Primary breast lymphoma: A single‐centre experience

Abstract

Background

Primary non‐Hodgkin lymphoma is an extremely rare entity, and this condition represents less than 0.5% of all malignant lesions involving the mammary gland. As such, there has been a paucity of relevant clinical data arising from Southeast Asia.

Aims

Our study aims to review the clinical presentation, diagnostic methods, treatment, and survival outcomes of all patients diagnosed with primary breast lymphoma in our institution between 2011 and 2017.

Methods and results

Patients who had histologically proven lymphoma involving the breast were identified from a prospectively collected database in a single institution between 2011 and 2017. All seven patients were female, with a median age of 65 years old, and had presented with unilateral large breast or axillary masses. All the histological diagnosis was achieved with adequate tissue diagnosis either through core, incisional, or excisional biopsy. Five patients had diffuse large B cell lymphoma, one had marginal zone lymphoma, and the other had follicular lymphoma. Based on Ann Arbor classification, one patient had stage 1, three had stage 2, one with stage 3, and two patients with stage 4 disease. Five patients had received standard CHOP regimen with rituximab. At the time of analysis, patients who had nondisseminated disease had a median survival of 57 months. The overall mean survival time for all seven patients was 47 months. With the standard systemic chemotherapy treatment regimen, the estimated 3‐year overall survival was found to be 64%.

Conclusion

Primary breast lymphoma, though uncommon, may present in a similar manner as breast carcinomas, but the main treatment modality remains nonsurgical with systemic chemotherapy. Hence, it is prudent to obtain accurate histological diagnosis of primary breast lymphoma. In this study, our patients with nondisseminated breast lymphoma have demonstrated a fairly good survival outcome following chemotherapy.

1. INTRODUCTION

Primary breast lymphoma (PBL) is a rare entity of malignant breast tumours which represents about 0.5% of breast cancers1, 2 and less than 1% of non‐Hodgkin's lymphoma.3, 4 Breast lymphoma often closely resembles breast carcinoma, both clinically and radiologically.5, 6

The standard definition of PBL was devised by Wiseman and Liao which consists of four criteria: (i) adequate histopathologic specimen, (ii) presence of breast tissue and lymphomatous infiltrates in close association, (iii) no evidence of disseminated lymphoma at diagnosis, and (iv) no previously known extramammary lymphoma disease at diagnosis.7 Involvement of ipsilateral axillary lymph node(s) is allowed.7

The commonest subtype of PBL is diffuse large B cell lymphoma (DLBCL)1, 2, 6, 8, 9, 10 which consists of 40% to 70% of cases.10 Other subtypes include marginal zone lymphoma, follicular lymphoma, and mucosa‐associated lymphoid tissue (MALT) which are less common.2

Multiple modalities have been used for the treatment of PBL, including chemotherapy, radiotherapy, and surgical resection. It has been established that mastectomy does not confer any treatment benefit1, 8, 11, 12, 13; hence, its use has declined since 1990s.6 Systemic chemotherapy remains the mainstay of treatment for PBL, with the consideration of adjuvant radiotherapy in patients with disease involvement in axillary lymph nodes.12

To date, there has been very limited data regarding the incidence, treatment outcomes, and prognosis of patients diagnosed with primary breast lymphoma in Southeast Asia.

2. AIM

This study aims to determine characteristics, diagnostic methods, treatment, and survival outcomes of all patients diagnosed with primary breast lymphoma in our institution between 2011 and 2017.

3. METHODS

3.1. Patient enrolment

A database for patients diagnosed with breast lymphoma in Changi General Hospital was created in January 2011. Every subsequent patient with confirmed histological diagnosis of breast lymphoma was added into this database. Approval for this study from the Institutional Review Board was obtained. Patients in the breast lymphoma database were reviewed in June 2017 for eligibility for this study. Patients presenting with breast or axillary mass with confirmed histological diagnosis of breast lymphoma are eligible to be included in this study. Patients with known or previous history of lymphoma are excluded. Seven patients from the database were identified to fulfil the inclusion and exclusion criteria. These patients were approached for recruitment for this study at their subsequent outpatient clinic visit or during their inpatient stay. All seven patients were agreeable to participate, and written informed consent was obtained.

A data collection spreadsheet was completed for each case through a retrospective review of their medical records. Descriptive statistics were used to summarize each case. These included patient demographics, tumour characteristics and histology, staging, treatment parameters, and outcomes. The diagnosis of PBL was based on Wiseman and Liao criteria.

All patients included in this study were followed up from time of diagnosis until death or time of analysis. The median follow‐up period was 25 months (range 2‐65 months). June 2017 was determined to be the time of analysis, when a review of the breast lymphoma database was performed.

