Recurrent Community-Associated Methicillin-Resistant Staphylococcus aureus Infections among HIV-Infected Persons: Incidence and Risk Factors

Abstract

Community-associated methicillin-resistant Staphylococcus aureus (CA-MRSA) infections are well described among the general population, but little is known regarding the incidence of and predictors for recurrent CA-MRSA infections among HIV-infected persons. We retrospectively evaluated HIV-infected patients seen at the Naval Medical Center San Diego from January 1, 2000 to June 30, 2007 for wound culture-proven MRSA infections defined as community-associated based on Centers for Disease Control and Prevention (CDC) criteria. Data on skin/soft tissue infections (SSTIs) following an initial CA-MRSA infection were collected by review of medical records and culture results. Patients with or without recurrent infections were compared for predictors of recurrence using multivariate logistic regression models. Thirty-one (6.8%) of 458 patients with HIV had wound culture-proven CA-MRSA SSTIs for an incidence rate of 12.3 infections per 1000 person–years. Those who developed a MRSA infection had a mean age of 40 years, 97% were male, 58% were Caucasian, 23% were Hispanic, 16% were African American, and 3% were other; demographics were similar to the overall study population. Fourteen (41%) HIV patients with an initial MRSA infection had recurrent SSTIs; of these, seven (21%) had culture-confirmed recurrent CA-MRSA. The median time between infection recurrences was 4 months (range, 1–20 months). Suppressed HIV-1 RNA levels of less than 1000 copies per milliliter (odds ratio [OR] 0.14, p = 0.03) was associated with a lower rate of SSTI recurrence. In summary, HIV-infected persons have a high incidence of CA-MRSA skin/soft tissue infections and a high rate of recurrence. HIV control may be associated with a reduced risk of recurrent skin/soft tissue infections.

Introduction

Community-associated methicillin-resistant Staphylococcus aureus (CA-MRSA) infections are well described among the general population and are distinct from health care-associated infections epidemiologically and genetically.^1,2 CA-MRSA infection rates in the general population have steady increased over the last several years and now are the major cause of skin infections seen in emergency rooms across the United States.³ The rapid spread of CA-MRSA, especially the USA 300 clone, is likely due to its smaller staphylococcal cassette chromosome (SCCmec IV), faster doubling time, and perhaps the presence of a novel virulence factor called the arginine catabolic mobile element.^4,5 Most cases occur sporadically, with outbreaks described among athletes, jail inmates, men who have sex with men (MSM), and other groups.⁶

Certain populations have been noted to be at higher risk of CA-MRSA colonization and infection, including HIV-infected persons.^7–10 For example, studies have shown that the rate of CA-MRSA is up to 18-fold higher among HIV-positive persons compared to the general population.^9,10 The reasons for the increased risk of CA-MRSA infections among HIV patients is likely multifactorial. Risk factors for the initial development of CA-MRSA infections among HIV patients have included recent use of β-lactam antibiotics, intravenous drug use, high-risk sexual behaviors, poor immune status, and lack of trimethoprim-sulfamethoxazole prophylaxis.^7,9,11,12

The natural history after an initial MRSA infection remains unclear; most commonly these infections resolve without recurrence.¹³ It has been suggested that HIV-infected persons may have a high recurrence rate of MRSA infections, despite appropriate initial treatment.¹⁴ Given the paucity of data on the incidence of and predictors for recurrent CA-MRSA skin and soft tissue infections (SSTIs) among HIV-infected persons, we performed a retrospective study to evaluate the natural history of CA-MRSA infections in this population.

Methods

We conducted a retrospective study among all HIV-infected patients attending our HIV clinic at the Naval Medical Center San Diego from January 1, 2000 to June 30, 2007 for wound culture-proven MRSA infections. All patients in our clinic were military beneficiaries (active duty members, retirees, or dependents) with open and free access to medical care. Our patient population consists mainly of persons who acquired HIV sexually; less than 1% use illicit drugs due to the presence of military policies and mandatory drug screening procedures.

