A Rheumatologic Approach to Granulomatous Mastitis: a Case Series and Systematic Review

Abstract

Objective:

Idiopathic granulomatous mastitis (IGM) is an enigmatic inflammatory breast disorder. IGM responds to immunomodulatory treatment and may be associated with systemic manifestations such as arthritis and erythema nodosum. These patients are increasingly referred to rheumatologists for management, yet IGM is rarely discussed in the rheumatology literature. The objective of this report is to familiarize rheumatologists with the treatment and systemic manifestations of IGM. We report here a case series of IGM at our institution, and a systematic literature review of IGM treated with methotrexate (MTX).

Methods:

IGM patients at our institution were identified and described using a retrospective chart review. A systematic literature review of PubMed and Google Scholar identified studies of IGM patients treated with MTX.

Results:

We identified 28 IGM patients at our institution. Inflammatory arthritis/arthralgia were present in 4 patients (14%), and 5 (18%) had erythema nodosum. Patients treated with MTX had the highest rates of relapse-free remission; relapse free remission occurred in 4/5 (80%) MTX-treated patients, versus 5/12 (42%) patients treated with steroids alone, and 2/3 (66%) patients treated with steroids and surgery. In the systematic literature review, 116 patients treated with MTX were identified, and the rate of relapse-free remission ranged from 58–100%. Arthritis/arthralgia and erythema nodosum were more common at our institution than reported in the literature.

Conclusion:

MTX is a promising treatment for IGM. Arthritis/arthralgias and erythema nodosum may be under-recognized when IGM patients are managed outside of rheumatology. Prospective studies are needed to characterize clinical features and optimal treatment of IGM.

Idiopathic granulomatous mastitis (IGM) is an inflammatory disease of breast tissue, which occurs primarily in young-middle aged women with a history of recent pregnancy and/or breastfeeding. (1, 2) IGM is a poorly understood disease, and there is no consensus regarding underlying cause, risk factors, or optimal treatment. Historically, breast surgeons have primarily managed this disease, however it is increasingly recognized as an inflammatory disease and is thus being referred to rheumatologists. The primary purpose of this article is to help rheumatologists recognize, understand, and manage this disease. We present here our single- academic medical center experience, including more frequent co-occurrence of arthritis and erythema nodosum than has been previously described, and the successful use of anti-rheumatic drugs in the treatment of IGM. We also report a literature review of the use of methotrexate (MTX) to treat IGM.

IGM typically presents with unilateral or bilateral tender firm breast masses accompanied by erythema, pain, and drainage. These lesions are often initially mistaken for either breast cancer or infection, and most patients undergo breast biopsy. The diagnosis is IGM is made by breast biopsy (ideally by core biopsy) showing a non-caseating granuloma with epithelioid histiocytes and multinucleated giant cells within breast lobules and may contain microabscesses (1, 2). Tissue needs to be sent for bacterial, mycobacterial and fungal cultures to exclude infection. Imaging findings are non-specific and may mimic breast carcinoma and infection. IGM may be unilateral or bilateral, a self-limited or persistent and recurring disease, and may leave disfiguring scars. Because of its rarity, patients often undergo several courses of antibiotics and surgical interventions prior to arriving at the correct diagnosis. (3)

It is not known what causes IGM, however risk factors may include trauma, hormones (oral contraceptives, pregnancy, and birth), breastfeeding (which may include both trauma and hormones), elevated prolactin levels, infections (particularly Corynebacterium), and autoimmunity. (1, 4) There additionally seems to be geographic and ethnic variance in the prevalence of IGM. The majority of large case series come from Asian and Mediterranean countries. (1) In the three largest case series from the USA, the majority of IGM patients were of Hispanic descent and IGM was less commonly seen in Caucasian and African American patients. (5–7) This geographic variation suggests that there could be an underlying unidentified infectious trigger or genetic predisposition, although neither has been identified.

