Skip Navigation

Salvage Treatment for Persistent Methicillin-Resistant Staphylococcus aureus Bacteremia: Efficacy of Linezolid With or Without Carbapenem

  1. Hee-Chang Jang1,
  2. Sung-Han Kim1,a,
  3. Kye Hyoung Kim1,
  4. Choong Jong Kim1,
  5. Shinwon Lee1,
  6. Kyoung-Ho Song1,
  7. Jae Hyun Jeon1,
  8. Wan Beom Park1,
  9. Hong Bin Kim1,
  10. Sang-Won Park1,
  11. Nam Joong Kim1,
  12. Eui-Chong Kim2,
  13. Myoung-don Oh1, and
  14. Kang Won Choe1
  1. 1Departments of Internal Medicine and Departments of Laboratory Medicine, Seoul National University College of Medicine, Seoul, Republic of Korea
  2. 2Departments of Laboratory Medicine, Seoul National University College of Medicine, Seoul, Republic of Korea
  1. Reprints or correspondence: Dr. Myoung-don Oh, Department of Internal Medicine, Seoul National University College of Medicine, 28 Yongun-dong, Chongro-gu, Seoul, 110–744, Republic of Korea (mdohmd{at}snu.ac.kr).
  1. Presented in part: 48th Annual Interscience Conference on Antimicrobial Agents and Chemotherapy/Infectious Diseases Society of America 46th Annual Meeting, Washington, DC, 25–28 October 2008 (abstract K-3485).

  • a Present affiliation: Department of Internal Medicine, University of Ulsan College of Medicine, Seoul, Republic of Korea.

Abstract

Background. Persistent methicillin-resistant Staphylococcus aureus (MRSA) bacteremia is associated with high mortality rates, but no treatment strategy has yet been established. We performed this study to evaluate the efficacy of linezolid with or without carbapenem in salvage treatment for persistent MRSA bacteremia.

Methods. All adult patients with persistent MRSA bacteremia for ⩾7 days from January 2006 through March 2008 who were treated at Seoul National University Hospital were studied. The results of linezolid salvage therapy with or without carbapenem were compared with those of salvage therapy with vancomycin plus aminoglycosides or rifampicin.

Results. Thirty-five patients with persistent MRSA bacteremia were studied. The early microbiological response (ie, negative results for follow-up blood culture within 72 hours) was significantly higher in the linezolid-based salvage therapy group than the comparison group (75% vs 17%; P=.006). Adding aminoglycosides or rifampicin to vancomycin was not successful in treating any of the patients, whereas linezolid-based therapy gave an 88% salvage success rate (P<.001). The S. aureus -related mortality rate was lower for patients treated with a linezolid salvage regimen than for patients continually treated with a vancomycin-based regimen (13% vs 53%; P=.030).

Conclusions. Linezolid-based salvage therapy effectively eradicated S. aureus from the blood for patients with persistent MRSA bacteremia. The salvage success rate was higher for linezolid therapy than for vancomycin-based combination therapy.

Staphylococcus aureus is a major cause of community-acquired and nosocomial infections [1]. The organism is the second most common cause of nosocomial bacteremia in the United States [2]. Glycopeptides have been used as the main treatment for serious methicillin-resistant S. aureus (MRSA) infections, including bacteremia. However, treatment failure in MRSA infections has increased and has become a major clinical problem [35].

Persistent S. aureus bacteremia raises concern because it occurs despite the administration of appropriate antibiotics shown by laboratory tests to be active in vitro. The microbiological and clinical characteristics of persistent S. aureus bacteremia have been studied [69], but the mechanism of persistence is still poorly understood, and no clinical indicator is available to predict glycopeptide treatment failure. Moreover, no treatment strategy has been established, despite the fact that failure is common and frequently leads to death [8, 9]. The approaches of infectious diseases consultants to treating persistent MRSA bacteremia have been examined [10], but no consensus has emerged in relation to antimicrobial therapy because few clinical data are available.

Linezolid, an oxazolidinone, is used to treat infections caused by resistant gram-positive pathogens. Its efficacy in treating skin and soft-tissue infections, pneumonia, and uncomplicated MRSA bacteremia is not inferior to that of vancomycin [11]. However, its efficacy in treating endocarditis or complicated or persistent bacteremia has not been clearly defined. Recent studies have shown that the combined use of linezolid and carbapenem has a synergistic bactericidal effect on S. aureus in vitro and in an animal endocarditis model [12, 13], but no clinical data are available. We performed this study to investigate the incidence and clinical features of persistent S. aureus bacteremia and to estimate the efficacy of linezolid with or without carbapenem for the salvage treatment of persistent S. aureus bacteremia.