MRSA infections were defined as community-associated (CA) based on the Centers for Disease Control and Prevention (CDC) criteria consisting of a positive culture for MRSA among persons outside of hospitals or health care facilities, or within 48 hours of admission. In addition, cases had not recently been hospitalized or undergone a medical procedure (e.g., dialysis, surgery, catheter) within the past year and had no history of MRSA prior to the study period.

Cases of MRSA were identified by searching computerized microbiologic results for any culture positive for MRSA. Subsequent infections following the initial MRSA infection were identified by computerized culture results as well as review of all subsequent clinic visits contained within the patient's medical record. Collected data included all wound culture data; number, type and location of SSTIs; demographics (age, gender, and self-reported race); CD4 cell counts; HIV-1 RNA levels; receipt of highly active antiretroviral therapy (HAART); antibiotic use; as well as incision and drainage procedure as part of the initial therapy for the first MRSA event.

Statistical analysis included descriptive statistics. Patients with or without recurrent infections were compared for predictors of recurrence using Fisher's exact and rank sum tests. Variables with a p value <0.20 were placed in a multivariate logistic regression model to evaluate for factors associated with recurrent SSTIs. Correlations between variables were assessed using Pearson's correlation tests. All analyses were conducted using STATA 9.0 (StataCorp, College Station, TX).

Results

Thirty-one (6.8%) of 458 patients with HIV had wound culture-proven CA-MRSA infection. The incidence rate of CA-MRSA among HIV-infected persons in our study was 12.3 infections per 1000 person–years (PYs). HIV-infected persons who developed a MRSA infection were a mean age of 40 years, 30 (97%) were male, and race was reported as Caucasian among 58%, Hispanic among 23%, African American among 16%, and other in 3%, which was not significantly different from the demographics of our study population. The mean CD4 cell count 445 cells/mm³ (range, 10–1348), HIV-1 RNA level was less than 1000 copies per milliliter in 48%, and 71% were receiving HAART at the time of the MRSA infection. The location of the initial MRSA infection was the lower extremity in 10 (32%), buttocks/scrotum in 8 (26%), upper extremity in 5 (16%), face in 4 (13%), trunk in 3 (10%); 1 patient had diffuse folliculitis (3%).

Fourteen (41%) of the HIV-infected patients with an initial MRSA infection had a recurrent SSTI. All recurrences occurred in unique location from the initial infection. Seven (21%) of these patients had a culture-confirmed recurrent CA-MRSA infection. The incidence rate for developing a recurrent SSTI after an initial CA-MRSA infection was 279 cases per 1000 PYs and for a recurrent culture-proven MRSA SSTI was 120 cases per 1000 PYs. The total number of culture-confirmed MRSA SSTIs was 48, and the number of SSTIs per patient ranged from 2–8. The median time between recurrent infections was 4 months (range, 1–20 months). Antibiotic resistance developed from time of the initial isolate to the subsequent isolate in two cases: one developed tetracycline resistance and one erythromycin resistance.

Patients with MRSA who developed recurrent SSTIs compared to those who did not develop recurrence had lower CD4 counts (334 versus 502 cells/mm³, p = 0.05), higher HIV-1 RNA levels (4.3 versus 1.8 log₁₀ copies per milliliter, p = 0.07), and lack of incision and drainage procedure at the time of the initial MRSA infection (65% vs. 36%, p = 0.15) in the univariate models (Table 1). Furthermore, more patients with a recurrent SSTI had a positive nares culture, but this was likely related to the differential testing practices after infection recurrence developed. Statistical analyses showed that CD4 cell count and HIV-1 RNA level were highly correlated (r = −0.41, p < 0.05), hence CD4 cell count was omitted in the final model. In the final multivariate model, HIV-1 RNA level less than 1000 copies per milliliter (OR 0.14, p = 0.03) was associated with a lower rate of SSTI recurrence. There was a trend toward having undergone an incision and drainage procedure at the time of the initial MRSA infection and a lower rate of recurrence (OR 0.19, p = 0.07).

