Re: Which antibiotic for hospital acquired pneumonia caused by MRSA?
We are disappointed to see the clinically irrelevant meta-analysis by Kalil et al published recently in BMJ Open1 and endorsed in this corresponding BMJ editorial.2 When comparing the relatively narrow spectrum antibiotics vancomycin and linezolid, the only relevant pathogen for hospital-acquired pneumonia (HAP) is methicillin-resistant Staphylococcus aureus (MRSA). Multiple other antibiotics have activity against the other Gram-positive pathogens in the spectrum of both antibiotics, including methicillin-sensitive S. aureus. In the era when antibiotic stewardship is increasingly important, neither agent should be used unless MRSA HAP is a concern. Therefore, Dr. Kalil and associates start with an erroneous rationale for their analysis by focusing on Gram-positive HAP rather than specifically on MRSA HAP.
To compound this initial error, Dr. Kalil and co-authors insist on the primary analysis being their designation of intent-to-treat populations despite up to 72% of patients in these cohorts having pneumonia with pathogens other than MRSA.3 A basic tenet of any comparison of two treatments is that they are studied in the disease for which they are intended. We assume Drs. Muscedere, Kalil and collaborators would agree that the effect of vancomycin or linezolid on outcome in Pseudomonas or Acinetobacter HAP is of questionable clinical relevance. Yet this in essence is what their conclusions based meta-analysis of the intention-to-treat population would have us believe.
Even more dubious is their inclusion of patients who don’t have pneumonia in their intention-to-treat analyses. The most egregious examples are the inclusion of the studies of Jaksic et al,4 who specifically studied febrile neutropenia and only listed pneumonia as the site of infection in 50/605 (8.3%) of patients randomized, and Kaplan et al in which only 39/321 (12.3%) of pediatric patients had pneumonia.5 Since the non-pneumonia patients in these two studies alone make up more than a quarter of the patients included in the intention-to-treat population, it is misleading to call this a meta-analysis of HAP.
Even their per protocol analysis is flawed and the statement in their conclusion that “the clinical response in the per-protocol patients with MRSA pneumonia likewise did not show differences between drugs” is a factual error. Per protocol cohorts included HAP due to pathogens other than MRSA in all but three of the studies.
If they had analyzed clinical response for only documented MRSA pneumonia, the results are seen in the Figure. Results of the MRSA cohorts for the two registration trials6 7 were available from a separate analysis.8 The study of Jaksic et al4 was excluded since only 9 total patients had MRSA and no information was available regarding whether they had HAP or other sites of infection. In this meta-analysis of almost exclusively MRSA VAP,3 5 8-12 linezolid has a statistically superior clinical success rate compared to vancomycin, with little heterogeneity. Mortality specifically from MRSA HAP could only be determined for 4 trials.3 8 10 Mortality was not statistically different, although the largest influence on the findings was the phase 4 trial,3 which occurred after linezolid was available for salvage therapy of vancomycin failures.13
The clinical superiority of linezolid in meta-analysis is not surprising, given that the largest series and the only one designed to compare the two drugs specifically in MRSA pneumonia found a statistically significant superiority.3 The purpose of a meta-analysis is to improve the power of small or inconclusive studies to answer specific questions, determine consistency of effect across different study designs, and explore differences.14 In 10-25% of cases, large clinical trials specifically designed to address the issues raised by meta-analysis of small trials or subgroups of larger trials refute the findings of the meta-analysis.15 Analysis of the pattern of these discrepant findings suggest that if the total number of patients in the smaller studies is less than the single larger study and if an endpoint other than the primary of the original study is analyzed, the meta-analysis is likely to be in error.15 These criteria characterize previous meta-analyses, including that of Kalil,16 as well as this one. Meta-analyses of infectious disease studies are particularly problematic.17 Because of the limitations of meta-analysis, some have favored large observational studies, which often yield results very similar to randomized controlled trials.18 Caffrey et al found nearly identical findings as our above analysis – greater clinical success and non-statistically lower mortality with linezolid - in a propensity-adjusted comparative-effectiveness analysis of 5271 MRSA HAP patients in Veterans Affairs Medical Centers.19
We were also surprised by the conclusion that vancomycin had equivalent renal toxicity as linezolid. The finding of Kalil et al contrasts with that of other meta-analyses of almost the same studies20 21 that suggest a greater nephrotoxicity with vancomycin than linezolid, particularly when targeting higher trough levels.22
We agree that further large randomized, controlled trials (RCTs) comparing linezolid to vancomycin for MRSA pneumonia are unlikely to be performed. A Bayesian network meta-analysis of MRSA HAP, which included data submitted to the Food and Drug Administration (FDA), found better efficacy and safety rankings for linezolid than vancomycin, questioning vancomycin’s role as the standard for comparison in future studies of MRSA HAP.21 Use of vancomycin in a future FDA registration trial of a new agent for MRSA HAP is unlikely given that the inferiority of vancomycin compared to linezolid would raise issues regarding its adequacy as the comparator.23 Conversely, on the ClinicalTrials.gov site, the only current prospective Gram-positive pneumonia registration trial in adults uses linezolid as the comparator.
