Clinical Infectious Diseases Advance Access published April 21, 2014
1
Viridans group streptococci in febrile neutropenic cancer patients: what should we fear?
Alison G. Freifeld1, Raymund R. Razonable2
cr ipt
1Division of Infectious Diseases, University of Nebraska Medical Center, Omaha, NE
an M pt ed ce Ac © The Author 2014. Published by Oxford University Press on behalf of the Infectious Diseases Society of America. All rights reserved. For Permissions, please e‐mail:
[email protected].
Downloaded from http://cid.oxfordjournals.org/ at Universidade de Vigo on April 27, 2014
us
68198 2Division of Infectious Diseases, and the William J von Liebig Center for Transplantation and Clinical Regeneration, Mayo Clinic, Rochester, MN 55905 Corresponding author: Alison Freifeld MD, University of Nebraska Medical Center, Omaha, NE 68198‐5400,
[email protected], tel: 402‐559‐8668, fax: 402‐559‐5514
2 In this issue of CID, Shelburne et al developed and validated a clinical
cr ipt
prediction model for ‐lactam resistance in viridans group streptococci (VGS)
causing bloodstream infection (BSI) and offer guidance for the use of vancomycin
(or Gram‐positive spectrum antibiotics) for empiric therapy of febrile neutropenia (FN) [1]. Although VGS cause a significant minority of BSI in FN patients, they are
us
syndrome (ARDS) in 7 to 39%, with mortality rates ranging from 2 to 21%. [2‐9]
an
VGSS/ARDS was initially described in the 1990s and linked to the presence of profound neutropenia, oral mucositis , high dose chemotherapy, and also to
M
fluoroquinolone or trimethoprim‐sulfamethoxazole prophylaxis and ceftazidime empiric therapy for FN. [5, 6, 8, 10, 11] Poor in vitro activity against Streptococcus mitis, the most common cause of VGSS/ARDS, is characteristic of these antibiotics.
pt ed
Later generation expanded‐spectrum ‐lactams with more reliable activity against VGS (and other gram‐positive pathogens) have largely supplanted ceftazidime for empirical antibiotic monotherapy and include cefepime, piperacillin‐tazobactam, and the antipseudomonal carbapenems, imipenem and meropenem.[12, 13]
ce
Increasingly, however, VGS bearing diminished susceptibility to penicillin and to the newer ‐lactams are being recognized. [2, 6, 10, 11] These ‐lactam resistant
Ac
isolates primarily concern Shelburne et al as potentially causing rapid and disastrous complications from VGS bacteremia in FN patients. Driving this study is the thought that improved outcomes (e.g., lower
incidence of VGSS/ARDS and mortality) will be afforded to FN patients with ‐
Downloaded from http://cid.oxfordjournals.org/ at Universidade de Vigo on April 27, 2014
associated with a shock syndrome (VGSS) and/or acute respiratory distress
3 lactam resistant VGS bacteremia if they receive vancomycin (or Gram‐positive antibiotics) as part of the empiric ‐lactam‐based regimen. In 1997, Spanik et al
cr ipt
observed higher mortality rates in FN patients with penicillin‐resistant VGS
bacteremia, while in another study published in the same year, Elting et al reported increased mortality in FN patients who did not receive vancomycin initially for documented alpha‐hemolytic streptococcal BSI. [8, 10] Both studies largely
us
two important risks for VGSS/ARDS. Fueled by these reports, some practitioners
an
have reflexively added vancomycin to the empiric ‐lactam regimen for all FN patients, presuming that it will prevent the potentially life‐threatening effects of VGS
M
BSI. Indeed, at the MD Anderson Cancer Center where Shelburne et al conducted their study, a stunning 96% of all FN patients in their validation cohort (2011‐2013) received vancomycin as part of the empiric therapy. In contrast, the 2010 IDSA
pt ed
guidelines recommend much stricter criteria for empiric vancomycin use, suggesting it be limited to FN patients with hemodynamic instability or other evidence of severe sepsis (an indicator for possible VGSS), documented Gram‐ positive BSI prior to identification, skin and soft tissue infection, pneumonia, or
ce
those with known colonization or suspected infection due to ‐lactam non‐ susceptible organisms (e.g. methicillin resistant Staphylococcus aureus, penicillin‐