IGM is a diagnosis of exclusion. Tuberculosis (TB) can cause a granulomatous mastitis and needs to be ruled out in addition to other fungal and bacterial infectious causes. Corynebacteria infection has been associated with a cystic neutrophilic granulomatous mastitis, which many authors suggest is a subtype of IGM that can be differentiated based on characteristic histopathology and positive Corynebacterium culture. (4) Rheumatic granulomatous diseases including sarcoidosis and granulomatosis with polyangiitis (GPA) have been reported to cause granulomatous inflammation of the breast, although both are exceedingly rare. (8, 9) It is interesting to note that there are reports of IGM patients with erythema nodosum and arthritis, which raises the question of whether this condition could be a manifestation of sarcoidosis, although other typical features of sarcoidosis, such as hilar lymphadenopathy, are not seen in IGM. Other non-infectious, inflammatory mimics to consider include squamous metaplasia of lactiferous ducts (associated with smoking), fat necrosis (associated with prior trauma or surgery) and foreign body reaction (associated with silicone and sutures). (4)

The optimal treatment of IGM in the literature is highly variable and often contradictory, ranging from surgical resection, observation alone, antibiotic therapy, steroids, and more targeted immunosuppressive therapy such as MTX and biologic therapy. Several small case series have found that MTX is an effective treatment in most patients, which will be reviewed later in this article. At least two larger series from Turkey and China suggested that complete surgical excision alone is the treatment of choice with very low recurrence rates (10, 11), while other authors have suggested that surgery is contraindicated due to poor wound healing. (12) In multiple case series, patients were initially treated with antibiotics but later needed further treatment, suggesting antibiotics alone have a low success rate. (13, 14) Iranian authors have suggested empiric antibiotics and NSAIDS, followed by corticosteroids for intractable disease, and consideration of surgical resection or even mastectomy in those with severe and persistent disease. (15) Finally, a recent large case series of 120 American patients (majority Hispanic) found that observation alone led to a high rate of remission over a mean period of 5 months. (6) However, the vast majority of these data come from the breast surgery literature, and rheumatologists’ perspective and experience is lacking.

Methods:

Retrospective Case Series:

The Oregon Health & Science University (OHSU) Cohort Discovery tool was used to identify subjects seen at OHSU carrying a diagnosis of “granulomatous mastitis”. Patients with known IGM seen in rheumatology clinic were also included. Included patients had been evaluated at OHSU between 2007–2018. A retrospective chart review was used to ensure each subject’s diagnosis was correct. Criteria for this diagnosis include a breast biopsy showing non-caseating granulomas, and evidence that infection was not the underlying etiology of disease. 30 patients were identified and 2 were excluded (1 ultimately felt to have granulomatous mastitis secondary to fat necrosis and 1 due to lack of adequate follow-up).

Clinical data were collected by chart review and included baseline characteristics (ethnicity, age at diagnosis, follow-up period, specialty of treating providers, smoking, gestational history, breastfeeding history, and use of hormonal birth control), IGM lesion details at time of diagnosis (lesion size, bilaterally, infectious studies), symptoms at diagnosis, presence of inflammatory arthralgia/arthritis, and erythema nodosum. Treatment data was collected in the following categories: high dose steroids alone (prednisone >20 mg/day), surgery plus high dose steroids, MTX plus high dose steroids, and other treatments. Remission was defined as a 3-month period without recurrence of symptoms or imaging suggestive of symptom recurrence. Relapse was defined as recurrence of disease after 3 months of remission. Persistent disease was defined as patients who never achieved remission. Patients whose treatment course was still ongoing or who were lost to follow-up were excluded from analysis of treatment courses/outcomes.

Literature review:

A Pubmed and Google Scholar search was done using the search terms “idiopathic granulomatous mastitis AND methotrexate” and “granulomatous mastitis AND methotrexate” to identify patients with IGM who received MTX (alone or in combination with other treatments). All full-length case series, cohort studies, or clinical trials with ≥3 patients were included. Studies were excluded if they did not use MTX as a treatment, if treatment outcomes were not available, if they were not written in English, or if they were abstracts only. Articles focusing only on radiographic or pathologic findings with no report of clinical or therapeutic outcomes were excluded. Review articles were excluded.

Results:

Retrospective Case Series:

Baseline Characteristics

30 patients seen between 2007 and 2019 were identified. Two were excluded (1 diagnosed with alternative condition and 1 without adequate follow-up). All patients were female, with a mean age at diagnosis of 32, the majority (61%) were Hispanic (Table 1). 89% were treated by providers in the surgical oncology/breast health center and 29% were treated by rheumatologists. Over 90% had a history of pregnancy and breastfeeding. Few patients used tobacco (14%) or were on hormonal-based contraception at time of diagnosis (29%).