For the 72% of HAP cases due to pathogens other than MRSA, we agree with Dr. Muscedere’s editorial and Kalil et al’s meta-analysis conclusion that vancomycin and linezolid are equivalent. For the clinically more relevant issue of MRSA HAP treatment, linezolid is consistently superior to vancomycin in clinical response with no difference in mortality. The evidence includes the only large study specifically addressing this issue and a large observational trial, as well as an appropriately performed meta-analysis.
1. Kalil AC, Klompas M, Haynatzki G, Rupp ME. Treatment of hospital-acquired pneumonia with linezolid or vancomycin: a systematic review and meta-analysis. BMJ Open 2013;3(10):e003912.
2. Muscedere J. Which antibiotic for hospital acquired pneumonia caused by MRSA? BMJ 2014;348:g1469.
3. Wunderink RG, Niederman MS, Kollef MH, Shorr AF, Kunkel MJ, Baruch A, et al. Linezolid in methicillin-resistant Staphylococcus aureus nosocomial pneumonia: a randomized, controlled study. Clin Infect Dis 2012;54(5):621-9.
4. Jaksic B, Martinelli G, Perez-Oteyza J, Hartman CS, Leonard LB, Tack KJ. Efficacy and safety of linezolid compared with vancomycin in a randomized, double-blind study of febrile neutropenic patients with cancer. Clin Infect Dis 2006;42(5):597-607.
5. Kaplan SL, Deville JG, Yogev R, Morfin MR, Wu E, Adler S, et al. Linezolid versus vancomycin for treatment of resistant Gram-positive infections in children. Pediatr Infect Dis J 2003;22(8):677-86.
6. Rubinstein E, Cammarata S, Oliphant T, Wunderink R. Linezolid (PNU-100766) versus vancomycin in the treatment of hospitalized patients with nosocomial pneumonia: a randomized, double-blind, multicenter study. Clin Infect Dis 2001;32(3):402-12.
7. Wunderink RG, Cammarata SK, Oliphant TH, Kollef MH. Continuation of a randomized, double-blind, multicenter study of linezolid versus vancomycin in the treatment of patients with nosocomial pneumonia. Clin Ther 2003;25(3):980-92.
8. Wunderink RG, Rello J, Cammarata SK, Croos-Dabrera RV, Kollef MH. Linezolid vs vancomycin: analysis of two double-blind studies of patients with methicillin-resistant Staphylococcus aureus nosocomial pneumonia. Chest 2003;124(5):1789-97.
9. Lin DF, Zhang YY, Wu JF, Wang F, Zheng JC, Miao JZ, et al. Linezolid for the treatment of infections caused by Gram-positive pathogens in China. Int J Antimicrob Agents 2008;32(3):241-9.
10. Wunderink RG, Mendelson MH, Somero MS, Fabian TC, May AK, Bhattacharyya H, et al. Early microbiological response to linezolid vs vancomycin in ventilator-associated pneumonia due to methicillin-resistant Staphylococcus aureus. Chest 2008;134(6):1200-7.