Table 1.

Univariate Analyses of HIV-Infected Persons with or Without Recurrent Skin/Soft Tissue Infections (SSTIs)

Characteristic	Recurrent SSTI n = 14 (%)	No recurrence n = 17 (%)	p value
Age (median, 95% CI), years	40.5 (37, 47)	39 (32, 47)	0.87
Gender, male	14/14 (100)	16/17 (94)	1.0
Race			0.32
Caucasian	7/14 (50)	11/17 (65)
African American	4/14 (29)	1/17 (6)
Hispanic	3/14 (21)	4/17 (23)
Other	0/14 (0)	1/17 (6)
Nares culture, positivea	7/14 (50)	2/17 (12)	0.04
Use of mupirocin	8/14 (57)	11/17 (65)	0.72
Trimethoprim-sulfamethoxazole   prophylaxis (yes/no)	0/14 (0)	1/17 (6)	1.0
Nadir CD4 cell count (median, 95% CI), cells/mm3	188 (76, 269)	261 (130, 323)	0.28
CD4 cell count at initial MRSA infection   (median, 95% CI), cells/mm3	334 (241, 403)	502 (79, 431)	0.05
HIV log10 viral load at initial MRSA infection   (median, 95% CI), copies/mL	4.3 (1.7, 4.6)	1.8 (1.7, 3.6)	0.07
Viral load <1000 copies/mL	4/14 (29)	11/17 (65)	0.07
Receipt of HAART at initial MRSA infection (yes/no)	8/14 (57)	14/17 (82)	0.23
Initial use of an antibiotic effective against   MRSA isolate (yes/no)b	11/13 (85)	16/17 (94)	0.57
Incision and drainage at time of initial MRSA infection	5/14 (36)	11/17 (65)	0.15

^aOnly taken on a subset; culture recorded was taken anytime with a larger number of cultures performed only among patients with recurrence.

^bData missing on one patient.

CI, confidence interval; MRSA, methicillin-resistant Staphylococcus aureus; HAART, highly active antiretroviral therapy.

Discussion

Our study confirms the high incidence rate of CA-MRSA infections among HIV-infected persons. Additionally, this study demonstrated that HIV-infected persons have a high rate of recurrent skin and soft tissue infections after the initial MRSA infection. This study is novel in it specifically examined factors associated with recurrent MRSA infections and suggested that improved HIV control may be associated with a reduced risk of recurrent infections.

The incidence of MRSA SSTIs infections among HIV-infected persons in our study was 12 cases per 1000 PYs. This number is similar to data from an HIV cohort in Chicago (9 cases per 1000 PYs), suggesting the rate of CA-MRSA infections among HIV-infected patients is substantially higher than the general population rate of 1–2 MRSA infections per 1000 PYs.¹⁰ At our institution, a study of the incidence of MRSA infections was conducted (1990–2004) among the general population served at our medical center¹³; during 2004, the incidence rate of CA-MRSA infections was 1.6 cases per 1000 PYs. This also implies that HIV-infected persons have a higher rate of MRSA infections, although the time periods of the evaluations varied. Prior studies have shown that immunodeficiency, high-risk sexual behaviors, use of illicit drugs, and increased use of antibiotics may account for the heighten risk of initial MRSA infections in the HIV population.^9,11,12

Our study showed that HIV-infected persons have a high risk of recurrent skin and soft tissue infections after the initial MRSA infection. Among our cohort of HIV positive persons, 41% developed a new SSTI and 21% had culture-proven recurrent CA-MRSA. Compared to the general population at our medical center who had a recurrence rate of approximately 9% during a prior investigation,¹³ HIV-infected persons appear to be at higher risk for recurrent CA-MRSA infections. Other studies have shown that HIV patients with CA-MRSA often have a history of SSTIs and/or CA-MRSA.^11,14–16 In our study, recurrent infections occurred relatively soon (median of 4 months) after the initial infection. In all cases, the first infection had completely resolved, and recurrence was noted at a novel site suggesting reinfection as the etiology. These data suggest that prevention strategies to reduce recurrent MRSA infections are needed.