Patients and Methods

Patients. All patients ⩾16 years old with persistent S. aureus bacteremia from 1 January 2006 through 31 March 2008 were studied at Seoul National University Hospital (Republic of Korea), a 1500-bed tertiary care university hospital and referral center. Patients were identified by reviewing the computerized records of the Clinical Microbiology Laboratory. Only the first episode of S. aureus bacteremia in a patient was included in the study.

Microbiological tests. S. aureus was identified and antibiotic resistance was determined with automated systems (Vitek 2, bioMérieux; and Microscan, Dade Behring). Susceptibility to arbekacin was determined by the disk diffusion method with 30-µg arbekacin disks (Eiken Chemical) and a breakpoint of ⩾18 mm. In cases of MRSA, the minimum inhibitory concentration (MIC) of vancomycin was determined by microdilution (BBL Mueller Hinton II Broth [cation adjusted]; BD Diagnostics) in accordance with the recommendations of the Clinical and Laboratory Standards Institute [14]. Screening tests for hetero-vancomycin-intermediate S. aureus (hetero-VISA) were performed using brain heart infusion agar plates containing 4 µg/mL of vancomycin [15].

Definitions. Persistent bacteremia was defined as the isolation of S. aureus in blood cultures obtained from peripheral veins on ⩾7 consecutive days despite appropriate antibiotic administration for ⩾5 days. S. aureus bacteremia was defined as community associated or as health care associated in accordance with Centers for Disease Control and Prevention definitions [16]. S. aureus infection was defined as catheter related if the catheter tip yielded >15 colonies for S. aureus or inflammation was present at the insertion site and no alternative source of infection was identified [17]. Endocarditis was defined by the modified Duke criteria [18]. Metastatic infection was defined as the presence of microbiological or radiographic evidence of S. aureus infection caused by hematogenous seeding [19]. Complicated infection was defined as a site of infection remote from the primary focus caused by hematogenous seeding, including endocarditis, or extension of the infection beyond the primary focus (eg, septic thrombophlebitis or abscess) [20]. Eradicable foci included surgically removable infections or drainable abscesses and indwelling foreign bodies, such as peripheral and central venous catheters. Noneradicable foci included unknown primary sites, pneumonia, endocarditis, and osteomyelitis or arthritis. Noneradicated foci comprised noneradicable foci and eradicable foci not actually eradicated [21]. Thirty-day mortality was defined as the death of a patient without clearance of bacteremia or within 30 days after clearance of bacteremia. Mortality was defined as S. aureus -related if there was no other definite cause of death.

Salvage attempt and outcome measure. Salvage attempt was defined as administration of a new antibiotic, active against the S. aureus isolate in laboratory susceptibility tests, for ⩾72 hours as a substitute or supplement. Linezolid was administered intravenously at a dosage of 600 mg every 12 hours as salvage therapy. The efficacy of salvage treatment was evaluated by 2 outcome measures: early microbiological response and salvage success. Early microbiological response was defined as conversion of positive blood culture results to negative within 72 hours of antibiotic initiation. Salvage was defined as successful if the agent used was not subsequently changed because of ineffectiveness and S. aureus -related death did not occur.

Statistical analysis. Categorical variables were compared using the Fisher exact test or Pearson χ2test, as appropriate, and continuous variables were compared using the Mann-Whitney U test. All tests of significance were 2-tailed, and P⩽.05 was considered statistically significant. Statistical analyses were performed with SPSS, version 12.0 (SPSS).

Results

Prevalence of persistent S. aureus bacteremia. We identified 377 cases of S. aureus bacteremia during the study period. Of the 377 cases of S. aureus bacteremia, 41 (11%) were persistent despite administration of appropriate antibiotic. Thirty-five (17%) of the 211 cases of MRSA bacteremia and 6 (4%) of the 166 cases of methicillin-sensitive S. aureus (MSSA) bacteremia were persistent (P<.001). The duration of persistence was 7–10 days (median, 7 days; mean, 7.6 days) for the cases of MSSA bacteremia and 7–168 days (median, 12 days; mean, 18.1 days) for the cases of MRSA bacteremia (P=.009). All the patients with persistent MSSA bacteremia were successfully treated using nafcillin-penicillin with or without aminoglycoside (plus rifampicin in the case of prosthetic valve endocarditis), so salvage attempt was not applied for them.