11. Kohno S, Yamaguchi K, Aikawa N, Sumiyama Y, Odagiri S, Aoki N, et al. Linezolid versus vancomycin for the treatment of infections caused by methicillin-resistant Staphylococcus aureus in Japan. J Antimicrob Chemother 2007;60(6):1361-9.
12. Stevens DL, Herr D, Lampiris H, Hunt JL, Batts DH, Hafkin B. Linezolid versus vancomycin for the treatment of methicillin-resistant Staphylococcus aureus infections. Clin Infect Dis 2002;34(11):1481-90.
13. Spellberg B, Talbot G. Recommended design features of future clinical trials of antibacterial agents for hospital-acquired bacterial pneumonia and ventilator-associated bacterial pneumonia. Clin Infect Dis 2010;51 Suppl 1:S150-70.
14. Borenstein M, Hedges LV, Higgins JP, Julien PT, editors. Introduction to Meta-Analysis. First ed: John Wiley & Sons, Ltd, 2009.
15. Ioannidis JP, Cappelleri JC, Lau J. Issues in comparisons between meta-analyses and large trials. Jama 1998;279(14):1089-93.
16. Kalil AC, Murthy MH, Hermsen ED, Neto FK, Sun J, Rupp ME. Linezolid versus vancomycin or teicoplanin for nosocomial pneumonia: a systematic review and meta-analysis. Crit Care Med 2010;38(9):1802-8.
17. Ioannidis JP, Lau J. State of the evidence: current status and prospects of meta-analysis in infectious diseases. Clin Infect Dis 1999;29(5):1178-85.
18. Benson K, Hartz AJ. A comparison of observational studies and randomized, controlled trials. N Engl J Med 2000;342(25):1878-86.
19. Caffrey AR, Morrill HJ, Puzniak LA, Laplante KL. Comparative Effectiveness of Linezolid and Vancomycin Among a National Veterans Affairs Cohort with Methicillin-Resistant Staphylococcus aureus Pneumonia. Pharmacotherapy 2014.
20. Jiang H, Tang RN, Wang J. Linezolid versus vancomycin or teicoplanin for nosocomial pneumonia: meta-analysis of randomised controlled trials. Eur J Clin Microbiol Infect Dis 2013;32(9):1121-8.
21. Bally M, Dendukuri N, Sinclair A, Ahern SP, Poisson M, Brophy J. A network meta-analysis of antibiotics for treatment of hospitalised patients with suspected or proven meticillin-resistant Staphylococcus aureus infection. Int J Antimicrob Agents 2012;40(6):479-95.
22. van Hal SJ, Paterson DL, Lodise TP. Systematic review and meta-analysis of vancomycin-induced nephrotoxicity associated with dosing schedules that maintain troughs between 15 and 20 milligrams per liter. Antimicrob Agents Chemother 2013;57(2):734-44.
23. Fleming TR, Powers JH. Issues in noninferiority trials: the evidence in community-acquired pneumonia. Clin Infect Dis 2008;47 Suppl 3:S108-20.
Competing interests: Dr. Wunderink serves on a Data Safety Monitoring Board and his institution has received research support from Pfizer unrelated to linezolid. He is a consultant to Bayer, Cubist, Accelerate Diagnostics, and Rempex. Dr. Shorr is a consultant, speaker, and/or investigator for Astellas, AZ, Bayer, Cubist, Cardeas, Pfizer, Theravance, and Tetraphase. Dr. Niederman is a consultant or speaker for: Pfizer, Theravance, Merck, Cubist, Bayer, Sanofi-Pasteur; his institution has received research support from: Cubist, Bayer. Dr. Kollef is a consultant for Cubist, Cardeas, and Accelerate Diagnostics and has served on the speaking bureau for Cubist and Merck. Dr. McGee is on a speakers bureau and has received research support from Pfizer. Jean Chastre has received honoraria for advice or public speaking from Pfizer, Astellas, Bayer/Nektar, Brahms, Cubist, Janssen-Cilag, and Sanofi Pasteur/KaloBios; and his institution has received research grants from Bayer/Nektar.