Reversion of Methicillin-Resistant Staphylococcus aureus Skin Infections to Methicillin-Susceptible Isolates

Importance  The rise of methicillin-resistant Staphylococcus aureus (MRSA) infections in the outpatient setting has led to a growing trend of empirical antibiotic treatment for MRSA. The limited oral antibiotics available and the growing resistance to these antibiotics make this a controversial practice.

Objective  To determine the frequency of patients with MRSA skin and soft-tissue infections (SSTIs) reverting to methicillin-susceptible Staphylococcus aureus (MSSA) positivity.

Design and Setting  Retrospective medical record review of inpatients and outpatients from our university hospital and clinics between January 1, 2000, and December 31, 2010.

Participants  Patients in our institutional microbiological database were included if they had a MRSA-positive SSTI and subsequent culture-proven S aureus SSTI more than 1 month later. No sociodemographic restrictions were applied. A sample of at least 200 patients meeting the above criteria was desired. The database was sorted by ascending medical record number, with the first 1681 patients’ medical records reviewed. Of these, 215 patients met our criteria.

Main Outcomes and Measures  Whether a patient remained MRSA positive in subsequent SSTIs or reverted to MSSA-positive infections.

Results  Of the total 215 patients, 64 (29.8%) had at least 1 incident of MSSA reversion, and 55 (25.6%) reverted to MSSA infections for the remainder of the study. We assessed various factors that might increase or decrease the likelihood of reversion. The presence of an invasive device was the only factor to demonstrate a statistically significant risk (relative risk, 1.20; 95% CI, 1.02-1.41; P = .03) toward remaining MRSA positive in subsequent infections.

Conclusions and Relevance  Patients with MRSA SSTIs demonstrated the ability to revert to subsequent MSSA SSTIs with a significant frequency. Further study regarding MRSA risk factors and their effects on subsequent infections would be valuable in guiding empirical treatment. Reculturing new infections in previously MRSA-positive patients is a prudent management strategy as we recognize that susceptibilities of the S aureus organisms change.

Staphylococcus aureus is an increasingly important pathogen in cutaneous infections. Both methicillin-susceptible S aureus (MSSA) and methicillin-resistant S aureus (MRSA) infections are becoming more prevalent in outpatient clinics and emergency departments.^1-3 Increasingly, dermatologists and other providers are responding to this trend by empirically treating MRSA infections.¹ The overall incidence of MRSA skin and soft-tissue infections (SSTIs) is increasing, but it is debatable whether the ratio of MRSA-positive to MSSA-positive staphylococcal infections is rising also.^4,5 The limited choice of oral antibiotics for MRSA infections and the changing resistance patterns to these antibiotics create a growing challenge in managing these outpatient infections.

Patients with atopic dermatitis have a higher colonization rate of S aureus and more frequent cutaneous infections compared with the general population.^6,7 This has been attributed to decreased antimicrobial peptides⁸ and antistaphylococcal lipids,⁹ impaired skin barrier,^10,11 and changes in the host immune response of lesional skin, leading to an inadequate response to staphylococcal colonization and invasion.¹² In atopic dermatitis, MRSA infections are not uncommon. Our clinical observations from patients with atopic dermatitis, however, have suggested that a significant number of such infections revert to MSSA after therapy.

We sought to investigate this observation among infections recorded in a large institutional database, focusing specifically on SSTIs. Our objectives were as follows: determine the frequency of MRSA reversion in patients with SSTIs in the institution’s microbiological database; identify factors that might increase or decrease the likelihood of MRSA reversion to MSSA; and seek evidence of re-reversion back from MSSA to MRSA in this population.

Data, derived from culture record log sheets at the Oregon Health & Science University microbiological laboratory, included both inpatient and outpatient culture results. Institutional review board approval from Oregon Health & Science University was obtained before the study for the use of the microbiological and patient medical records. All MRSA-positive microbiological results between January 1, 2000, and December 31, 2010, that were classified as either wound or tissue cultures were included in the initial database. This database was edited to include cultures with documented sites from skin, wound, or soft tissue. All suspected SSTI MRSA-positive cultures were verified by medical record review and then served as an index case. Only patients who had at least 1 additional S aureus culture-positive SSTI at least 1 month after the index case were included in the analyses. Sentinel colonization cultures were identified via medical record review as cases in which the documented infection site differed from the culture site or no active infection was noted. These colonization cultures were excluded from the study.