Ac
resistant Streptococcus pneumoniae). [12] While the IDSA guidelines recognize VGS BSI as a cause of shock and ARDS, there is no recommendation for broadly adding empiric vancomycin to an initial anti‐pseudomonal agent, in an attempt to treat VGS pre‐emptively. Why is there such disparity, then, between clinical practice and the
Downloaded from http://cid.oxfordjournals.org/ at Universidade de Vigo on April 27, 2014
included patients treated with fluoroquinolone prophylaxis and ceftazidime therapy,
4 IDSA guidelines regarding how vancomycin should be most effectively used (or not used) in fever and neutropenia? It may reflect differences in local epidemiology and
cr ipt
antibiogram patterns at various centers. Furthermore, no predictive tool has been available to accurately identify FN patients at greatest risk of VGSS/ARDS. A
prevailing fear among some clinicians is that FN patients will rapidly succumb to
sudden and unrelenting shock due to VGS bacteremia unless vancomycin (or other
us
utilization driven by fear is certainly a dilemma for antibiotic stewardship. As
an
Shelburne et al suggest, a re‐evaluation of the data may help to change this pattern. Underlying this study were several assumptions about VGS bacteremia: 1)
M
that a penicillin MIC≥2g/ml (defined by the authors as penicillin non‐ susceptibility) adequately identifies non‐susceptibility to later generation ‐
pt ed
lactams; 2) that a penicillin MIC≥2g/ml is a surrogate marker for increased risk of VGSS and mortality; and 3) that vancomycin administered at the onset of fever in neutropenic patients with VGS bacteremia will improve outcome. The authors reviewed all VGS bloodstream isolates from FN patients at their center over a 13
ce
year period, using 569 cases (2000‐2010) in a derivation set to identify factors correlating with non‐susceptible MICs for penicillin, and then prospectively assessed their predictive value using a validation cohort of 163 VGS cases (2011‐
Ac
2013). Three variables correlated with isolating VGS with a penicillin MIC≥2 g/ml: ‐lactam use in the prior 30 days, ‐lactam prophylaxis, and inpatient status at onset of FN. Only 1% of patients who lacked all 3 criteria were infected with a penicillin‐resistant VGS. The authors concluded that the presence of one or more of
Downloaded from http://cid.oxfordjournals.org/ at Universidade de Vigo on April 27, 2014
Gram‐positive drugs such as daptomycin) is administered “up front.” Antibiotic
5 the 3 elements could predict for penicillin non‐susceptible VGS with high sensitivity, but relatively low specificity and a positive predictive value of only 0.34. They
cr ipt
further calculated that 42% of all FN patients without these risks could have been spared receipt of empiric vancomycin. This finding is clearly a step in the right direction to curb vancomycin overuse, and we applaud the authors for their
evidence‐based proposal of narrowed indications for empiric vancomycin.
us
analysis were truly the ones at highest risk for VGSS/ARDS – and not just high MIC’s
outcome?
an
to penicillin – and if so, would empiric vancomycin have made a difference in overall
M
The study assumes that a penicillin MIC≥2ug/ml serves as a good surrogate for non‐susceptibility to later generation ‐lactams among VGS isolates. However, using E‐test, they found that all VGS with a penicillin MIC equal to 2ug/ml remained
pt ed
fully susceptible to cefepime and piperacillin‐tazobactam, the most commonly used empiric therapies; only a small proportion were resistant to meropenem. Thus, defining cefepime and piperacillin‐tazobactam as non‐susceptible at a penicillin MIC=2 ug/ml in the validation analysis creates a bias against their efficacy.
ce
Understandably, the authors wanted to “cast a wide net” to ensure that all resistant VGS were captured by their definition, but by doing so they may have overdrawn
Ac
conclusions about the circumstances in which vancomycin would be needed, when in fact, many VGS isolates would have been covered by standard cefepime or piperacillin‐tazobactam monotherapy.
