Read an Excerpt

BMJ Clinical Review General Surgery

BPP Learning Media Ltd

Treatment of breast infection

J Michael Dixon, professor of surgery and consultant surgeon, Lucy R Khan, specialty registrar breast surgery

A cohort study of American women reported that 10% of women who breast feed have mastitis, and a recent Cochrane review reported the incidence to be as high as 33%. Breast abscesses are seen less often, but when they do develop delays in referral to a specialist surgeon may occur. A recent survey in the United Kingdom found that many surgical units have no clear protocols for managing patients with breast infection who are referred to hospital. Some surgeons aspirate breast abscesses under local anaesthesia, whereas others use general anaesthesia. The management of breast infection has evolved over the past two decades, with advances in both diagnosis and treatment. A new concept is bedside ultrasound, and this plays an important part in current management.

We review management of breast infection in the primary care setting and after hospital referral. The review is based on our current practice and the best quality evidence available. Few randomised controlled trials deal with this topic, and most breast specialists have adopted their own protocols for clinical management, loosely based on published algorithms, and largely dictated by their specific patient population and their clinical practice setting. This review provides a resource for those who see breast infection infrequently. Appropriate timely referral will help avoid unnecessary morbidity for patients.

What kinds of breast infection are there?

Infection can occur in the parenchyma of the breast or the skin overlying the breast (fig 1). Parenchymal breast infections can occur in lactating and non-lactating breasts. One cross sectional analysis of 89 patients with breast abscesses requiring surgical intervention found that 14% were lactational and 86% were non-lactational.

Which micro-organisms are implicated?

An up to date retrospective case series shows that during lactation the most common organism responsible is Staphylococcus aureus, including strains of meticillin resistant S aureus (MRSA), particularly if the infection was acquired in hospital. Other organisms responsible include streptococci and Staphylococcus epidermidis. Organisms responsible for non-lactating breast infections include bacteria commonly associated with skin infections but also include enterococci and anaerobic bacteria such as Bacteroides spp and anaerobic streptococci. Patients with recurrent breast abscesses have a higher incidence of mixed flora (20.5% in those with recurrence v 8.9% with a single episode), including anaerobic organisms (4.5% v 0%).

Investigating and managing breast infection in lactating women

Who gets it and how do they present?

Lactating breast infection is most commonly seen within the first six weeks of breast feeding, although it can develop during weaning. The infection arises initially in a localised segment of the breast and can spread to the entire quadrant and then the whole of the breast if untreated.

A review of 946 cases of lactational mastitis in the United States found that women often gave a history of difficulty with breast feeding and many had experienced engorgement, poor milk drainage, or an excoriated nipple. Population based studies have shown that risk factors for abscess formation include maternal age over 30 years, gestational age greater than 41 weeks, and a history of mastitis. The examining doctor may see erythema, localised tenderness, localised engorgement, or swelling. Some women present with fever, malaise, and occasionally rigors.

A cohort study estimated that 2-10% of breastfeeding women get mastitis but only 0.4% develop an abscess. A prospective study of 128 women reported that 5-10% of women with mastitis developed a breast abscess, possibly because of suboptimal management of their mastitis.

How to treat mastitis

Guidelines from the World Health Organization and numerous reviews of the condition recommend treating lactating women with mastitis by prescribing appropriate oral antibiotics and encouraging milk flow from the engorged segment (by continuation of breast feeding or use of a breast pump). Such measures reduce the rate of abscess formation and thereby relieve symptoms. A Cochrane review found only one reported randomised trial of antibiotic treatment versus breast emptying alone conducted among women with lactational mastitis that showed faster clearance (mean 2.1 v 4.2 days) of symptoms in women using antibiotics. Oral antibiotics are usually sufficient, and only rarely do patients with sepsis require hospital admission and intravenous antibiotics. Lactating infection can be treated by flucloxacillin, co-amoxiclav, or a macrolide such as erythromycin or clarithromycin (in patients who are allergic to penicillin), given for at least 10 days. Tetracycline, ciprofloxacin, and chloramphenicol should not be used to treat lactating breast infection because these drugs can enter breast milk and harm the baby.