The medical records were reviewed for all subsequent MRSA and MSSA SSTIs through June 1, 2011, when data collection commenced. All sentinel MRSA cultures were obtained, however, between January 1, 2000, and December 31, 2010.

Medical records were also reviewed for other factors that might be predictive of MSSA reversion, including the following: basic demographics (age and sex), dermatologic information (diagnoses related to skin, mucosa, or wounds and whether the culture was associated with a flare of the cutaneous disease), health status (comorbidities, medications, and recent antibiotic use), hospital-acquired MRSA risk factors (procedures or hospitalizations since the first culture, history of an invasive device [an indwelling catheter, prosthesis, port, or other implanted stimulator or pain control device], and previous positive MRSA cultures), community-acquired MRSA risk factors (high-risk residence, participation in contact sports, and men having sex with men), and for each culture (site, date, culture result, infection vs colonization, relation to skin diagnosis, inpatient vs outpatient, and invasive device status).

Patients were then classified into 2 groups: those who remained MRSA positive in all subsequent infections and those who reverted to MSSA in at least 1 subsequent infection. Those who reverted to MSSA were further divided into a group that remained MSSA positive for all subsequent SSTIs available in the Oregon Health & Science University microbiological database and those who switched back to MRSA positivity at least once in future culture screenings during the study.

Relative risks were calculated and 2-tailed Fisher t tests were performed to determine the P values and 95% confidence intervals.

Among 1681 culture records reviewed in our database, 215 patients had at least 2 culture-positive SSTIs meeting our criteria. Demographic information is included in Table 1.

Patients with MRSA SSTIs demonstrated the ability to revert to subsequent MSSA SSTIs with a significant frequency. Of the 215 total patients, 160 (74.4%) had a final status of MRSA positivity, while 55 (25.6%) reverted to MSSA infections for the remainder of the study. The last culture within this 10-year period was considered the patient’s final MRSA status.

A total of 151 (70.2%) remained MRSA positive in all subsequent cultures, while 64 (29.8%) had at least 1 incident of MSSA reversion. Nine of these patients went on to develop future MRSA infections. Eighty-six percent of the patients showing a reversion to MSSA remained MRSA free throughout the study.

Patients with atopic dermatitis showed about the same reversion rate to MSSA as the larger sample, with 2 of 6 (33%) patients reverting. Patients with flaring atopic dermatitis showed similar incidence rates of MSSA reversion to the larger sample and their nonflaring counterparts (1 of 3 patients reverted). The small number of patients with atopic dermatitis analyzed, however, makes firm conclusions difficult.

We next assessed various factors that might increase or decrease the likelihood of reversion (Table 2). Among a variety of possible factors, only 1, the presence of an invasive device, demonstrated a statistically significant risk (relative risk [RR], 1.20; 95% CI, 1.02-1.41; P = .03) toward remaining MRSA positive in subsequent infections. Three other factors demonstrated trends toward remaining MRSA positive: taking immunosuppressive medications (RR, 1.19; 95% CI, 1.00-1.41; P = .17), having a sentinel culture taken as an inpatient (1.15; 0.97-1.36; P = .21), and using an antibiotic within 1 week of a culture (1.13; 0.96-1.33; P = .25). It should be noted, however, that relatively small numbers of patients constituted these various subcategories, and subsequent studies would require larger populations to adequately assess these factors.

The identified MRSA bacteria were not speciated, and it is unknown which were hospital-acquired or community-acquired strains. However, of the 215 total patients, only 36 were inpatients at the time of the sentinel culture. Having a sentinel inpatient culture was not statistically significant in increasing the likelihood of remaining MRSA positive, although it did show a trend toward less likely reversion in the inpatient setting. Not all inpatients had their follow-up cultures in the hospital, but for those who did, the likelihood of remaining MRSA positive was similar to those who had follow-up cultures in the outpatient setting (RR, 0.96; 95% CI, 0.71-1.28; P > .99). Furthermore, the 179 patients with sentinel outpatient cultures similarly demonstrated that having the subsequent culture as an inpatient did not affect the likelihood of remaining MRSA positive (RR, 1.04; 95% CI, 0.74-1.45; P > .99).

