Breakthrough bacteraemia with a susceptible Enterococcus faecalis during tigecycline monotherapy

Sir, Despite its efficacy in abdominal1 and soft tissue infections2 and community-acquired pneumonia,3 there have been concerns about tigecycline use in bacteraemia and endocarditis.4 These relate to the low serum levels that occur due to its large volume of distribution, and its bacteriostatic action against some bacteria, including Enterococcus species. Following the case we now report, we reviewed the literature on tigecycline and bacteraemia. Our patient was middle aged, with a history of type 2 diabetes mellitus, long-term low-dose steroid use for polymyalgia rheumatica and had undergone mitral valve repair and coronary artery bypass grafting 2 years previously. Following surgery, Enterococcus faecalis bacteraemia developed, secondary to a peripheral venous catheter infection. This was treated, successfully, with 4 weeks of linezolid. The patient had also had a skin infection of the hand (without positive microbiological tests) treated 2 months before this presentation. There was a history of allergies to penicillin (severe rash) and to the glycopeptide teicoplanin, administration of which had previously been associated with dyspnoea. Presentation was with a 6 week history of back pain. Systemic inflammatory response syndrome criteria were absent and there was no cardiac murmur. C-reactive protein (CRP), however, was elevated at 161 mg/L. Three blood cultures, on sequential days, grew E. faecalis (speciation by PCR) susceptible to amoxicillin, vancomycin, linezolid and tigecycline (MIC 0.064 mg/L repeatedly by local Etest, 0.25 mg/L by reference laboratory Etest), but resistant to rifampicin and tetracycline, and highly resistant to gentamicin. Transthoracic echocardiogram suggested an aortic vegetation; imaging showed discitis at the L5/S1 level. The patient initially refused a trans-oesophageal echocardiogram (TOE). Penicillin desensitization and closely monitored therapy with amoxicillin or vancomycin were considered, but after discussion the patient refused these options. At that time, evidence for daptomycin use in enterococcal endocarditis was lacking and daptomycin was not available at our hospital. Linezolid 600 mg intravenously (iv) every 12 h was therefore commenced to good clinical and biochemical (CRP) effect, but this had to be stopped after 4 weeks because of intractable nausea, despite antiemetics. Tigecycline 50 mg iv every 12 h was commenced, with a further initial fall in CRP. In the fourth week, however, the patient developed rigors. Multiple blood cultures grew strains of tigecycline-susceptible E. faecalis (MIC of 0.25 mg/L by agar dilution at the reference laboratory), the antibiotic susceptibility pattern of all isolates being identical to those of initial isolates. Vancomycin iv was prescribed, but marked deterioration in renal function resulted. TOE confirmed mitral and aortic valve endocarditis. The patient was unfit for valve replacement surgery and therefore received amoxicillin 2 g iv every 4 h. Despite the past history of a severe rash with penicillin, this was well tolerated. The patient subsequently developed Clostridium difficile-associated diarrhoea and cardiac failure, however, and died as a result. This case is in keeping with reports that suggest that tigecycline should be used cautiously in bacteraemia and endocarditis, even when the causal pathogen is susceptible in vitro. The phenomenon of breakthrough bacteraemia, despite a relatively low MIC, is likely to occur because of the pharmacokinetic profile of tigecycline; it has excellent tissue penetration, but its large volume of distribution (9 L/kg) and high protein binding (71%) result in low serum levels.5 Curcio4 suggested that tigecycline monotherapy be used for Gram-positive bacteraemia only when the MIC is ≤0.5 mg/L (the steady-state serum concentration of tigecycline is 0.6–0.7 mg/L with 50 mg every 12 h). Caution is recommended by the manufacturer when treating bacteraemia as few cases have occurred in trials.5 To our knowledge, there have been no reports of successful treatment using tigecycline alone for endocarditis. Successful therapy of enterococcal endocarditis in combination with daptomycin has been reported.6,7 Regarding specific use for Enterococcus, tigecycline has shown excellent results in vitro, including against vancomycin-resistant species, and in rat and rabbit models of enterococcal endocarditis.8 In the study by Florescu et al.9 that specifically set out to recruit patients with vancomycin-resistant enterococcal infections, there was only one clinically evaluable patient with bacteraemia treated with tigecycline, who was cured. We therefore performed a PubMed literature search using the keywords: tigecycline; bacteraemia; sepsis; and endocarditis. We identified 114 additional patients with all-cause bacteraemia treated with tigecycline, of which 78 were treated with monotherapy (Table 1). In these 78 patients, cure was achieved in 64 patients (82%). In 24 patients with Gram-positive bacteraemias (one enterococcal) treated with monotherapy, 22 (92%) achieved cure. Although this may appear to be an adequate cure rate, there is a considerable likelihood of reporting bias. Experience of tigecycline use in all-cause bacteraemia and endocarditis MDR, multidrug resistant; SOT, solid organ transplant; MS, multiple sclerosis; SLE, systemic lupus erythematosus; ESBL+, extended-spectrum β-lactamase positive; CABG, coronary artery bypass graft; DM, diabetes mellitus; MI, myocardial infarction; IPPV, already on mechanical ventilation. Experience of tigecycline use in all-cause bacteraemia and endocarditis MDR, multidrug resistant; SOT, solid organ transplant; MS, multiple sclerosis; SLE, systemic lupus erythematosus; ESBL+, extended-spectrum β-lactamase positive; CABG, coronary artery bypass graft; DM, diabetes mellitus; MI, myocardial infarction; IPPV, already on mechanical ventilation. In conclusion, there were no reports of successful treatment in vivo for infective endocarditis using tigecycline monotherapy, and clinical experience in bacteraemia is limited. Although our literature review suggests that tigecycline monotherapy can be used successfully in selected cases of Gram-positive and Gram-negative bacteraemia, the presented case demonstrates that physicians should only use this approach with caution when other reasonable therapeutic options are exhausted. There were no external or internal sources of specific funding for this paper. None to declare. We acknowledge: Mrs Polly Kaufmann (Laboratory of HealthCare Associated Infection, Centre for Infections, Health Protection Agency, 61 Colindale Avenue, London, UK) for the species-specific PCR result; and Dr Robert Hill and Dr Neil Woodford (Antibiotic Resistance Monitoring and Reference Laboratory, Health Protection Agency, 61 Colindale Avenue, London, UK) for details of the method used for tigecycline MIC estimation.