3.2. Staging

All patients had presented with either a breast or axillary mass, and they had undergone initial radiological imaging such as mammogram and ultrasound. Histological confirmation of the mass was obtained with either core, incisional, or excisional biopsy. Staging investigations had included computed tomography scan (thorax, abdomen, and pelvis) or photon emission tomography (PET) scan. All were subsequently staged according to the Ann Arbor classification.14 International Prognostic Index (IPI)15 and age‐adjusted IPI scores for each individual were also calculated.

3.3. Treatment and survival analysis

The chemotherapy regimen received by each patient was examined. Response to chemotherapy was evaluated with follow‐up PET scan and assessed as either remission, progressive, or complete response. Overall survival (OS) was measured from time of diagnosis to time of analysis or death. The OS rate was calculated by the Kaplan‐Meier method using the SPSS statistical software, version 19.0.

4. RESULTS

4.1. Cohort description

This study comprised of seven female patients with no history of breast malignancy or lymphoma. The demographics and characteristics of the seven patients are shown in Table 1. They had a median age of 65 years at time of diagnosis. At the time of analysis, two patients had passed away.

Table 1.

Patient demographics and characteristics

Patient Number	Age at Diagnosis (years)	Sex	ECOG Grade	Presenting Complaint	Side	Primary Cancer	Ipsilateral Nodes	Other Nodes	Other Visceral Involvement	Initial Tumour Size (cm)
1	23	F	0	Axillary lump	Right	Yes	Yes	Yes	No	5.2 × 2.5
2	60	F	1	Breast mass	Right	Yes	Yes	No	No	7 × 6.9
3	50	F	0	Breast mass	Left	Yes	No	No	No	3 × 3.2
4	76	F	1	Axillary lump	Left	Yes	Yes	Yes	No	1.4 × 0.8
5	74	F	1	Breast mass	Left	Yes	Yes	No	No	8.3 × 4.7
6	65	F	2	Breast mass	Left	Yes	Yes	Yes	No	4.8 × 4.5
7	74	F	2	Breast mass	Right	Yes	Yes	Yes	Yes	8 × 10.3

Abbreviations: F, female; ECOG, Eastern Cooperative Oncology Group.

4.2. Presentation

The most common clinical presentation was breast lump with a frequency of 71.4%, followed by axillary lump with a frequency of 28.6%. Left breast involvement was more frequent than the right (57.1% vs 42.9%). All patients presented with unilateral lesion, and none had bilateral disease involvement. Four patients had a tumour size larger than 5 cm. The lumps' sizes ranged from 3 to 8.3 cm. Out of the five patients who presented with breast mass, three had fungating breast tumours and the other two had painless, firm breast lumps. Out of these two patients, one presented with erythematous and inflamed skin overlying the lump. None had complaints of nipple discharge nor nipple inversion. None had peau d'orange appearance on their breasts.

4.3. Diagnosis and pathological features

As shown in Table 2, histology was obtained from these patients either through core, incisional, or excisional biopsy. All of them had B cell lymphoma. Two had indolent lymphoma subtypes, and the remaining five had aggressive lymphoma subtype. Five out of seven patients had DLBCL. One other patient had MALT (14.3%), and the other had follicular (14.3%) subtype.

Table 2.

Patient characteristics

Patient Number	Method of Diagnosis	Histology	Ann Arbor Classification	LDH Level on Diagnosis	IPI	Age‐Adjusted IPI
1	Excisional biopsy	Grade 1 follicular lymphoma	IIIE	354	‐	‐
2	Excisional biopsy	DLBCL	IIE	397	1	0
3	Incisional biopsy	Marginal zone lymphoma (MALT)	IE	1219	‐	‐
4	Core biopsy	DLBCL	IIE	1198	2	‐
5	Incisional biopsy	DLBCL	IIE	874	2	‐
6	Core biopsy	DLBCL	IVE	1482	5	‐
7	Incisional biopsy	DLBCL	IVE	729	5	‐

Abbreviations: DLBCL, diffuse large B cell lymphoma; IPI, International Prognostic Index; LDH, lactate dehydrogenase.

For the five patients with DLBCL, their breast or axillary lump histopathology report shows diffuse proliferation of large lymphoid cells with vesicular nuclei. Tumour necrosis is present in all. All had high mitotic activity with Ki‐67 proliferation index above 60%. CD 20 and CD 79a cell surface markers, along with bcl‐2 and bcl‐6 oncogenes, were expressed in all these patients' biopsy specimens.