Downloaded from http://cid.oxfordjournals.org/ at Universidade de Vigo on April 27, 2014
However, the key questions remain as to whether the patients identified by this
6 The idea that a higher penicillin MIC portends a worse outcome from VGS bacteremia is not consistently supported in the literature. Contrasting to the older
cr ipt
studies mentioned earlier, there are a number of studies that have found no
correlation between the VGS penicillin or ‐lactam susceptibility pattern and clinical outcome. [3‐7] Han et al recently reported 202 episodes of VGS bacteremia in FN children and adults, finding only 40% susceptible to penicillin but 80%
us
antibiotic susceptibility patterns. [3] Indeed, Shelburne et al also found no
an
difference in the incidence of VGSS and ARDS among patients with VGS showing a penicillin MIC≤1 g/ml (susceptible) compared to those with MIC≥2 g/ml
M
(resistant). These findings suggest that in vitro antibiotic susceptibility does not necessarily predict prognosis in VGS bacteremia.
pt ed
Finally, there are data to suggest an outcomes advantage when vancomycin is given early in patients with VGS with diminished susceptibility to ‐lactam. [9, 10] However, abundant data exist from randomized controlled trials (RTC) affirming that the empirical routine addition of anti‐Gram positive treatment with
ce
glycopeptides does not improve the outcomes of FN patients with cancer. [14, 15] In a meta‐analysis of 13 RCTs comparing empiric ‐lactams alone or with additional anti‐Gram‐positive antibiotic, similar rates of mortality and treatment failure were
Ac
observed. [15] The authors concluded that glycopeptides can be safely deferred until documentation of a resistant Gram‐positive bacterial infection is made. Similarly, Shelburne et al detected no significant difference in mortality related to vancomycin use. [1] In the vast majority of cases, VGS bacteremia pursues a
Downloaded from http://cid.oxfordjournals.org/ at Universidade de Vigo on April 27, 2014
susceptibility to cefepime; rates of VGSS, ARDS and death were independent of
7 relatively indolent course; only in a minority will it lead to VGSS/ARDS. The preponderance of data support the addition of vancomycin to empirical ‐lactam
cr ipt
therapy for FN if clinical deterioration occurs. [2, 12, 14, 15]
In conclusion, the current study provides a clinical prediction model to help
clinicians focus the prescription of empirical vancomycin (and other Gram‐positive antimicrobial agents) more sharply. This achievement contributes significantly to
us
pressure that can lead to bacterial resistance. However, we suggest moving beyond
an
using ‐lactam MICs as surrogates for high risk outcomes. A focus on factors correlating with the clinically relevant outcomes of shock, ARDS and mortality
M
should be a next step. Risk factors associated with VGS bacteremia, including profound neutropenia, oral mucositis, fluoroquinolone prophylaxis, inpatient status
pt ed
and prior ‐lactam use might be incorporated into a clinical model for prediction of the devastating outcomes related to VGS bacteremia. In the meantime, we urge adherence to current evidence‐based guidelines in an effort to stem the rising tide of vancomycin overuse and resistance.
Ac
ce
Notes: The authors have no reported conflicts of interest.
Downloaded from http://cid.oxfordjournals.org/ at Universidade de Vigo on April 27, 2014
the goals of antimicrobial stewardship, seeking to limit the burden of antibiotic
8
References
cr ipt
1. Shelburne SA, Lasky R, Sashasrabhojane P, Tarrand JR, Rolston K. Development and validation of a clinical model to predict the presence of β‐lactam resistance in viridans group streptococci causing bacteremia in neutropenic cancer patients. Clinical Infectious Diseases, 2014 (in press).
