Incidence of Methicillin-Resistant Staphylococcus aureus in Postpartum Breast Abscesses

Introduction

Breast abscesses are a relatively rare but serious complication of mastitis. Lactational mastitis is estimated to occur in anywhere from 2% to 33% of lactating women. Current estimates support a lactational breast abscess incidence of 0.1–0.4%, meaning that 10–20% of women with mastitis will develop an abscess.^1–4

Mastitis is a cellulitis of the interlobular connective tissue within the mammary gland. ³ Symptoms of mastitis include temperature of greater than 38.5°C, flu-like aches and chills, and red, painful areas around the breasts. Symptoms seldom appear before the end of the first week postpartum but most commonly present within 6 weeks of delivery. ⁵ Risk factors for mastitis include poor breastfeeding technique, increased stress, sleep deprivation, a previous history of mastitis, and cracked nipples. ⁴ Risk factors for breast abscess development include primiparity, delivery after 41 weeks, and delayed treatment of mastitis.^1,2

Staphylococcus aureus is the most commonly isolated organism from breast abscesses; coagulase-negative Staphylococcus and Streptococcus viridans are less frequent offenders. ⁴ S. aureus is an organism transmitted by contact, and it is therefore postulated that the infecting bacteria are transmitted from the infant's nares to the mother during the skin-to-skin contact of breastfeeding. ³ Abscess formation is more common with S. aureus infection. Clinical suspicion for abscess formation should increase if a patient fails to defervesce within 48–72 hours of antibiotic treatment for mastitis or if patients develop a fluctuant palpable mass.

Treatment of breast abscesses has traditionally comprised antibiotic therapy and incision and drainage in the operating room. Ultrasound-guided drainage has proven to be an effective and less invasive alternative to incision and drainage. High success rates have been reported, especially in small abscesses. ⁶

From 30% to 50% of healthy adults are colonized with S. aureus. From 10% to 20% are persistently colonized. Methicillin-resistant S. aureus (MRSA), which was first isolated in the United Kingdom in 1961, has become an increasingly important healthcare concern. Risk factors for colonization include prolonged hospitalization, ventilatory support, dialysis, residence in long-term care facilities, and indwelling catheters. ⁷ Community-acquired MRSA (CA-MRSA) occurs in patients without these traditional risk factors, and CA-MRSA is found commonly in large cities. CA-MRSA strains have susceptibility patterns with resistance to fewer classes of antimicrobial drugs, including clindamycin, fluoroquinolones, and trimethoprim-sulfamethoxazole.^5,8 CA-MRSA strains have a genotype and phenotype that are distinct from those of hospital-acquired MRSA strains.^4,5

The incidence of MRSA in breast abscesses is not well defined. We attempt here to determine the incidence of MRSA in postpartum breast abscesses in two Houston, TX hospitals.

Subjects and Methods

This study is a retrospective chart review of women hospitalized for postpartum breast abscesses at Memorial Hermann Hospital and at Woman's Hospital in Houston between January 1, 2000 and December 31, 2006. The study was approved by the Institutional Review Board at the respective hospitals. Study participants were identified by searching for admission records containing ICD-9 codes relating to breast abscesses.

Medical records of women with cultured postpartum breast abscesses were reviewed. Data collected were de-identified, and no patients were contacted for data collection. Demographic characteristics, time to presentation, parity, medical history, social history, mode of delivery, culture results, and pertinent procedures were recorded.

Statistical analyses included the Fisher exact test for categorical data and Student's t test for continuous variables. A value of p < 0.05 was considered statistically significant. Data were analyzed by STATA version 7.0 statistical software (Stata Corp., College Station, TX).

Results

Thirty-three abscesses were identified: 19 from Memorial Hermann Hospital and 14 from Woman's Hospital of Texas. Twelve of the 19 abscesses from Memorial Hermann Hospital were MRSA positive (63%), whereas nine of the 14 abscesses from Woman's Hospital were MRSA positive (64%). S. aureus was the only other causative bacterium identified by culture from the remaining abscesses. During this time period 52 women presented to the above institutions with a diagnosis of breast abscess. Of the 19 patients not included in this study, one left the hospital against medical advice, seven abscesses were nonpuerperal, and the remaining 11 patients had no culture results available.

Demographic characteristics of the two groups are listed in Table 1. The average age of our patients was 25.9 years in both groups. Forty-five percent of patients were Caucasian, 39% were African-American, 11% were Hispanic, and 6% were Asian. Most patients had no relevant medical history. One patient had diabetes mellitus, two had hypothyroidism, and one had chronic hypertension. Patients with MRSA breast abscesses presented to the hospital an average of 5.5 weeks after delivery (ranging from 2 weeks to 6 months), whereas those with S. aureus presented within an average of 4 weeks (ranging from 2 weeks to 7 weeks). This difference was not statistically significant with p = 0.35.