For patient number 3 (Table 2) who has marginal zone lymphoma (MALT type), her core biopsy histology report showed interstitial lymphoid infiltrate, featuring a polymorphous admixture of small lymphocytes and centrocyte‐like cells. There was expression of CD 20 and CD 79a cell surface markers, whereas CD 10 was negative.

As for patient number 1 (Table 2) with follicular lymphoma, her lump biopsy histology report showed diffuse effacement of architecture by closely packed neoplastic follicles of small lymphocytes. There were two to four centroblasts per high field power (hpf), thus corresponding to grade 1. CD 10, CD 20, and CD 23 cell surface markers and bcl 2 and bcl 6 oncogenes were expressed, and CD 5 was negative.

4.4. Staging

All patients underwent positron emission tomography‐computed tomography (PET‐CT) scan to stage their disease except for patient numbers 2 and 3 (Table 3). Patient number 2 did not undergo PET‐CT scan prior to her surgery as there was no suspicion of lymphoma. Her PET‐CT scan performed postsurgery showed no residual disease. Patient number 3 also did not undergo PET‐CT scan as she opted for expectant management. The PET‐CT scan findings for the other five patients are shown in Table 3. All the five patients who underwent PET‐CT scan had flurodeoxyglucose (FDG) avid breast masses with maximum standardized uptake values (SUVmax) ranging from 4.6 to 24.0. Disease staging for patients 2 and 3 was done using pan computed tomography scan.

Table 3.

PET‐CT scan findings

Patient Number	PET‐CT Scan Findings	Breast Mass SUVmax
1	• 1.1 × 0.3 cm hypermetabolic right breast mass	22.6
	• Multiple intensely hypermetabolic right subpectoral and axillary lymphadenopathy
	• Multiple hypermetabolic abdominopelvic and inguinal lymphadenopathy
2	‐	‐
3	‐	‐
4	• A 0.8 × 0.8 cm hypermetabolic left breast mass	4.6
	• Multiple enlarged left axillary, supraclavicular, subpectoral, and subcarinal nodes
5	• A 8.3 × 4.7 cm hypermetabolic left breast mass	11.6
	• Multiple enlarged left axillary nodes
6	• Multiple FDG avid necrotic enhancing masses up to 7 × 7 cm in the left breast	24.0
	• Bilateral FDG avid enlarged axillary and left subpectoral adenopathy
	• A 3.6 × 2.7 cm FDG avid enhancing posterior nasopharyngeal mass crossing the midline
	• Multiple enlarged FDG avid cervical, abdominopelvic, and retroperitoneal lymph nodes
7	• Large 8.0 × 10.3 cm hypermetabolic right breast mass with focus of calcification	13.4
	• Diffuse thickening and FDG uptake in the gastric mucosa
	• Several enlarged right axillary and right subpectoral nodes
	• Multiple hypermetabolic perigastric, mesenteric, and peritoneal nodes
	• Multiple hypermetabolic nodes in bilateral lungs and a hypermetabolic focus in the left lobe of the liver

Abbreviations: FDG, fludeoxyglucose; SUVmax, maximum standardized uptake value.

Based on the Ann Arbor classification, one was stage IE, three were stage IIE, one was stage IIIE, and two were stage IVE. Two of the stage IIE patients had ipsilateral axillary lymph node involvement, and one had supraclavicular, subcarinal, and subpectoral lymph node involvement as well. For the stage IVE patients, one had lymphomatous infiltration in the retropharyngeal space, while the other had gastric, lung, and liver involvement. Five patients had elevated LDH level at the time of diagnosis. Among them, two scored two points for IPI, indicating low to intermediate risk and two scored 5 points which indicate high risk.

4.5. Treatment and outcome

Six patients in this study underwent chemotherapy, as shown in Table 4. Five of them underwent 6 cycles of standard cyclophosphamide, doxorubicin, vincristine, prednisolone (CHOP) with rituximab regimen, and one had cyclophosphamide, vincristine, and prednisolone regimen. Patient number 3 (Table 4) opted for expectant management.

Table 4.

Treatment regimen and outcome

Patient Number	Chemotherapy Regimen	Response	Outcome	Length of Survival from Diagnosis (months) at Time of Analysis
1	R‐CHOP	Remission	Alive	65
2	R‐CHOP	Complete response	Alive	59
3	‐	‐	Alive	57
4	R‐CHOP	Remission	Alive	11
5	R‐CHOP	Remission	Alive	24
6	R‐CHOP	Progressive	Deceased	25
7	Vincristine/Cyclophosphamide/Prednisolone	Progressive	Deceased	2

Abbreviation: R‐CHOP, rituximab, cyclophosphamide, doxorubicin, vincristine, prednisolone.