The factors associated with recurrent SSTIs among HIV positive persons in our study included higher HIV-1 RNA levels, suggesting that better HIV control may be important in the prevention of recurrent infections. Secondly, initial performance of an incision and drainage procedure had a trend toward reduced risk of SSTI recurrence. Other studies also have noted the importance and effectiveness of incision and drainage in SSTI management;¹⁵ prospective studies are needed to further investigate this finding. The use of an antibiotic effective against MRSA, however, was not significantly associated with a reduced risk of recurrence, concurrent with other studies.¹⁴ Of note, in our population, the majority of patients received an effective antibiotic to treat MRSA.

Other possible factors associated with recurrent infections include persistent colonization. Several studies have shown that HIV patients have an elevated risk of nasal carriage of MRSA.^7,8 The high rate of perineal infections in our study suggests the possibility of MRSA carriage at this site, as well, and perhaps that sexual activities are related to these infections.¹⁷ Studies examining novel carriage sites in addition to the effectiveness of decolonization strategies and their impact on recurrent infections are needed. Finally, trimethoprim-sulfamethoxazole use (typically for Pneumocystis prophylaxis) has been associated with a reduction in MRSA colonization and infection in other studies.^11,12,18 Although it is likely that this agent would reduce recurrent SSTIs, our study did not demonstrate a protective effect, likely due to the small number of patients (n = 1) receiving this medication during the HAART era.

The clinical significance of recurrent MRSA infections is numerous. The morbidity, increased costs, and potential mortality related to these infections are noteworthy. Furthermore, persons experiencing recurrent MRSA infections may serve as an important reservoir for MRSA within the community. Finally, Diep et al.¹⁶ recently demonstrated a strong relationship between a history of MRSA infection and infection with a multidrug-resistant USA 300 strain. In our study, two MRSA isolates acquired resistance from the time of the initial infection to recurrence. Trends in increasing resistance may evolve particularly in the setting of recurrent infections and repeated course of antibiotics.

As with all studies, our study had potential limitations. This study was retrospective, so causality could not be temporally established between patient characteristics and recurrent MRSA infections. However, our study does provide important insights into factors which may be associated with MRSA infections, which may be validated by prospective investigations. While our study evaluated a large HIV-positive cohort, we did not have data available among HIV-negative persons during the exact same study period for direct comparison; however, information collected at our institution as part of an earlier study¹³ was discussed. Although our study showed that elevated HIV-1 RNA levels are associated with recurrent MRSA infections, it is unclear if HIV control itself is related or if this is a marker for other conditions or behaviors. Finally, although our study examined a large HIV clinic, the overall numbers of MRSA infections were small; larger studies are advocated.

In summary, HIV-infected persons appear to have an elevated risk of initial and recurrent CA-MRSA skin and soft-tissue infections. Improved HIV control may be associated with a reduced risk of CA-MRSA infections in this population; prospective studies on preventive strategies are needed.

Acknowledgments

The opinions or ascertains contained herein are the private views of the authors and are not to be construed as official or as reflecting the views of the Departments of the Army, Navy, or Air Force, or the Department of Defense. The authors have no commercial or other association that might pose a conflict of interest in this work.

Support for this work was provided by the Infectious Disease Clinical Research Program (IDCRP), Uniformed Services University of the Health Sciences (USUHS), Bethesda, MD, of which the TriService AIDS Clinical Consortium (TACC) is a component. The IDCRP is a DoD tri-service program executed through USUHS and the Henry M. Jackson Foundation for the Advancement of Military Medicine in collaboration with HHS/NIH/NIAID/DCR through Interagency Agreement HU0001-05-2-0011.

Author Disclosure Statement

No competing financial interests exist.