s second episode of meningitis was treated with a tripledrug regimen (vancomycin, rifampin, and ciprofloxacin), and the treatment duration was double that of the previous course of treatment (8 weeks prior to shunt reinsertion and 2 weeks thereafter). The clinical and microbiological responses were excellent, and 11 year passed with no signs of infection, but close follow-up is still essential because SCV-SA infections have been associated with recurrence after intervals as long as 54 years [6, 7]. With the increasing use of invasive medical devices, SCV-SA infections are likely to become more common. Failure to identify these agents can be particularly disastrous for patients with CID 2005:41 (1 September) e51
CNS infection. In cases of this type, particularly when the infection seems to be recurrent, SCVs should be suspected whenever pinpoint colonies are noted on routine microaerobic cultures, even in the presence of normal S. aureus colonies. These pathogens must then be actively sought by use of appropriate selective media and growth conditions [16]. Our experience highlights the low reliability of automated systems and other conventional methods for detection of methicillin resistance in these strains. Species-level identification and meca detection based on real-time PCR are simple procedures that provide reliable results in 4 h. The time saved by use of the PCR-based method can have both clinical and financial benefits in the management of SCV-SA infection. Acknowledgments We thank Marian Kent for editorial assistance. Financial support. This study was financed in part by the Italian Ministry for Universities and Scientific Research (grant ex MURST 2003 2004). Potential conflicts of interest. All authors: no conflicts. References 1. Proctor RA, Bates DM, McNamara PJ. Electron transport-deficient Staphylococcus aureus small-colony variants as emerging pathogens. In: Scheld WM, Craig WA, Hughes JM, eds. Emerging infections 5. Washington, DC: ASM Press, 2001:95 110. 2. Proctor RA, van Langevelde P, Kristjansson M, Maslow JN, Arbeit RD. Persistent and relapsing infections associated with small-colony variants of Staphylococcus aureus. Clin Infect Dis 1995; 20:95 102. 3. Kahl B, Herrmann M, Everding AS, et al. Persistent infection with small colony variant strains of Staphylococcus aureus in patients with cystic fibrosis. J Infect Dis 1998; 177:1023 9. 4. Von Eiff C, Bettin D, Proctor RA, et al. Recovery of small colony variants of Staphylococcus aureus following gentamicin bead placement for osteomyelitis. Clin Infect Dis 1997; 25:1250 1. 5. Seifert H, Wisplinghoff H, Schnabel P, von Eiff C. Small colony variants (SCVs) of Staphylococcus aureus as a cause of pacemaker-related infection. Emerg Infect Dis 2003; 9:1316 8. 6. Proctor RA, Peters G. Small colony variants in staphylococcal infections: diagnostic and therapeutic implications. Clin Infect Dis 1998; 27:419 22. 7. Proctor RA, Kahl B, von Eiff C, Vaudaux PE, Lew DP, Peters G. Staphylococcal small colony variants have novel mechanisms for antibiotic resistance. Clin Infect Dis 1998; 27(Suppl 1):S68 74. 8. Seifert H, von Eiff C, Fätkenheuer G. Fatal case due to methicillinresistant Staphylococcus aureus small colony variants in an AIDS patient. Emerg Infect Dis 1999; 5:450 3. 9. Kipp F, Becker K, Peters G, von Eiff C. Evaluation of different methods to detect methicillin resistance in small-colony variants of Staphylococcus aureus. J Clin Microbiol 2004; 42:1277 9. 10. NCCLS. Performance standards for antimicrobial susceptibility testing: 14th informational supplement. NCCLS document M100-S14. Wayne, PA: NCCLS, 2004. 11. Yugueros J, Temprano A, Berzal B, et al. Glyceraldehyde-3-phospate dehydrogenase-encoding gene as a useful taxonomic tool for Staphylococcus spp. J Clin Microbiol 2000; 38:4351 5. 12. Brakstad OG, Aasbakk K, Maeland JA. Detection of Staphylococcus aureus by polymerase chain reaction amplification of the nuc gene. J Clin Microbiol 1992; 30:1654 60. 13. Goh SH, Byrne SK, Zhang JL, Chow AW. Molecular typing of Staphylococcus aureus on the basis of coagulase gene polymorphisms. J Clin Microbiol 1992; 30:1642 5. 14. Geha DJ, Uhl JR, Gustaferro CA, Persing DH. Multiplex PCR for identification of methicillin-resistant staphylococci in the clinical laboratory. J Clin Microbiol 1994; 32:1768 72. 15. McDougal LK, Steward CD, Killgore GE, Chaitram JM, McAllister SK, Tenover FC. Pulsed-field gel electrophoresis typing of oxacillin-resistant Staphylococcus aureus isolates from the United States: establishing a national database. J Clin Microbiol 2003; 41:5113 20. 16. Kipp F, Kahl BC, Becker K, et al. Evaluation of two chromogenic agar media for recovery and identification of Staphylococcus aureus smallcolony variants. J Clin Microbiol 2005; 43:1956 9. 17. Kipp F, Ziebuhr W, Becker K, et al. Detection of Staphylococcus aureus by 16S rrna directed in situ hybridisation in a patient with a brain abscess caused by small-colony variants. J Neurol Neurosurg Psychiatry 2003; 74:1000 2. 18. Tunkel AR, Hartman BJ, Kaplan SL, et al. Practice guidelines for the management of bacterial meningitis. Clin Infect Dis 2004; 39:1267 84. 19. McGirt M, Zaas A, Fuchs HE, George TM, Kaye K, Sexton DJ. Risk factors for pediatric ventriculoperitoneal shunt infection and predictors of infectious pathogens. Clin Infect Dis 2003; 36:858 62. 20. Chuard C, Vaudaux PE, Proctor RA, Lew DP. Decreased susceptibility to antibiotic killing of a stable small colony variant of Staphylococcus aureus in fluid phase and on fibronectin-coated surfaces. J Antimicrob Chemother 1997; 39:603 8. e52 CID 2005:41 (1 September)