One report of using Lactobacillus fermentum and Lactobacillus salivarius as an alternative treatment has shown them to be as effective as antibiotics. Further studies are needed before they can be used as an alternative to appropriate antibiotics.

There is an alarming trend towards believing that fungi are important in the aetiology of breast infection and deep breast pain associated with breast feeding, despite a lack of good quality evidence. The prescription of antifungals, such as fluclonazole, is common despite the lack of good quality clinical evidence to support their use.

A case series describes several patients with breast pain during breast feeding who did not have mastitis but Raynaud's disease of the nipple and who responded to nifedipine. Prescription of anti-inflammatory drugs and the application of cold compresses or ice packs can help to alleviate pain. One small trial compared the effectiveness of chilled or room temperature cabbage leaves with ice packs and both produced identical symptom relief.

We have found that it is not uncommon for patients to be referred late to hospital with established large volume abscesses (fig 2). Reasons for this include failure to refer infection that does not settle rapidly after one course of antibiotics; a lack of continuity of care in the community; use of inappropriate antibiotics; and delays as a result of using other treatment modalities, such as antifungal agents and cold compresses alone.

Investigating a suspected breast abscess

Ultrasound will establish the presence of pus and should be performed in any patient whose infection does not settle with one course of antibiotics, whether a breast abscess is suspected or not (fig 3). Even when clinical examination shows obvious signs of an abscess, ultrasound is useful because it may identify more than one collection of pus that might otherwise be missed.

Draining an abscess

In our specialist practice we have developed and evaluated the following approach to the management of breast abscesses. We base our approach to draining the abscess on the appearance of the skin overlying the abscess (fig 4).

If the overlying skin is normal, we recommend aspiration of the abscess under ultrasound guidance using adequate local anaesthesia. A 21 gauge needle is introduced through the skin some distance away from the abscess and 1% lidocaine with 1:200 000 adrenaline is infiltrated into the skin and into the breast tissue under ultrasound image guidance. When reaching the abscess cavity (fig 3B), if the pus is thin enough it can be aspirated with the same needle. Once the pus has been aspirated the syringe is changed and the abscess cavity is irrigated with as much as 50 mL of 1% lidocaine and adrenaline. On ultrasound imaging the abscess cavity should be seen to expand and collapse as fluid is injected and aspirated to dryness (fig 3C).

If the pus is very thick and cannot be aspirated through a 21 gauge needle, then having waited for local anaesthetic to be effective, a larger gauge needle may be advanced through the skin and breast tissue into the cavity. The pus is diluted with local anaesthetic and adrenaline, after which this is aspirated. We find that using a combination of lidocaine and adrenaline in solution reduces pain and minimises bleeding and subsequent bruising. Irrigation is continued until all the pus is aspirated and the fluid used to irrigate comes back clear. The net effect of this procedure is to control pain by a combination of providing local anaesthesia and reducing the pressure within the abscess cavity by aspirating all the pus. We send a sample of pus to the microbiology department for culture and continue appropriate oral antibiotics and analgesia until the abscess resolves.

We review the patient every two to three days and repeat aspiration under ultrasound guidance if fluid is present in the abscess cavity. We continue with this approach until no further fluid is visible in the abscess cavity or the fluid aspirated does not contain pus. Few abscesses require more than two to three aspirations, although very large collections may require more. Characteristically, the fluid aspirated changes from pus to serous fluid and then to milk over a few days. Most abscesses in lactating breasts can be managed successfully in this manner.

If the skin overlying the abscess is compromised and is thin and shiny or necrotic we perform mini-incision and drainage (fig 5). Local anaesthetic is infiltrated into the skin overlying the abscess and left for a minimum of seven to eight minutes, and then a small stab incision with a number 15 blade is made into the abscess over the point of maximum fluctuation. If the point of maximum fluctuation is not clear, ultrasound can help to define the best site for incision. We excise any necrotic skin. Once the contents of the abscess cavity are drained, we irrigate the cavity thoroughly with local anaesthetic solution and repeat every two to three days until there is no evident leakage from the abscess, the wound closes, and no further pus is draining. Most patients whose abscess needs to be incised and drained can have the procedure performed under local anaesthesia in the outpatient clinic. Large incisions are not necessary to drain breast abscesses, and the cosmetic results of the small incisions needed are usually excellent. The placement of drains and insertion of packing have no role in the modern day management of breast abscesses.