The mean time to reversion was 29.7 months (median, 22 months; range, 4-95.5 months). The mean time between first and second cultures was 13.3 months (median, 7 months; range, 0.5-77 months) for patients remaining MRSA positive and 23.2 months (median, 17 months; range, 0.5-95.5 months) for those patients who reverted.

This study demonstrates the tendency of patients with MRSA SSTIs to revert to MSSA infections with considerable frequency (29.8%). The findings reinforce the importance of culturing recurrent purulent infections whenever possible, even in patients with a history of MRSA infections.

Little is known about the basic microbial mechanisms that promote reversion. One possibility is that both MRSA and MSSA strains persist, but that antimicrobial use promotes the growth of MRSA over MSSA, making it the dominant organism in laboratory cultures. After the antimicrobials are withdrawn, MSSA colonies regain dominance. Our data, however, did not show any correlation between antibiotic use within 1 week of culturing and MRSA persistence.

Similarly, a phenomenon has been described with MRSA susceptibility to glycopeptides, in which vancomycin hydrochloride exposure causes the development of a glycopeptide-resistant strain of bacteria, which reverts to glycopeptide susceptibility when vancomycin is discontinued.¹³ A proposed mechanism for the observed reversion is one of extreme selection in which exposure to antibiotics generates colonies that are 100% MRSA, but the withdrawal of antibiotics favors reversion to the MSSA phenotype, with the bacteria mutating back to this state. This has been demonstrated both in vitro and in vivo with glycopeptide-resistant S aureus.^13,14 In 1 study, the resistant strains of bacteria reverted to glycopeptide susceptibility when exposed to nonselective media.¹⁴ No identifiable colonies of glycopeptide-resistant bacteria remained among the reverted populations.

The mechanism for reversion may, alternatively, be that patients clear their MRSA colonization and are reinfected with a new organism. Patients who remain MRSA positive may have persistent colonization or are prone to MRSA reinfection. Contributing to MRSA persistence could be differing host immune responses to MRSA and MSSA, as well as variations in host immunity that predispose certain patients to a MRSA infection. The length of time between the first and second cultures can have implications as to the patients’ comorbidities and immune status as well as the virulence of the MRSA strain. The shorter mean time between first and second cultures in patients who remained MRSA positive may indicate an immunologic vulnerability to more frequent MRSA infections. Although much research has been done on the immune response to S aureus infections, the differences, if any, between the immune response to MRSA and MSSA have yet to be delineated.¹⁵

Also adding to the predisposition to or the persistence of MRSA infections could be the patient’s skin microbiome. Grice and Segre¹⁶ demonstrated the importance and effect of commensal organisms on subsequent cutaneous infections in patients with chronic wounds.

In an outpatient setting, during a 10-week observation, Fernandez et al¹⁷ demonstrated that colonization is associated with an increased risk of infection and suggest a higher conversion rate in patients colonized with MRSA compared with those colonized with MSSA. In addition, molecular similarity between S aureus strains isolated from nasal carriers and their infected wounds has been shown.¹⁸ In contrast, Kauffman et al¹⁹ were unable to relate a decrease in the colonization rate to a drop in the infection rate, although this may have been due to a low rate of infections overall. In summary, the relationship between colonization and infection in outpatient SSTIs is still being delineated.

Other studies tracking systemic infections and colonization in hospitalized patients showed that more than 80% of MRSA carriers decolonized without active treatment,^20,21 although those patients who were actively treated decolonized at a faster rate²⁰ and those with skin wounds had a higher risk of remaining colonized.²² Approximately 50% of MRSA carriers decolonized in 1 year. The rate of decolonization slows after this first year, but 80% of patients can decolonize by 5 years.²¹ This ability of patients to decolonize at a significant rate without treatment may be evidence that subsequent infections are due to a new rather than a persistent organism, particularly when the time between infections is more than 1 year. Interestingly, our data showed that patients reverted to MSSA infections a mean 29.7 months after the sentinel MRSA culture, when the suspected colonization rate would be about 30%, favoring the likelihood of a new colonizing or infecting organism for infections at this point.²¹ As MRSA carriers decolonize, they are more likely to become colonized by MSSA, as shown in the general population.²³ For those patients remaining MRSA positive, it could be hypothesized that they are clearing their MRSA colonization more slowly than average and are being reinfected by the same organism or, alternatively, this might relate to host immune response or their skin microbiome, making them more susceptible to MRSA reinfection. Speciating the organisms in future studies would clarify whether subsequent MRSA infections were recurrent or new bacterial strains.

