Previous Article | Next Article ![]()
Antimicrobial Agents and Chemotherapy, September 2007, p. 3298-3303, Vol. 51, No. 9
0066-4804/07/$08.00+0 doi:10.1128/AAC.00262-07
Copyright © 2007, American Society for Microbiology. All Rights Reserved.

Division of Infectious Diseases, Department of Medicine, University of Arkansas for Medical Sciences, and Central Arkansas Veterans Healthcare System, Little Rock, Arkansas
Received 21 February 2007/ Returned for modification 27 March 2007/ Accepted 12 June 2007
|
|
|---|
|
|
|---|
Potential treatment options for patients who do not require intravenous therapy include trimethoprim-sulfamethoxazole, clindamycin, and the tetracyclines (i.e., minocycline, doxycycline, and tetracycline). However, few data are currently available to guide clinicians in the optimal use of these agents. The expanded-spectrum tetracyclines doxycycline and minocycline demonstrate excellent oral bioavailability, tissue penetration, and tolerability (14). Published reports on the use of these agents for the treatment of MRSA SSTI suggest their effectiveness but are limited to case studies (2, 3, 23). Thus, more data are needed. The primary objective of this study was to describe the clinical characteristics and outcomes of a large cohort of patients that received expanded-spectrum tetracyclines for tetracycline-susceptible MRSA SSTI acquired in an outpatient setting. As a secondary aim, we compared these patients to a patient cohort that shared very similar clinical characteristics but was treated with ß-lactam monotherapy within the same time period.
|
|
|---|
Definition of terms and study design. A soft tissue infection was defined as having a community onset if the causative organism was isolated from a tissue specimen obtained in an outpatient setting or within 48 h of hospital admission. Patients with community onset MRSA SSTI were identified by review of the clinical microbiology laboratory records at both hospitals. An episode was further classified as a community-associated MRSA (CA-MRSA) SSTI if none of the following previously defined epidemiological criteria for a health care-associated infection were fulfilled: (i) prior history of an MRSA infection or colonization; (ii) the presence of an invasive device such as a central venous catheter; and (iii) dialysis, surgery, hospitalization, or residence in a long-term care facility in the 12 months preceding the culture (20).
SSTI episodes were grouped into three categories: (i) primary cellulitis which extended from a visible focus of infection, such as folliculitis, a minor abscess, or a nonchronic skin ulcer; (ii) furuncles or carbuncles; and (iii) cutaneous abscesses. Minor infections, such as impetigo and folliculitis, which generally require no or only topical antibiotic therapy, were excluded, as were complicated infections that involved deep structures, such as tendons, fascia, and muscle, as demonstrated by operative or radiological findings. Postoperative wound infections and chronic processes, including diabetic foot infections and nonhealing skin ulcers, were also excluded, as were cases in which multiple organisms were isolated from a wound culture.
The time zero of the study was defined as the day of the first incision and drainage procedure. For those patients on whom surgical drainage was not performed, the date of the first positive culture from a wound exudate, needle aspirate, or skin biopsy specimen was designated time zero. Treatment failure was the primary outcome of interest and was defined by the presence of two criteria: (i) documented worsening of clinical signs and symptoms of infection, such as increased purulence, erythema, induration, and/or tenderness at least 2 days after time zero, and (ii) the performance of a second, not previously planned incision and drainage procedure and/or admission to the hospital, both by at least 2 days after time zero.
Doxycycline and minocycline were both given orally at a dose of 100 mg twice daily. ß-Lactam therapy consisted of intravenous agents such as cefazolin or piperacillin-tazobactam and oral agents such as cephalexin, dicloxacillin, or amoxicillin-clavulanate administered in standard doses to the majority of patients. Patients without signs of treatment failure who had received at least 3 days of ß-lactam monotherapy and whose therapy was subsequently changed to another agent (in response to results from antimicrobial susceptibility testing in the majority of cases) were included as ß-lactam treatment successes.
The study was approved by the institutional review boards of both hospitals.
Data review. The inpatient, outpatient, emergency department, and pharmacy records of all patients with MRSA cultured from a normally nonsterile site within the observation period were reviewed to ensure enrollment of all episodes that fulfilled our inclusion criteria. Sociodemographic and clinical data were obtained with the use of a standardized data extraction instrument. For abscesses, furuncles, and carbuncles, the largest diameter of the lesion was recorded. Comorbidity was assessed with the Charlson comorbidity score (5, 17). The systemic inflammatory response syndrome (SIRS) was diagnosed based on previously described criteria (1). We also determined whether the incision and drainage procedure at time zero was performed by a surgical service or a nonsurgical service, including the emergency medicine department. Patient follow-up was recorded as the number of clinical encounters that occurred between days 2 and 21 after time zero. If a patient experienced more than one eligible SSTI episode during the observation period, the later episode was included as an independent event if it occurred at least 1 month after the first presentation and involved a different body site.