Table 1:

Baseline Characteristics of 28 patients with Idiopathic Granulomatous Mastitis

Baseline Characteristic1	N (%)
Mean age at diagnosis	32 years (range 16–42)
Mean follow-up period	27 months (range 5–63)
Female	28 (100)
Race/Ethnicity
White, Hispanic	17 (61)
White, non-Hispanic	6 (21)
Asian	2 (7.1)
Other	3 (11)
Specialties of treating MDs2
Surg Onc/Breast Health	25 (89)
Rheumatology	8 (29)
Other3	5 (18)
Smoking (ever)	4 (14)
Gestational history
Ever	26 (93)
In past 5 years prior to symptom onset	19 (83)
Breastfeeding history
Ever	23 (92)
In past 5 years prior to symptom onset	15 (83)
Hormonal-based birth control
Ever	14 (56)
At time of symptom onset	6 (29)

^1.Data not available in the following numbers of patients, as listed by baseline characteristic: mean follow-up period- 2; Gestational history/in past 5 years prior to symptom onset- 5; breastfeeding history/ever- 3; breastfeeding history/in past 5 years prior to symptom onset- 10; hormonal-based birth control/ever- 3; hormonal-based birth control/at time of symptom onset-4

^2.Patients sometimes treated by more than one specialist

^3.Other= primary care (3), general surgery (1), obstetrics-gynecology (1)

Clinical Features of Patients with IGM

Most women had pain on presentation (96%) and many others had overlying erythema (57%; Table 2). Interestingly, over the course the disease 14% had inflammatory arthritis/arthralgia and 18% had erythema nodosum. The presence of both of these features were often inferred by review of the reported symptoms and clinical exam—there was not often documented synovitis or a picture of a rash. The mean max size of the lesion at diagnosis was 4.7 cm and most were unilateral, although 25% had bilateral involvement throughout the course of the disease. 7 patients (26%) had positive bacterial culture and/or gram stain of breast tissue and of these, 2 had positive Corynebacterium cultures and 2 had gram staining with gram positive bacilli that is suggestive of Corynebacterium. Of these 4 patients, 2 had the diagnosis of cystic neutrophilic granulomatous mastitis on pathology. None of the patients who had positive bacterial cultures responded to antibiotics alone. 4 patients had positive tuberculin skin test (TST) and/or interferon gamma release assay (IGRA) (2 with history of treated tuberculosis (TB), 2 with no prior TB treatment) but none of these patients had TB identified on acid fast bacilli (AFB) staining of the breast biopsy tissue. Of the 2 patients with positive TB testing without history of prior TB treatment, 1 patient had positive TST but negative chest x-ray and negative AFB staining of the IGM biopsy specimen and 1 patient had positive TST and IGRA as well as chest x-ray with transient hilar lymphadenopathy and axillary lymphadenopathy. In the latter case, biopsy of both the axillary lymph node and breast tissue were negative for AFB by staining and culture. This patient received treatment for active TB as well as IGM treatment and it is notable that breast symptoms did not completely resolve after TB treatment.

Table 2:

Clinical Features of Idiopathic Granulomatosis Mastitis at Time of Diagnosis in 28 Patients

Clinical Feature1	N (%)
Max diameter at diagnosis (mean)	4.7 centimeters
Bilateral disease
At time of diagnosis	1 (3.6)
Ever	7 (25)
Positive bacterial culture or gram stain
Any positive	7 (25)
Corynebacterium2	4 (14)
Positive tuberculosis3	4 (14)
Symptoms at diagnosis
Pain	27 (96)
Discharge	5 (18)
Erythema	16 (57)
Inflammatory arthritis/arthralgia	4 (14)
Erythema nodosum	5 (18)

^1.Data not available in the following numbers of patients, as listed by clinical feature: max size at diagnosis- 6; positive bacterial culture or gram stain/any positive-4 had no gram stain or bacterial culture although these 4 patients had AFB and fungal staining; positive tuberculosis-2 had no AFB stain or culture

^2.Defined as positive Corynebacterium culture or gram stain showing gram positive bacilli in breast tissue

^3.Defined as positive acid fast bacilli (AFB) culture, interferon gamma release assay (IGRA), tuberculin skin test (TST), and/or chest x-ray (CXR) with suggestive findings.

Treatment Outcomes in IGM patients

Of the 28 patients in this study, treatment outcome data is available for 25 (Table 3). The 3 patients that were excluded had insufficient follow-up to determine outcomes. The majority of patients were treated with high doses steroids alone (n=12), which was defined as a starting dose of >20 mg/day of prednisone. 5 patients were treated with MTX, at a dose of 15–20 mg/week oral. All patients treated with MTX had also received high dose steroids but in all cases, they had not had a complete response to steroids and 4 of these patients had also undergone surgery for IGM lesion. 1 of the patients treated with MTX also had concurrent treatment for active tuberculosis (TB). She had a positive TB interferon-gamma release assay and tuberculin skin test with axillary lymphadenopathy, however biopsy of the axillary lymph node and breast tissue was negative for acid fast bacilli.