Clinical characteristics of persistent MRSA bacteremia. Of the 35 cases of MRSA bacteremia, 34 were health care associated (30 nosocomial) and 1 was community associated. The clinical features of the persistent cases of MRSA bacteremia are given in tables 1and 2. The mean age of the patients was 59 years in the vancomycin-continued group and 70 years in the linezolid salvage group (P=.01). For the 35 patients, diabetes mellitus (in 12 patients [34%]), end-stage renal disease (in 10 [29%]), solid cancers (in 10 [29%]), and the presence of vascular grafts (in 11 [31%]) were common as underlying disease. Endovascular infection (in 19 patients [54%]), especially of a central catheter (in 15 [43%]), was the most common primary site of infection. Twenty-eight (80%) of the cases were complicated S. aureus infections. Infected venous thrombus and abscess were detected in 7 (20%) and 14 (40%) of the cases, respectively. Transthoracic echocardiography and transesophageal echocardiography were performed in 30 (86%) and 13 (37%) of the patients with MRSA bacteremia, respectively. Endocarditis was detected in 6 patients (17%). Metastatic foci were observed in 20 patients (57%), and common metastatic foci were endovascular structures (endocardium or vascular grafts) (in 9 patients), spine (in 8), brain (in 7), lung (in 5), and pleura (in 2). Although adjunct therapy to remove foci was performed in 25 patients (71%), noneradicated foci remained in 30 patients (86%). Underlying diseases, primary site of infection, rates of complicated and metastatic infection, and noneradicated foci were not significantly different in the vancomycin-continued and linezolid salvage groups.

Table 1

Clinical Features and Outcomes of Persistent Methicillin-Resistant Staphylococcus aureus Bacteremia in 19 Patients Treated with Vancomycin-Based Regimens

Table 2

Clinical Features and the Consequences of Linezolid-Based Salvage Therapy in 16 Patients with Persistent Methicillin-Resistant Staphylococcus aureus Bacteremia

Among the 35 MRSA isolates causing persistent bacteremia, the vancomycin MIC was 2 µg/mL in 3 isolates. No VISA was observed. Four isolates (11%) showed growth on brain heart infusion agar with 4 µg/mL of vancomycin. Resistance to gentamicin and rifampicin was observed in 25 (71%) and 3 (9%) of the cases, respectively. All the isolates were susceptible to linezolid and arbekacin. MRSA bacteremia was persistent, although the serum vancomycin trough concentration exceeded 10 µg/mL in 31 (91%) and 15 µg/mL in 19 (56%) of the 34 cases evaluated in our study.

Salvage antibiotic attempts and outcomes for patients with persistent MRSA bacteremia. The salvage attempts against persistent MRSA bacteremia and their consequences are summarized in tables 3and 4. Early microbiological responses and salvage success rates were significantly higher with the linezolid-based regimen than with the comparators (P=.006 and P<.001, respectively). Adding aminoglycosides or rifampicin to vancomycin was not successful in any case. However, linezolid-based therapy had a salvage success rate of 88%. A linezolid-based regimen was introduced for 6 patients who had treatment failure despite the addition of aminoglycosides or rifampicin. One of these patients died without a microbiological response, but 4 of the others became culture negative within 72 hours and the remaining patient within a week. These 5 patients were successfully treated.

Table 3

Results of 28 Salvage Attempts for 35 Patients with Persistent Methicillin-Resistant Staphylococcus aureus Bacteremia

Table 4

Mortality Rates in 35 Patients with Persistent Methicillin-Resistant Staphylococcus aureus Bacteremia

The 30-day mortality rate was 40% (14 patients) and the S. aureus -related mortality rate was 34% (12 patients) for the 35 patients with persistent MRSA bacteremia. The S. aureus -related mortality rate was significantly lower for patients who were treated with a linezolid salvage regimen than for patients who were continuously treated with a vancomycin-based regimen (53% vs 13%; P=.03; table 4).

Adverse reactions to linezolid-based therapy. Seven (58%) of 12 evaluable patients had linezolid-associated thrombocytopenia during use of the antibiotic. Thrombocytopenia developed 7–21 days after the initiation of antibiotic treatment. Linezolid-based regimens were changed to vancomycin with or without rifampicin in 7 patients. Of these, 5 were successfully treated. However, recurrence of bacteremia was observed in 2 patients, and they were successfully treated by readministration of linezolid with or without carbapenem.

Discussion

In the present study, we found that linezolid-based salvage therapy was effective in eradicating S. aureus from the blood within 72 hours for patients with persistent MRSA bacteremia. We also showed that the salvage success rate was higher with linezolid-based therapy than with vancomycin-based combination therapy. The rate of persistence of S. aureus bacteremia was 11% during the study period in our hospital. This figure had not increased: it was 13% (31 of 238 patients with S. aureus bacteremia) in our previous investigation from January 1998 to October 2001 [22]. The prevalence in our hospital was similar to that in another center, with 11% of persistent bacteremia for ⩾7 days [9]. Persistence was more common in MRSA bacteremia than in MSSA bacteremia, which agrees with previous studies [8, 9].