While the original observations occurred in patients with atopic dermatitis, few were detected in this retrospective medical record review, making it difficult to draw any conclusions regarding their MRSA reversion rates. This may reflect the culturing and treatment practices of our dermatologists, combined with the small number of patients with atopic dermatitis who were MRSA positive. Trends noted in San Diego, California, and Toronto, Canada, suggest that pediatric patients with atopic dermatitis have a lower infection rate with MRSA,^24,25 and the rate of increase of MRSA infections is slower than in the general pediatric population.²⁵

The clinical observation that patients are able to revert to subsequent MSSA infections after having an index MRSA SSTI raises many questions discussed earlier. How this should affect clinician practices remains to be seen, but at this time, reculturing new infections in previously MRSA-positive patients is a prudent management strategy as we recognize that susceptibilities of the organisms change, particularly as elapsed time from the last MRSA culture increases. Miller et al²⁶ demonstrated that it is not possible to predict serious MRSA SSTIs based on the presence or absence of risk factors. This is consistent with our evidence that few, if any, MRSA risk factors showed trends toward persistence. Future areas of research include exploring potential differences in host immune responses to MRSA and MSSA, characterization of the human microbiome of patients with recurrent MRSA vs MSSA infections, and molecular typing of the mec resistance genes for each infection to help indicate if these are recurrent or new infections. Epidemiologic studies that would be of interest include tracking a cohort of patients’ susceptibility patterns of their cutaneous S aureus infections. Furthermore, a cohort of patients with dermatologist-diagnosed atopic dermatitis would be of interest to study S aureus colonization and subsequent infection rates. One of the difficulties in the current study was identifying patients with atopic dermatitis in the database. Having a central database with culture information for patients with atopic dermatitis would greatly aid in the research of their infections.

Limitations of our study include the retrospective medical record review method and access only to university system records. We had hoped to identify patterns of infection in patients with atopic dermatitis but found infrequently documented dermatologic diagnoses, and few patients had seen a dermatologist.

In conclusion, this study demonstrates the ability of patients with MRSA to have subsequent MSSA SSTIs, although the mechanism of this change and whether it demonstrates a reversion or new infection remains to be clarified. We can conclude, however, that SSTIs should be cultured when possible and that further research into this topic could help identify risk factors for subsequent MRSA infections.

Corresponding Author: Anisha B. Patel, MD, Department of Dermatology, University of Texas Medical Branch, 301 University Blvd, Galveston, TX 77555 (anishabpatel@gmail.com)

Accepted for Publication: April 12, 2013.

Published Online: August 14, 2013. doi:10.1001/jamadermatol.2013.4909.

Author Contributions: Drs Patel and Hanifin had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Patel, Simpson, Hanifin.

Acquisition of data: Patel, Hill.

Analysis and interpretation of data: Patel, Simpson, Hanifin.

Drafting of the manuscript: All authors.

Critical revision of the manuscript for important intellectual content: Patel, Simpson, Hanifin.

Statistical analysis: Patel.

Obtained funding: Patel.

Administrative, technical, and material support: Hill, Hanifin.

Study supervision: Hanifin.

Conflict of Interest Disclosures: Dr Hanifin reported serving as a consultant to Chugai Pharma USA, GlaxoSmithKline, Leerink Swann, LEO Pharma, Meda Pharmaceuticals, Merck Sharp & Dohme, Novartis, Pfizer, Valeant Elidel Advisory Board, and ZymoGenetics/BMS. Dr Simpson reported serving as a consultant to Alcimed, Asubio, Brickell Biotech, Clarion Healthcare, Galderma Laboratories, L.E.K. Consulting, Medicis Pharmaceutical Corp, Panmira Pharmaceuticals LLC, Regeneron, and Versant Ventures. Dr Simpson reported serving as a speaker for Galderma Laboratories.

Funding/Support: This study was supported in part by the Department of Dermatology, Oregon Health & Science University.

Role of the Sponsors: The sponsors had no role in the design and conduct of the study; in the collection, analysis, and interpretation of data; or in the preparation, review, or approval of the manuscript.

Additional Contributions: We thank the Department of Dermatology, Oregon Health & Science University, Samuel Bremmer, MD, and Christine Carocci.