Empirical therapy for diabetic foot infections
3
EDITORIAL 10.1111/j.1469-0691.2007.01697.x Empirical therapy for diabetic foot infections: are there clinical clues to guide antibiotic selection? B. A. Lipsky 1,2 1 University of Washington and 2 Primary Care Clinic, VA Puget Sound Health Care System, Seattle, WA, USA ABSTRACT Initial antibiotic therapy for diabetic foot infections is usually empirical. Several principles may help to avoid selecting either an unnecessarily broad or inappropriately narrow regimen. First, clinically severe infections require broad-spectrum therapy, while less severe infections may not. Second, aerobic Gram- positive cocci, particularly Staphylococcus aureus (including methicillin-resistant S. aureus (MRSA) for patients at high-risk) should always be covered. Third, therapy should also be targeted at aerobic Gram- negative pathogens if the infection is chronic or has failed to respond to previous antibiotic therapy. Fourth, anti-anaerobe agents should be considered for necrotic or gangrenous infections on an ischaemic limb. Parenteral therapy is needed for severe infections, but oral therapy is adequate for most mild or moderate infections. Keywords Antibiotic therapy, diabetic foot infections, empirical therapy, foot infections, guidelines, treatment Clin Microbiol Infect 2007; 13: 351–353 Foot infections are common in individuals with diabetes. Among the decisions that the clinician treating such a patient must make is to select the most appropriate antibiotic regimen, usually in the face of inadequate microbiological informa- tion. If too narrow a spectrum is chosen, there is a risk of missing a pathogen in these often polymi- crobial infections, potentially leading to a poor clinical outcome. Opting for unnecessarily broad- spectrum therapy contributes to the growing problem of antibiotic resistance, and potentially increases the risk of drug toxicity and treatment expense. Considering that initial treatment is empirical in about two-thirds of cases, how is a responsible clinician to decide? Some basic principles can provide guidance. First, criteria established by the Infectious Dis- eases Society of America [1] or the International Working Group on the Diabetic Foot [2] should be used to determine the severity of the infection. Broad-spectrum therapy is usually needed only for patients with severe infections; these require an adequate antibiotic regimen until the results of culture are available. Mild, and most moderate, infections can often be treated with an agent with a more narrow spectrum of activity. Even if some organisms isolated from an infection are found to be resistant to the selected regimen, most patients with non-severe infections will improve (or at least stabilise) if they receive proper supportive care, debridement, pressure off-loading and wound care. Second, therapy should virtually always in- clude coverage for aerobic Gram-positive cocci, especially Staphylococcus aureus, which is both the most frequent and virulent pathogen isolated. Whether or not empirical coverage for methicillin- resistant S. aureus (MRSA) should be provided is a growing concern worldwide [3,4]. This decision depends largely on the overall local prevalence of MRSA, and the presence or absence of risk-factors for MRSA infection, e.g., recent hospitalisation or residence in a healthcare facility, recent antibiotic therapy or a requirement for renal dialysis. Agents that cover S. aureus will usually cover the b-haemolytic streptococci that are also relat- ively frequent pathogens. While enterococci are often isolated from diabetic foot infections, they are rarely primary pathogens, and are often Corresponding author and reprint requests: B. A. Lipsky, Primary Care Clinic, VA Puget Sound Health Care System, 1660 S. Columbian Way, Seattle, WA 98108-1597, USA E-mail: [email protected] Ó 2007 Copyright by the European Society of Clinical Microbiology and Infectious Diseases
description
Lipsky
Transcript of Empirical therapy for diabetic foot infections
EDITORIAL 10.1111/j.1469-0691.2007.01697.x
Empirical therapy for diabetic foot infections: are there clinical clues toguide antibiotic selection?B. A. Lipsky1,2
1University of Washington and 2Primary Care Clinic, VA Puget Sound Health Care System, Seattle,WA, USA
ABSTRACT
Initial antibiotic therapy for diabetic foot infections is usually empirical. Several principles may help toavoid selecting either an unnecessarily broad or inappropriately narrow regimen. First, clinically severeinfections require broad-spectrum therapy, while less severe infections may not. Second, aerobic Gram-positive cocci, particularly Staphylococcus aureus (including methicillin-resistant S. aureus (MRSA) forpatients at high-risk) should always be covered. Third, therapy should also be targeted at aerobic Gram-negative pathogens if the infection is chronic or has failed to respond to previous antibiotic therapy.Fourth, anti-anaerobe agents should be considered for necrotic or gangrenous infections on an ischaemiclimb. Parenteral therapy is needed for severe infections, but oral therapy is adequate for most mild ormoderate infections.