Table 3:

IGM Treatment Outcomes in 25 patients

	All treatment groups1 N=25 N(%)	High dose steroids + surgery N=3 N(%)	High dose steroids alone N=12 N (%)	MTX N=5 N(%)	Other2 N=5 N(%)
Relapse-free remission	14 (56)	2 (66)	5 (42)	4 (80)	2 (40)
Relapse	7 (28)	0	5 (42)	0	2 (40)
1 relapse	5 (20)	0	4 (33)	n/a	1 (20)
> 1 relapse	2 (8)	0	1 (8)	n/a	1 (20)
Mean length of time until 1st remission (mo)	7.8	15	7.3	8.3	5.3
Mean length of time until 1st relapse (mo)	11.4	n/a	9.6	n/a	16
Persistent disease	5 (20)	1 (33)	2 (17)	1 (20)	1 (20)
Treatment-free remission for ≥6 mo	20 (80)	2 (66)	10 (83)	4 (80)	4 (80)

^1.All patients received varying courses of antibiotics, without significant clinical improvement except in 1 case (listed in “other”)

^2.Other treatments were: observation, surgery alone, antibiotics/surgery alone, antibiotics alone

In total, 56% of patients had remission without relapse. Of those, patients treated with MTX had the highest rate of remission without relapse (80%). The duration of follow-up for MTX-treated patients ranged from 13–34 months. The one MTX-treated patient with persistent disease actually responded well to MTX and was able to taper prednisone from 40 mg daily to 2.5 mg daily over the course of 4 months. At the end of these 4 months, she had an increase in symptoms, was seen by a breast surgeon and a new 3.3 cm mass was seen in the breast that had been affected by IGM. She was tapered off of all immunosuppressants and she underwent a lumpectomy and was found to have a phyllodes tumor although there was also residual IGM at one of the surgical margins. She has been followed for 15 months off of immunosuppresants, and has become pregnant in the interim. She has some mild residual pain over the left breast and ultrasound shows a residual 1.2 cm nodule as well as heterogeneous areas of decreased echogenicity in the site of prior IGM that may be ongoing residual IGM.

Literature Review:

22 studies were identified, 14 were excluded (9 case reports with <3 patients, 3 review articles, 1 without patient outcomes reported and 1 written in non-English language). (3, 7, 12, 13, 16–19) All studies were retrospective case series except for one, which was a prospective case series. (3) In total, there were 116 patients treated with MTX. The majority of these studies originated from Middle Eastern countries, 1 study was from the United States. 14 (12%) patients had co-existent inflammatory arthritis or erythema nodosum. Several studies used MTX as monotherapy, although most patients were also on steroids. (7, 17, 18) The majority of patients had tried/failed other treatments including steroids, anti-tuberculosis treatment and/or surgery. One study used MTX as initial therapy (16) and in another, 7 out of the 12 patients received MTX as initial therapy. (3) Patients were treated with doses of MTX ranging from 7.5–25 mg weekly. The average rate of relapse free remission was 79% (range 58–100%). Duration of follow-up ranged from 2–36 months (Table 4).

Table 4:

Literature Review of Case Series of IGM Patients Treated with MTX

						Outcomes2
Article	# patients	Country of origin	Mean follow-up (mo)	Extra-mammary manifestations n (%)1	MTX regimen, mg/week	Remission without relapse n (%)	Persistent disease n (%)
Sheybani et al 2015	12	Iran	11.9 +/− 4.4	9 (41%) Inflammatory arthritis/arthralgia 2 (9%) Erythema nodosum	7.5–10 escalate to max of 15	12 (100%)	0 (0%)
Postolova et al 2019	19	USA	36 (12–84)	2 (11%) Pre-existing rheumatologic conditions including tenosynovitis and erythema nodosum	10–15, escalate to max of 25	11 (58%)	1 (5.3%)
Haddad et al 2019	13	Iran	16.4 +/− 9.2	1 (8%) Bilateral ankle arthritis and erythema nodosum	unknown dose	10 (77%)	0 (0%)
Akbulut et al 2011	4	Turkey	2–9	none	7.5	4 (100%)	0 (0%)
Akbulut, S., Arikanoglu, Z. et al 2011	4	Turkey	8.6 (6–14)	none	7.5–15	4 (100%)	0 (0%)
Kim et al 2003	5	Australia	n/a	none	10 –15	3 (60%)	0 (0%)
Aghajanzadeh et al 2015	56	Iran	3–18	none	7.5–10	40 (71%)	16 (29%)
Raj et al 2004	3	UK	6–12+	none	7.5–15	2 (66%)	0 (0%)

^1.Extra-mammary manifestations were based on all patients included in a series, regardless of whether they received methotrexate or not.