Microbiological methods. The Staphylococcus aureus isolates' susceptibilities to oxacillin, tetracycline, ciprofloxacin, trimethoprim-sulfamethoxazole, gentamicin, rifampin, and vancomycin were determined by the broth microdilution method with the Vitek 2 system (bioMérieux, Durham, NC). The double-disk diffusion assay was performed to test for inducible clindamycin resistance in accordance with the recommendations of the Clinical and Laboratory Standards Institute (6, 7).
Statistical analysis.
Bivariate analyses were conducted by Pearson's
2 test, Fisher's exact test, and the Wilcoxon rank sum test to compare categorical and continuous (but not normally distributed) variables (P < 0.05). We calculated odds ratios (ORs) in order to facilitate comparison to the results from multivariate analyses. Logistic regression analyses that included interaction terms were conducted to derive independent risk factors for treatment failure. Factors with a P value of <0.2 on bivariate analysis were included in the model. SPSS for Windows software, version 11.0 (SPSS), was used for all statistical analyses.
|
|
|---|
Baseline demographics and comorbidities. The two treatment groups were similar with regard to baseline demographic characteristics and comorbidities (Table 1). For the entire cohort (282 episodes), the median patient age was 48 years and 236 patients (84%) were male. A total of 184 SSTI episodes (65%) were treated at the Central Arkansas Veterans Healthcare System. Twelve patients (4%) used intravenous drugs. The administration of antibiotics within the 12 months prior to presentation was common; ß-lactams (36%) and fluoroquinolones (18%) were the most commonly prescribed agents. One hundred ninety-three episodes (68%) fulfilled the criteria of a CA-MRSA SSTI. The levels of susceptibility to non-ß-lactam antibiotic classes were 5% for macrolides (13 of 282 isolates), 71% for ciprofloxacin (199 of 279 isolates tested), 99.5% for trimethoprim-sulfamethoxazole (217 of 218 isolates tested), 100% for rifampin and gentamicin (219 isolates tested, each), 100% for vancomycin (all 282 isolates tested), and 99% for clindamycin (278 of 282 isolates tested). The double-disk diffusion test was performed routinely only toward the end of the study period; only 1 (3%) of 32 tested isolates displayed inducible clindamycin resistance.
|
View this table: [in a new window] |
TABLE 1. Baseline demographic characteristics and comorbid conditions of patients with community onseta MRSA SSTI treated with expanded-spectrum tetracyclines or ß-lactams
|
Clinical presentation and therapy. Table 2 describes the clinical characteristics and patient outcomes categorized by type of antimicrobial therapy. For the entire cohort (282 episodes of SSTI), abscesses were the most commonly encountered entity (75%). Only 35 (12%) of the 282 episodes were classified as cellulitis which originated from a purulent focus of infection. The maximum diameter of induration was documented for 206 (86%) of 247 abscesses, furuncles, or carbuncles; the median lesion size was 4 cm (interquartile range, 3 to 5 cm), and diameters did not differ significantly between the two treatment groups. A second pathogen grew in four clinical specimens; a group B streptococcus, Proteus vulgaris, a coagulase-negative staphylococcus, and a viridans group streptococcus were present in one clinical specimen each. A surgical drainage procedure was performed on the majority of patients in both treatment groups. One hundred ninety-three (91%) of 211 abscesses required incision and drainage. Of note, most patients with cellulitis (66%) also underwent incision and drainage of the primary focus of infection. Needle aspiration of abscess fluid, manual expression of pus, or spontaneous wound drainage was considered adequate in the 57 episodes (20%) managed without formal incision and drainage.
|
View this table: [in a new window] |
TABLE 2. Clinical characteristics of and outcomes for patients with community onseta MRSA SSTI treated with expanded-spectrum tetracyclines or ß-lactams
|
Clinical outcome and predictors of treatment failure. Treatment failure occurred for 28 (10%) of 282 episodes at a median of 3 days (range, 2 to 21 days; interquartile range, 2 to 3 days) after time zero. The majority of clinical failures (23 of 28) consisted of patients who required a repeat incision and drainage procedure due to worsening clinical findings on a follow-up visit, with or without concomitant hospital admission. The other five patients required subsequent hospital admission due to failing antibiotic therapy and improved after treatment with an active agent (an agent to which the organism demonstrated in vitro susceptibility) and additional wound care were initiated.
Table 3 shows predictors of treatment failure on bivariate analyses. A variety of clinical parameters, such as the severity of an underlying illness, the presence of a SIRS, or the site of infection, were not associated with a negative outcome. After adjustment for potential confounders on logistic regression analysis, receipt of a ß-lactam agent remained the only clinical predictor of treatment failure (adjusted OR, 3.94; 95% confidence interval [CI], 1.28 to 12.15; P = 0.02).
|
View this table: [in a new window] |
TABLE 3. Association of demographic and clinical characteristics with treatment failure in patients with community onseta MRSA SSTI
|
|
|
|---|
Older studies found minocycline to have a more-potent in vitro antistaphylococcal activity than doxycycline (16, 18). Minocycline was also used successfully in a few reported cases of serious infections, such as endocarditis or osteomyelitis (15, 34). These data on its clinical utility are supported by results from a rabbit endocarditis model that used a single MRSA strain (21). In this study, reductions of the bacterial density of the vegetation were similar among vancomycin- and minocycline-treated animals, combined with better tissue penetration by the latter agent. In our study, the vast majority of patients (97%) received doxycycline monotherapy. Therefore, we were not able to compare its therapeutic efficacy to that of minocycline. Both agents were well tolerated; mild gastrointestinal side effects or a rash occurred in three patients, but these symptoms did not necessitate a modification from doxycycline therapy to a different agent.