Keywords Antibiotic therapy, diabetic foot infections, empirical therapy, foot infections, guidelines, treatment
Clin Microbiol Infect 2007; 13: 351–353
Foot infections are common in individuals withdiabetes. Among the decisions that the cliniciantreating such a patient must make is to select themost appropriate antibiotic regimen, usually inthe face of inadequate microbiological informa-tion. If too narrow a spectrum is chosen, there is arisk of missing a pathogen in these often polymi-crobial infections, potentially leading to a poorclinical outcome. Opting for unnecessarily broad-spectrum therapy contributes to the growingproblem of antibiotic resistance, and potentiallyincreases the risk of drug toxicity and treatmentexpense. Considering that initial treatment isempirical in about two-thirds of cases, how is aresponsible clinician to decide?
Some basic principles can provide guidance.First, criteria established by the Infectious Dis-eases Society of America [1] or the InternationalWorking Group on the Diabetic Foot [2] should beused to determine the severity of the infection.Broad-spectrum therapy is usually needed onlyfor patients with severe infections; these require
an adequate antibiotic regimen until the results ofculture are available. Mild, and most moderate,infections can often be treated with an agent witha more narrow spectrum of activity. Even if someorganisms isolated from an infection are found tobe resistant to the selected regimen, most patientswith non-severe infections will improve (or atleast stabilise) if they receive proper supportivecare, debridement, pressure off-loading andwound care.
Second, therapy should virtually always in-clude coverage for aerobic Gram-positive cocci,especially Staphylococcus aureus, which is both themost frequent and virulent pathogen isolated.Whether or not empirical coverage for methicillin-resistant S. aureus (MRSA) should be provided isa growing concern worldwide [3,4]. This decisiondepends largely on the overall local prevalence ofMRSA, and the presence or absence of risk-factorsfor MRSA infection, e.g., recent hospitalisation orresidence in a healthcare facility, recent antibiotictherapy or a requirement for renal dialysis.Agents that cover S. aureus will usually coverthe b-haemolytic streptococci that are also relat-ively frequent pathogens. While enterococci areoften isolated from diabetic foot infections, theyare rarely primary pathogens, and are often
Corresponding author and reprint requests: B. A. Lipsky,Primary Care Clinic, VA Puget Sound Health Care System,1660 S. Columbian Way, Seattle, WA 98108-1597, USAE-mail: [email protected]
� 2007 Copyright by the European Society of Clinical Microbiology and Infectious Diseases
simply colonisers [5] that do not usually requirespecifically targeted therapy. One exception maybe patients who are failing to respond to treat-ment with a cephalosporin, as these agents canpredispose to infection with Enterococcus spp.
Third, the time to target aerobic Gram-negativebacilli must be considered. These microorganismsare isolated in up to two-thirds of infections, butusually from patients with a chronic infectionwho have recently received antibiotic therapy.Various Enterobacteriaceae are the most frequentGram-negative isolates, and these are covered bymost commonly used advanced broad-spectrumagents. The main issue with Gram-negative cov-erage is to decide when to specifically targetPseudomonas aeruginosa. This organism is usuallyfound in water, and is thus most often isolatedfrom patients who have been receiving some formof hydrotherapy. Pseudomonas is among the morecommonly isolated organisms in some popula-tions [6,7], but is usually found as part of apolymicrobial infection [5,8]. In general, antibiotictherapy does not need to be directed at Pseudo-monas isolates unless they are the predominantpathogens from a deep tissue specimen. Coverageof all Gram-negative isolates, especially extended-spectrum b-lactamase producers and resistantPseudomonas spp., would require a broader-spec-trum regimen than is usually necessary.
Fourth, the role of obligate anaerobic bacteriashould be considered. Studies that have properlyobtained and processed specimens to allow isola-tion of anaerobes show that they are present in asubstantial minority of cases [9–11]. However, aswith Pseudomonas, they are usually isolated in amixed infection with aerobes, rather than as a sole
pathogen [5]. Anaerobes are isolated more oftenfrom patients with limb ischaemia and conse-quent necrosis or gangrene. A clinical clue to thepresence of anaerobes is the feculent odour thatthey produce. Debriding necrotic material, andthereby removing many of the anaerobes andexposing those remaining to air, may be all that isneeded to treat these potential pathogens. Thus,an antibiotic regimen aimed specifically at anaer-obes is required only when there is strong clinicalevidence of anaerobic infection, i.e., the classic‘fetid foot’ [12].
