Persistent MRSA Bacteremia in a Patient with Low Linezolid Levels

SIR—We read with interest the report by Ruiz et al. [1] of 2 cases of endocarditis caused by methicillin-resistant Staphylococcus aureus (MRSA) that failed to respond to therapy with intravenous linezolid. Failure of therapy due to the development of linezolid resistance has been reported by Tsiodras et al. [2], but the 2 patients described by Ruiz et al. [1] had isolates that remained susceptible to linezolid. At the 2001 Infectious Diseases Society of America annual meeting, we described a patient who had persistent MRSA bacteremia and low linezolid blood levels during both oral and intravenous linezolid therapy [3]. A 49-year-old man, who weighed 85 kg and had a history of insulin-dependent diabetes mellitus, coronary artery disease, hepatitis C, and intravenous drug abuse, and who was receiving oral methadone at the time of admission, initially had cellulitis of the toe, with drainage that grew MRSA on culture. He developed a femoral pseudoaneurysm infected with MRSA, which required an axillopopliteal bypass graft and subsequent revision with a ring Gore-Tex graft (W. L. Gore & Associates). Because of a rash due to vancomycin, treatment with linezolid, 600 mg po q12h, was initiated. After 2 weeks, the patient developed renal insufficiency that required dialysis. (Subsequent antibiotic doses were not administered immediately prior to the start of dialysis.) Therapy was changed to quinupristin/dalfopristin, 700 mg iv q12h, to complete a 4-week total treatment course. Two weeks after completion of the first quinupristin/dalfopristin course, the patient developed tenderness at the dialysis catheter site and purulence at the pseudoaneurysm site. Cultures of blood samples, fluid from the pseudoaneurysm site, and the dialysis catheter were positive for MRSA. Treatment with quinupristin/dalfopristin, 700 mg iv q8h, was restarted. Blood cultures were positive for MRSA at day 4 but negative at day 7 of the quinupristin/dalfopristin course. Eight days after quinupristin/dalfopristin therapy was restarted, a new rash appeared, and quinupristin/dalfopristin was changed to linezolid, 600 mg po q12h, to complete a 3-week total treatment course. A transesophageal echocardiogram (TEE) revealed no vegetations. A renal biopsy revealed crescentic glomerulonephritis. Methylprednisolone succinate therapy was started, and, 2 weeks later, blood cultures were again positive for MRSA, and the dialysis catheter site was tender. Linezolid therapy, 600 mg iv q12h, was initiated. Both the catheter tip and specimens from the catheter site grew MRSA on culture. Because of poor venous access, the linezolid dosage was changed to 600 mg po q12h. Blood cultures were persistently positive, despite catheter removal and 3 weeks of oral linezolid therapy. A second TEE revealed no vegetations. Radiographic evaluation for a graft infection by CT scan and WBC-labeled nuclear medicine scan yielded inconclusive results. The patient and surgeon declined surgical intervention. Susceptibility testing confirmed that the isolate was sensitive to linezolid (MIC, <4 µg/mL; diffusion disk inhibition zone diameter, >21 mm). A Schlichter test performed during oral linezolid therapy at the time of peak drug levels revealed the lack of bacteriostatic or bactericidal activity at a serum dilution of 1 : 2. Because of persistent bacteremia, linezolid levels were measured by high-pressure liquid chromatography assay (see table 1). Linezolid blood levels were also measured subsequently, while the patient was receiving linezolid at a dosage of 600 mg iv q12h and when the dosage was increased to 900 mg iv q12h. The patient was eventually treated with linezolid, 900 mg iv q12h, for 5 weeks, with resolution of bacteremia. Linezolid blood levels achieved at various dosages in a patient treated with linezolid every 12 hours. In this patient, a vascular graft infection may have led to persistent MRSA bacteremia. Since the MIC90 of linezolid for S. aureus is usually in the range of 2–4 µg/mL (the MIC of linezolid for the isolate obtained from the patient was <4 µg/mL by disk diffusion), the low levels achieved in the patient during both oral and intravenous linezolid therapy at a dosage of 600 mg q12h may have contributed to his persistent bacteremia. It is unclear why the linezolid levels were low. Dialysis may remove ∼30% of a dose, but for this patient the drug was not administered immediately prior to the start of dialysis. Low linezolid levels such as those we observed, although uncommon, may have contributed to the results described by Ruiz et al. [1]. In patients who do not respond to linezolid therapy, it may be prudent to check linezolid levels. At this time, we would not recommend use of higher dosages or decreased dosage intervals without further study.