2. Han SB, Bae EY, Lee JW, et al. Clinical characteristics and antibiotic susceptibility of viridans streptococcal bacteremia in children with febrile neutropenia. Infection, 2013; 41(5): 917‐24.
us
an
4. Husain E, Whitehead S, Castell A, Thomas EE, Speert DP. Viridans streptococci bacteremia in children with malignancy: Relevance of species identification and penicillin susceptibility. Pediatr Infect Dis J, 2005; 24(6): 563‐6.
M
5. Marron A, Carratala J, Gonzalez‐Barca E, Fernandez‐Sevilla A, Alcaide F, Gudiol F. Serious complications of bacteremia caused by viridans streptococci in neutropenic patients with cancer. Clin Infect Dis, 2000; 31(5): 1126‐30.
pt ed
6. Marron A, Carratala J, Alcaide F, Fernandez‐Sevilla A, Gudiol F. High rates of resistance to cephalosporins among viridans‐group streptococci causing bacteraemia in neutropenic cancer patients. J Antimicrob Chemother, 2001; 47(1): 87‐91. 7. Gassas A, Grant R, Richardson S, et al. Predictors of viridans streptococcal shock syndrome in bacteremic children with cancer and stem‐cell transplant recipients. J Clin Oncol, 2004; 22(7): 1222‐7.
Ac
ce
8. Spanik S, Trupl J, Kunova A, et al. Viridans streptococcal bacteraemia due to penicillin‐resistant and penicillin‐sensitive streptococci: Analysis of risk factors and outcome in 60 patients from a single cancer centre before and after penicillin is used for prophylaxis. Scand J Infect Dis, 1997; 29(3): 245‐9. 9. Jaffe D, Jakubowski A, Sepkowitz K, et al. Prevention of peritransplantation viridans streptococcal bacteremia with early vancomycin administration: A single‐ center observational cohort study. Clin Infect Dis, 2004; 39(11): 1625‐32. 10. Elting LS, Rubenstein EB, Rolston KV, Bodey GP. Outcomes of bacteremia in patients with cancer and neutropenia: Observations from two decades of epidemiological and clinical trials. Clin Infect Dis, 1997; 25(2): 247‐59.
Downloaded from http://cid.oxfordjournals.org/ at Universidade de Vigo on April 27, 2014
3. Han SB, Bae EY, Lee JW, et al. Clinical characteristics and antimicrobial susceptibilities of viridans streptococcal bacteremia during febrile neutropenia in patients with hematologic malignancies: A comparison between adults and children. BMC Infect Dis, 2013; 13: 273,2334‐13‐273.
9 11. Bochud PY, Eggiman P, Calandra T, Van Melle G, Saghafi L, Francioli P. Bacteremia due to viridans streptococcus in neutropenic patients with cancer: Clinical spectrum and risk factors. Clin Infect Dis, 1994; 18(1): 25‐31.
cr ipt
12. Freifeld AG, Bow EJ, Sepkowitz KA, et al. Clinical practice guideline for the use of antimicrobial agents in neutropenic patients with cancer: 2010 update by the infectious diseases society of america. Clin Infect Dis, 2011; 52(4): e56‐93.
13. Pfaller MA, Marshall SA, Jones RN. In vitro activity of cefepime and ceftazidime against 197 nosocomial blood stream isolates of streptococci: A multicenter sample. Diagn Microbiol Infect Dis, 1997; 29(4): 273‐6.
us
an
15. Paul M, Dickstein Y, Borok S, Vidal L, Leibovici L. Empirical antibiotics targeting gram‐positive bacteria for the treatment of febrile neutropenic patients with cancer. Cochrane Database Syst Rev, 2014; 1: CD003914.
Ac
ce
pt ed
M
Downloaded from http://cid.oxfordjournals.org/ at Universidade de Vigo on April 27, 2014
14. Paul M, Borok S, Fraser A, Vidal L, Leibovici L. Empirical antibiotics against gram‐positive infections for febrile neutropenia: Systematic review and meta‐ analysis of randomized controlled trials. J Antimicrob Chemother, 2005; 55(4): 436‐ 44.