Finally, having considered the likely aetiologi-cal agents and appropriate antibiotic choices, theroute of therapy must also be decided. Severeinfections require parenteral therapy to achievereliable blood levels quickly. The advent of highlybioavailable oral antimicrobial agents, e.g., fluoro-quinolones and linezolid, has made oral therapymore acceptable. Several studies of the treatmentof diabetic foot infections have revealed excellentoutcomes with oral antibiotics alone [13,14]. Theimportant issue is whether or not an adequatelevel of antibiotic reaches the infected tissue, nothow it arrives there. Thus, many patients can betreated with a quick switch to oral agents whenthey clinically improve, or even with initial oraltherapy. For some mildly infected and superficialwounds, topical therapy may be another option[15–17].
In conclusion, as shown in Table 1, relativelynarrow-spectrum therapy directed against aerobicGram-positive cocci, e.g., a semi-synthetic peni-cillin or first-generation cephalosporin, may beadequate for many acute infections in patientswho have not recently been treated with an
Table 1. Guide to selecting an empirical antibiotic regimen for a diabetic foot infection
Type of infection Likely pathogen(s) Class of antibiotic(s)
Acute, antibiotic-naive;low-risk MRSA
Aerobic GPC Penicillins; first-generation cephalosporins
Healthcare-associated; MRSA Co-trimoxazole; doxycycline;high local rates of MRSA clindamycin; glycopeptide;
linezolid; daptomycin
Chronic, previousantibiotic treatment
GPC + GNR± anaerobes
b-Lactam, b-lactamase inhibitor;second- or third-generationcephalosporin; group 1 carbapenem; FQ
Necrotic, gangrenousischaemic limb;foul odour
GPC + GNBobligate anaerobes
Clindamycin (± FQ);metronidazole (+ FQ);b-lactam, b-lactamase inhibitor;carbapenem
Hydrotherapy;green–blue-coloureddrainage
Pseudomonasaeruginosa
Anti-pseudomonal FQ,penicillin or cephalosporin
GPC, Gram-positive cocci; GNB, Gram-negative bacilli; FQ, fluoroquinolone; MRSA, methicillin-resistant Staphylococcus aureus.
352 Clinical Microbiology and Infection, Volume 13 Number 4, April 2007
� 2007 Copyright by the European Society of Clinical Microbiology and Infectious Diseases, CMI, 13, 351–353
antibiotic. Where MRSA is a concern, community-acquired infections may be treated effectivelywith co-trimoxazole, doxycycline or clindamycin;hospital-acquired strains often require a glyco-peptide, linezolid or daptomycin. As infectionsbecome chronic, and particularly if they fail torespond to antibiotic therapy, they tend tobecome polymicrobial, with the addition ofGram-negative bacilli. In these instances, broad-er-spectrum treatment, accomplished either withappropriate single or combination agents, isjudicious. Necrotic or gangrenous wounds, espe-cially in a patient with poor arterial flow, are ofteninfected with anaerobes that may act synergicallywith the aerobes, and thus require anti-anaerobictherapy, e.g., with clindamycin, metronidazole ora broader-spectrum agent with good anaerobiccoverage.
Other clues may help to predict the probablepathogen(s) in a diabetic foot infection. Anyavailable previous wound culture results shouldbe checked, and the possibility of recent antibiotictherapy or hospitalisations that could increase thelikelihood of drug-resistant bacteria should beinvestigated [18]. Gram’s-stained smears ofwound specimens should be reviewed [19,20].While there are few published data concerningpatients with diabetic foot infections, the Gram’sstain results should inform the clinician about themorphological and tinctorial properties of iso-lates. Specifically, when the smear shows a poly-microbial infection, broader coverage isappropriate; otherwise, a narrower spectrum ofantimicrobial coverage may suffice. Using theprinciples cited above, clinicians can offer apatient with a diabetic foot infection the ‘justright’ choice of empirical antibiotic therapy.