Primiparous patients were 67% (n = 18) of our study group, and multiparous patients were 33% (n = 9). Parity information was unavailable for the six remaining women. There was no statistically significant difference in the incidence of MRSA based on parity.

Fifteen patients (45%) were either uninsured or had some form of Medicaid, and 18 patients (55%) had private insurance. There was no statistically significant difference in the incidence of MRSA between patients with different types of insurances.

Mode of delivery was classified as vaginal (which included vacuum, forceps, and vaginal birth after cesarean deliveries) or cesarean. Twenty-two patients were delivered vaginally, and five (19%) by cesarean section. Delivery information for the remaining six patients was unavailable. Again, no statistically significant difference was found between these two groups.

The susceptibility patterns of the isolated strains of MRSA from patients at Memorial Hermann Hospital and at Woman's Hospital are listed in Table 2. As the hospitals had different antibiotic susceptibility testing, not all drugs were tested at each hospital. Those drugs that were not tested at the respective hospital are identified with a dash.

All MRSA isolates were sensitive to rifampin, trimethoprim/sulfamethoxazole, and vancomycin. At Memorial Hermann Hospital, MRSA was likely to be sensitive to fluoroquinolones, gentamicin, and clindamycin but resistant to erythromycin. Woman's Hospital MRSA isolates were resistant to cefazolin, amoxicillin/clavulanate, oxacillin, and ampicillin. They were likely to be sensitive to tetracycline, clindamycin, and gentamicin.

Four of the 33 patients (12%) were treated with ultrasound-guided aspiration of their abscesses, whereas the remainder underwent incision and drainage in the operating room.

Discussion

Results from our review suggest that MRSA is a significant pathogen with an incidence of 63% in postpartum breast abscesses in our population.

The prevalence of S. aureus and MRSA in pregnant women has been reported as 17.1% and 0.5%, respectively, in a 2006 study by Chen et al. ⁸ Women colonized with MRSA are at higher risk for infection by these strains. Interestingly, S. aureus colonization was associated with group B streptococcal co-colonization in this study.

Although no molecular studies were performed in our study, the susceptibility patterns are consistent with CA-MRSA. Most patients were discharged home on clindamycin or a fluoroquinolone. Ciprofloxacin and ofloxacin have been approved by the American Academy of Pediatrics for use in breastfeeding mothers and are reasonable options. Ofloxacin levels in breastmilk are consistently lower than ciprofloxacin levels in breastmilk; therefore, ofloxacin may be preferred if a fluoroquinolone is required. ⁹ Routine nasal culturing of patients with MRSA was not performed at the study institutions during the time of this study.

Most patients from our study underwent incision and drainage in the operating room. Ultrasound-guided drainage, with and without indwelling catheter placement, has proven to be a well-tolerated and successful treatment of postpartum breast abscesses. Because of better cosmetic results, this method is preferred to the more invasive incision and drainage. ⁶ Continued breastfeeding may also be facilitated with less invasive procedures. Despite evidence supporting ultrasound-guided drainage, few women were treated with this less invasive method in our study population.

Limitations of our study include our limited sample size. Using an incidence of 0.1% for breast abscesses, we would expect a yield of at least 70 abscesses over the 6-year study period. Our lower yield is due in part to the fact that women admitted with breast abscesses that were not cultured were not included in the study. Some patients with breast abscesses may have been managed at home; however, we doubt that this practice is frequently used by physicians in our area and would likely not be a reason for our lower incidence. Another explanation could be that an incorrect ICD-9 code was used for their admission. The similarity in percentage of MRSA positivity (63% and 64%) suggests that this is an accurate assessment.

This study involved retrospective chart review, and no information about continued breastfeeding after treatment or prior antibiotic therapy before admission was available. In future studies, information about the impact of different treatment options for breast abscess on continued lactation would be useful.

Conclusions

Despite the limitations in our study, we believe that MRSA is a significant pathogen in postpartum breast abscesses, and a high level of suspicion is warranted. Local susceptibility patterns should guide empirical therapy. ¹⁰ Empirical treatment of mastitis and/or abscesses without culturing breastmilk or infiltrate may not be appropriate in all settings.

Further studies are needed to determine the incidence of MRSA in mastitis and the impact on continued lactation, but until these are done, physicians should remain cognizant of the frequency of MRSA in their own practice setting and should use that knowledge to guide treatment.