Some investigators have reported cases of persistent bacteremia caused by VISA or hetero-VISA [23, 24]. However, we found no VISA, and only 4 isolates were positive on the hetero-VISA screening test. In addition, only a few isolates had a vancomycin MIC of 2 µg/mL. Similar findings were obtained in other studies [7, 8]. Adequate vancomycin serum concentration is important to eradicate MRSA from infected sites. However, MRSA bacteremia persisted even though serum vancomycin trough concentrations exceeded 10 µg/mL in most of the patients evaluated in our study. Some authors have proposed that it is important to achieve serum vancomycin trough concentrations of >15 µg/mL to achieve better treatment outcomes in MRSA infections [25]. However, 56% of our patients had persistent bacteremia despite the fact that their serum vancomycin trough concentrations were >15 µg/mL.

Approximately 70% of infectious diseases consultants in the United States reported that they preferred to continue administration of vancomycin and add aminoglycoside or rifampicin for persistent MRSA bacteremia if the vancomycin MIC did not exceed 2 µg/mL [10]. We also used these agents in addition to vancomycin for patients with persistent MRSA bacteremia that did not response to vancomycin monotherapy. Of the aminoglycosides, gentamicin was used for 2 patients, whereas arbekacin was used for the patients for whom the isolates were resistant to gentamicin but susceptible to arbekacin [26]. However, these attempts were unsuccessful: combination therapy was ineffective for most of the patients for whom vancomycin alone had failed to eradicate S. aureus from the blood.

The success rate of salvage attempts with linezolid-based regimens was 88% in the present study, and similar success rates have been reported in previous studies [24, 27]. Although the number of cases was small, linezolid seemed to be effective as salvage therapy in our study, even though the regimen was introduced for patients for whom vancomycin-based combination therapy had already failed or who had had persistent MRSA bacteremia for a prolonged period or who had multiple metastatic sites. We chose ertapenem for combination therapy in our study because it has the narrowest spectrum among carbapenems.

Various new antibiotics could be considered candidates for the treatment of persistent MRSA bacteremia [28]. However, good penetration of the antibiotic into the tissues of the body, especially lung, bone, and brain, and if possible into biofilms, abscesses, thrombi, and cardiac vegetations is essential, because devices, abscesses, thrombi, endocarditis, and metastatic foci at various sites were shown in our study and in others [8, 9] to be implicated in bacteremic persistence. We tried linezolid and carbapenem, which synergize, because both have good tissue-penetration properties.

Although the efficacy of linezolid-based regimens was good, thrombocytopenia limited their prolonged use for the ⩾4 weeks that are needed to treat complicated S. aureus bacteremia [1]. The incidence of thrombocytopenia in our study was high, probably because of the high proportion of serious underlying diseases and end-stage renal disease for patients with persistent bacteremia [29]. Although the duration of linezolid therapy was shortened by thrombocytopenia, the subsequent use of vancomycin for a total of 4–6 weeks after negative blood culture results were achieved was successful.

The present study has methodological limitations because it was retrospective in design. Hence, regimen, starting point, and duration of salvage therapy were not uniform. Because the factors influencing the physicians' choice of antibiotics were not determined, they may have influenced our results as unmeasured confounding factors in the analysis. Also, the sample size was limited and too small to distinguish between the effectiveness of linezolid-carbapenem combination therapy and linezolid monotherapy. The prevalence of endocarditis was possibly underestimated because transesophageal echocardiography was performed in only 37% of the patients. Despite such limitations, our data provide some valuable information that can affect treatment strategy. No clinical data were previously available on the outcome of salvage therapy that compared the use of alternative agents with the addition of aminoglycosides or rifampicin to vancomycin for patients with persistent MRSA bacteremia. Moreover, the clinical outcome of combination treatment with linezolid and carbapenem for patients with persistent MRSA bacteremia was defined for the first time.

In summary, for treating persistent MRSA bacteremia, the substitution of alternative agents for vancomycin appeared to be preferable to the addition of rifampicin or aminoglycoside to vancomycin, even if the isolate involved has been reported by current laboratory tests to be susceptible to vancomycin. Linezolid with or without carbapenem can be an effective salvage option, producing better outcomes for patients with persistent MRSA bacteremia.

Acknowledgments

Potential conflicts of interest. All authors: no conflicts.

  • Received October 21, 2008.
  • Accepted March 18, 2009.

References

| Table of Contents