REFERENCES
1. Lipsky BA, Berendt AR, Deery HG et al. IDSA guidelines:diagnosis and treatment of diabetic foot infections. ClinInfect Dis 2004; 39: 885–910.
2. Lipsky BA. A report from the international consensus ondiagnosing and treating the infected diabetic foot. DiabetesMetab Res Rev 2004; 20 (suppl 1): S68–S77.
3. Wagner A, Reike H, Angelkort B. Highly resistant patho-gens in patients with diabetic foot syndrome with specialreference to methicillin-resistant Staphylococcus aureusinfections (in German). Dtsch Med Wochenschr 2001; 126:1353–1356.
4. Tentolouris N, Petrikkos G, Vallianou N et al. Prevalenceof methicillin-resistant Staphylococcus aureus in infectedand uninfected diabetic foot ulcers. Clin Microbiol Infect2006; 12: 186–189.
5. Lipsky BA, Armstrong DG, Citron DM, Tice AD, Mor-genstern DE, Abramson MA. Ertapenem versus pipera-cillin ⁄ tazobactam for diabetic foot infections (SIDESTEP):prospective, randomised, controlled, double-blinded,multicentre trial. Lancet 2005; 366: 1695–1703.
6. Shankar EM, Mohan V, Premalatha G, Srinivasan RS, UshaAR. Bacterial etiology of diabetic foot infections in SouthIndia. Eur J Intern Med 2005; 16: 567–570.
7. Chincholikar DA, Pal RB. Study of fungal and bacterialinfections of the diabetic foot. Ind J Pathol Microbiol 2002;45: 15–22.
8. Cunha BA. Antibiotic selection for diabetic foot infections:a review. J Foot Ankle Surg 2000; 39: 253–257.
9. El-Tahawy AT. Bacteriology of diabetic foot. Saudi Med J2000; 21: 344–347.
10. Goldstein EJ, Citron DM, Nesbit CA. Diabetic foot infec-tions. Bacteriology and activity of 10 oral antimicrobialagents against bacteria isolated from consecutive cases.Diabetes Care 1996; 19: 638–641.
11. Gerding DN. Foot infections in diabetic patients: the roleof anaerobes. Clin Infect Dis 1995; 20 (suppl 2): S283–S288.
12. Fierer J, Daniel D, Davis C. The fetid foot: lower-extremityinfections in patients with diabetes mellitus. Rev Infect Dis1979; 1: 210–217.
13. Lipsky BA, Pecoraro RE, Larson SA, Hanley ME, AhroniJH. Outpatient management of uncomplicated lower-extremity infections in diabetic patients. Arch Intern Med1990; 150: 790–797.
14. Lipsky BA, Itani K, Norden C. Treating foot infections indiabetic patients: a randomized, multicenter, open-labeltrial of linezolid versus ampicillin–sulbactam ⁄ amoxicillin–clavulanate. Clin Infect Dis 2004; 38: 17–24.
15. Apelqvist J, Ragnarson-Tenvall G, Larrson J. Topicaltreatment of diabetic foot ulcers: an economic analysis oftreatment alternatives and strategies. Diabet Med 1995; 12:123–128.
16. Lipsky BA, McDonald D, Litka P. Treatment of infecteddiabetic foot ulcers: topical MSI-78 vs. oral ofloxacin.Diabetologia 1997; 40 (suppl 1): 482.
17. Nelson EA, O’Meara S, Golder S, Dalton J, Craig D, Igle-sias C. Systematic review of antimicrobial treatments fordiabetic foot ulcers. Diabet Med 2006; 23: 348–359.
18. Kandemir O, Akbay E, Sahin E, Milcan A, General R. Riskfactors for infection of the diabetic foot with multi-antibi-otic resistant microorganisms. J Infect 2007; in press, e-pub.
19. Schmidt K, St Debus ESJ, Ziegler U, Thiede A. Bacterialpopulation of chronic crural ulcers: is there a differencebetween the diabetic, the venous, and the arterial ulcer?Vasa 2000; 29: 62–70.
20. Viswanathan V, Jasmine JJ, Snehalatha C, RamachandranA. Prevalence of pathogens in diabetic foot infection inSouth Indian type 2 diabetic patients. J Assoc PhysiciansIndia 2002; 50: 1013–1016.
Lipsky Empirical therapy of diabetic foot infections 353
� 2007 Copyright by the European Society of Clinical Microbiology and Infectious Diseases, CMI, 13, 351–353