At our institutions, the rates of susceptibility to the tetracyclines among MRSA isolates have remained high and stable at approximately 95% over the last 4 years. This number includes nosocomial isolates that generally tend to be less susceptible (20). During our 53-month observation period, patients experiencing eight episodes of tetracycline-resistant MRSA SSTI not enrolled in this study were treated with a tetracycline; one (13%) failed treatment. Staphylococcus aureus resistance to the tetracyclines is frequently conferred by the chromosomally located Tet M determinant, which mediates protective modifications on the ribosome; its presence generally renders all class members, including minocycline, ineffective (25, 28). However, some MRSA strains that display resistance to tetracycline contain only the Tet K determinant, coding for drug efflux; the growth of these isolates is still inhibited by minocycline. Because commonly used clinical microbiology laboratory methodologies are not capable of differentiating these genotypes, clinicians should consider all isolates that display nonsusceptibility to tetracycline to be resistant to minocycline also (31). A recent large study that enrolled patients with acute, purulent SSTI from 11 urban emergency departments in the United States described results similar to ours: 207 (92%) of 226 MRSA isolates were susceptible to the tetracyclines (19). Clinicians who consider prescribing an expanded-spectrum tetracycline for the treatment of a suspected MRSA SSTI should be familiar with their local rates of resistance, as these rates vary within North America and worldwide (8, 10, 12). This drug class may be especially useful in patients with a sulfonamide allergy, an entity diagnosed for about 5% of patients in a large general-practice database (27).
The importance of the pulsed-field-type USA300 MRSA strain as the most common cause of community-associated SSTI in the United States has recently been documented (19). This strain displays previously described molecular characteristics, such as the presence of a specific mobile gene complex that confers methicillin resistance (SCCmecA type IV) and the Panton-Valentine leukocidin (30). Pulsed-field gel electrophoresis data were not available in our study; instead, we used the epidemiological definition of a community onset infection as supported by previous reports (13, 20). However, a recent study performed by the Arkansas Department of Health also demonstrated USA300 to be the most frequently isolated strain from staphylococcal SSTI in central Arkansas (24). Furthermore, antimicrobial susceptibility patterns to non-ß-lactam agents such as the fluoroquinolones or trimethoprim-sulfamethoxazole in our cohort were very similar to those observed in other studies (9, 13). Thus, we conclude that the microbiological and patient characteristics in our study may be applicable to other geographical regions.
As a second objective of our study, we compared the 90 patients treated with expanded-spectrum-tetracycline monotherapy to a group of 192 subjects who received ß-lactams. Both cohorts were very similar with regard to a variety of underlying demographic and clinical characteristics, including surgical wound drainage at time zero. ß-Lactam use was associated with a higher rate of treatment failure than was expanded-spectrum-tetracycline use (12.5% versus 4.4%) and remained a significant predictor of a negative outcome on logistic regression analysis (P = 0.02). Interpreted differently, 13 patients would have to be treated with expanded-spectrum tetracyclines in order to prevent one additional treatment failure associated with inactive antistaphylococcal therapy. Therefore, our results support the recommendations of current guidelines that emphasize the primary importance of surgical drainage, which may then be followed by the administration of agents that are active against methicillin-resistant staphylococcal isolates. More data are needed to determine which patients would benefit most from adjunctive antimicrobial therapy.
Our study has disadvantages in addition to those already mentioned. Patients were not randomly assigned to either type of antimicrobial therapy. Thus, our results may have been influenced by differences between the two groups that were not accounted for in the analysis. However, the high degree of similarity with regard to a variety of important clinical characteristics, such as underlying disease, clinical presentation, and rates of surgical drainage, among both treatment groups makes this less likely. Additionally, the difference in outcome between the two subgroups may have been overestimated by the exclusion of a number of patients who did not present for follow-up, presumably due to treatment success not requiring a second visit. The relatively small number of excluded patients limits this possibility. Furthermore, the retrospective design may have introduced additional forms of bias, such as misclassification and outcome bias, which we attempted to limit by the use of clear and objective definitions of the main study parameters, such as time zero and treatment outcome.
In summary, the expanded-spectrum tetracyclines appear to be a promising treatment option for patients who present with a suspected CA-MRSA SSTI that does not require intravenous therapy. Current local antimicrobial resistance patterns should be taken into account in therapeutic decision making. Studies that compare the efficacies of these agents to those of other treatment options for CA-MRSA SSTI, such as clindamycin, linezolid, or the sulfonamides, are needed to guide clinicians in the management of this emerging clinical entity.
We thank Philip Mwangi and Kim Hoffman for their technical assistance.
Published ahead of print on 18 June 2007. ![]()
|
|
|---|
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2009 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»