If infection fails to regress with appropriate management, carry out further imaging combined with needle core biopsy of any suspicious abnormality to exclude an inflammatory cancer.

Breast feeding after breast infection

Although women are encouraged to continue breast feeding after treatment of mastitis or an abscess, it may be difficult to do so from the affected side. If the infant cannot relieve breast fullness during nursing, the woman may use hand expression or a breast pump to encourage and maintain milk flow until breast feeding can resume. Although most women are able to continue breast feeding even if they have excoriation of the nipple and pain, a few experience continuous and disabling pain (fig 6). If after discussion a woman chooses to stop breast feeding so that the breast infection can be controlled and the breast can heal, lactation can be suppressed using cabergoline.

Investigating and managing breast infection in nonlactating women

Who is at risk?

People at highest risk of developing an infection of breast tissue when not lactating are those who smoke and those with diabetes. A recent retrospective analysis found that patients with non-lactating skin associated abscesses who have diabetes or who smoke (or both) are likely to have recurrent episodes of breast infection. Infections are categorised as central or subareolar infections and peripheral infections — each has different causes and treatments. Infections that occur in the skin of the breast are usually secondary to an underlying lesion such as a sebaceous cyst or hidradenitis suppurativa.

Types of infection

Central or subareolar infection

This is usually secondary to periductal mastitis, a condition in which the subareolar ducts are damaged and become infected, often by anaerobic bacteria. Patients may present initially with subareolar inflammation (with or without an associated mass) or with an established abscess (fig 7A). Associated features include nipple retraction and a discharge from the nipple. Periductal mastitis predominantly affects young women, the average age being 32 years, and smoking is a major causative factor, with 90% of patients being smokers. Periductal mastitis and can also occur in men. Substances in cigarette smoke — such as lipid peroxidise, nicotine, and cotinine — concentrate in the breast and are found at much higher concentrations in subareolar ducts than in plasma. Either the toxic substances in cigarette smoke damage the ducts directly or local hypoxia causes subareolar duct damage and subsequent inflammation and infection. Patients with periductal mastitis can have bilateral disease, and some women present with bilateral fistulas and nipple changes on both sides. Smokers who have nipple piercing can develop persistent and troublesome infection. Breast abscesses can affect men as well as women.

Peripheral non-lactating infection

This is less common than central infection. Peripheral infection has been associated with diabetes, rheumatoid arthritis, steroid treatment, trauma, and granulomatous lobular mastitis but often there is no underlying cause. Occasionally, comedo ductal carcinoma in situ can become infected and present with inflammation or as an abscess; we therefore recommend that patients over 35 years with peripheral infection and no obvious cause undergo bilateral mammography once the infection has resolved.

Granulomatous lobular mastitis

One cause of peripheral infection is granulomatous lobular mastitis, a condition of unknown aetiology. It can present as a peripheral inflammatory mass that masquerades as cancer or as an area of infection with or without overlying skin ulceration. Although this condition mostly affects young parous women, who develop multiple and recurrent abscesses, it is seen in nulliparous women as well. It has been suggested that Corynebacterium spp play a part in this condition, but antibiotics effective against these organisms rarely lead to resolution of disease and thus they are unlikely to have a major aetiological role.

Skin associated infection

Sebaceous cysts are common over the skin of the breast and these can become infected to form local abscesses. Cellulitis of the breast with or without abscess formation is common in patients who are overweight, have large breasts, or have had breast surgery or radiotherapy. It occurs in the lower half of the breast and also under the breast where sweat accumulates and intertrigo develops. Intertrigo may be a recurrent problem in women with large ptotic breasts. Staphylococcus aureus is the usual causative organism. Although antifungal creams are commonly prescribed, there is no evidence that fungi play an aetiological role in this condition. Hidradenitis suppurativa commonly affects the axilla and groin and can also affect the skin of the lower half of the breast, resulting in recurrent episodes of infection and abscess formation.

---

Excerpted from BMJ Clinical Review General Surgery by Gopal K Mahadev, Eleftheria Kleidi. Copyright © 2015 BPP Learning Media Ltd. Excerpted by permission of BPP Learning Media Ltd.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.

---