^2.Other patients had relapse and then remission, were still on treatment at the conclusion of the study, were lost to follow-up or stopped methotrexate due to side effects

Discussion:

Consistent with previous reports, the patients in our series were young-middle aged females, mostly of Hispanic ancestry, and the vast majority had recent history of childbirth and/or lactation. Interestingly, we found in our cohort a stronger association with inflammatory arthritis and erythema nodosum than has been previously described. A recent systematic review of 3060 cases found the incidence of erythema nodosum to be 8% (20). Our cohort’s higher incidence of erythema nodosum and arthritis may be partly explained by the fact that rheumatologists are more likely to identify these features, however it adds evidence that this disease is a systemic inflammatory condition and may benefit from rheumatologists’ care.

Rheumatologists care for other granulomatous diseases such as sarcoidosis, and are familiar with an arsenal of immunomodulatory therapy, which may not be within the realm of expertise for breast surgeons. While corticosteroids have been used by many as a first line therapy for IGM, they are associated with significant risks when used long-term and it is imperative to have additional treatment options for IGM. The utility of MTX in the treatment of IGM was also highlighted by the systematic literature review and current case series. In our case series, the rate of remission free relapse with MTX treated patients was 80%, which is within the range of 57–100% seen in the systematic literature review

There are several limitations to this report. Our case series was a retrospective analysis and thus some data had to be inferred from chart review. An exact time to remission or relapse could not be determined due to the retrospective nature of the data and the varied intervals at which patients were seen in follow-up. Additionally, precisely defining remission is challenging due to limitations in confidently identifying active IGM on history, exam, and breast imaging, particularly in patients with prior breast surgery or multiple breast conditions.

Infection needs to be excluded prior to the diagnosis of IGM. TB infection is of particular concern in patients from regions where TB is endemic. In our retrospective data, there was not a standardized way that infection was excluded; in many cases, breast tissue was stained for bacterial, fungal and acid fast bacilli, but cultures were not always done. While we feel that none of the cases in this series were likely to have infection as the primary driver of breast disease given lack of response to antibiotics alone and in many cases, an excellent response to immunosuppressant agents, there were 7 patients who had positive gram stain and/or bacterial culture of breast tissue. This could be due to super-infection of inflamed and draining breast tissue rather the primary etiology of disease. We did have 4 patients with confirmed or suspected Corynebacterium on breast tissue. Of these 4 patients, 2 had pathology consistent with cystic neutrophilic granulomatous mastitis. Further studies are still needed to define the role of Corynebacterium in IGM and if cystic neutrophilic granulomatous mastitis is a subset that is best treated with antibiotic therapy.

Although our data suggest that MTX treated patients have a higher rate of remission, randomized prospective studies are urgently needed to understand optimal IGM treatment. The literature review included studies that were also retrospective, not randomized, and were highly variable in terms of diagnostic evaluation, treatment, and follow up. A prospective, randomized trial comparing MTX to prednisone for the treatment of this condition would be very useful. Tumor necrosis factor (TNF) inhibitors also may have a role in treating resistant cases. (21) Lastly, azathioprine should be explored as a treatment, particularly because it can be used in the context of pregnancy and breast-feeding.

IGM is an enigmatic orphaned disease. Patients diagnosed with this condition often suffer through cancer-scares, numerous biopsies, and multiple rounds of unsuccessful treatments. Once they are finally diagnosed, these patients are faced with a frustrating paucity of high-quality data to guide their management, and a lack of certainty who should be treating them. We have found at our center that signs of systemic inflammation (erythema nodosum and arthritis) may be more common than previously thought, and that MTX is a very useful treatment. Together, these findings suggest that rheumatologists can and should be partners in the management of these patients. Management of IGM is an ideal opportunity for interdisciplinary care, leading to timely diagnosis, rapid treatment, and ultimately an increase in expertise and understanding of this challenging disease.

Significance and Innovations:

• IGM is an inflammatory breast disease, which is increasingly seen and managed by rheumatologists.

• Extra-mammary findings such as arthritis and erythema nodosum are more common in IGM than previously thought, and may be more recognizable by a rheumatologist.

• MTX may be an effective treatment for granulomatous mastitis.