One patient with aggressive lymphoma had complete response, three (42.9%) were in remission, and two (28.6%) had progressive disease. Patient number 3 who opted for expectant management (Table 4) experienced slow progressive enlargement of her axillary lump from 3 to 5 cm over her follow‐up period of 57 months. The two patients with progressive disease had passed away at the point of analysis. One passed away at 2 months, and the other 25 months after being diagnosed with the disease. Both patients passed away from sepsis secondary to pneumonia, one whom was still undergoing chemotherapy.

At the time of analysis, patients who had nondisseminated disease had a median survival of 57 months and mean survival time of 47.4 months. The overall survival at 3 years was calculated to be 64.3% (Figure 1).

Figure 1.

Overall survival for patients with primary breast lymphoma. The 3‐year overall survival rate is 64.3%. The mean survival time is 47.4 (95% CI 27.4‐67.4) months

5. DISCUSSION

Breast has been recognized to be an uncommon primary site for lymphoma. The paucity of cases in all centres worldwide has resulted in wide variations of information reported in current literature with regards to its presentation, staging, optimal treatment modality, and prognosis.

The definition of primary and nonprimary breast lymphoma still remains controversial. Some literature define PBL as lymphoma involving strictly the breast and ipsilateral axillary lymph nodes, whereas others define PBL more broadly as lymphoma involving the breast and lymph node, involvement beyond the ipsilateral axilla, ie, supraclavicular lymph nodes, or bone marrow involvement.16 According to the Wiseman and Liao classification, lymphoma which arises primarily from the breast that could metastasize in its natural history but was diagnosed early can still be considered PBL.7, 17 In our study, we included two patients with stage IVE Ann Arbor classification and classified them as primary breast lymphoma as the first clinical site of presentation for lymphoma was in the breast. Disseminated disease was only discovered after the primary diagnosis of breast lymphoma was made.

Vignot et al reported no pertinent differences between the clinical or histological presentation between primary and nonprimary breast lymphoma as well as their subsequent treatment indication and prognosis.17 Our study finding echoes similar findings when comparing patients with localized and disseminated disease. We found that both groups had similar patient and tumour characteristics such as age, size of tumour, and histology subtype. Both groups had received systemic chemotherapy as recommended by medical oncologists.

When comparing the clinical presentations and examination findings of our patients in this study with patients presenting with breast carcinoma in our centre, there are no distinguishing features between these two cancers. The fungating breast tumour that patient number 7 (Table 1) presented with appeared very similarly to fungating breast carcinomas, therefore rendering it impossible to distinguish between breast lymphoma and breast carcinoma via clinical examination.

The median age of diagnosis of our patients is 65 years, which is consistent with majority of other published studies where the median age ranged between 60 and 70 years.3, 6, 17, 18, 19, 20 In terms of clinical presentation, there is no predominance in the laterality of breast or axillary mass involvement although four out of seven of our patients presented with a left‐sided mass. Few studies have reported a right‐sided tumour predominance8, 10 although presently the significance is unknown. There have been reports of this disease manifesting in both breasts,8, 21, 22 although none in our study had bilateral breast involvement. The majority of lymphoma histology type in our study is diffuse large B cell, which is also a common finding in most literature.10, 20, 23, 24, 25

In the earlier decades, PBL was treated with mastectomy8; however, subsequent studies have shown that there is no improvement in the overall survival of patients who underwent mastectomy.8, 12, 21, 26, 27 Avenia et al recommended that mastectomy should be avoided unless for palliation purposes, and surgical intervention should be limited to diagnostic purposes.2, 11 None of the patients in our study underwent mastectomy. All had chemotherapy, and those with nondisseminated disease are currently in remission or had pathologic complete response. This further supports the notion of nonsurgical treatment in the control of this disease, which has been recommended in other smaller studies.6, 20

One of our patients who had presented with breast mass and subsequently underwent wide excision of the mass might have had a preoperative presumptive diagnosis of breast carcinoma. She subsequently underwent chemotherapy after surgical resection. In cases like this, Aviv et al strongly stated in their report that systemic therapy should still be administered to patients in whom surgery has been performed.2 Breast lymphoma is a systemic disease,28 and the mainstay of treatment is chemotherapy,8, 17 regardless whether nondisseminated (stages I and II) or disseminated (stages III and IV).21

Our study seems to suggest that IPI is a significant prognostic factor for overall survival. Both our patients with IPI score of 5, indicating high risk, had progressive disease and have passed away at the time of diagnosis. Other larger case series have also validated IPI as a prognostic factor.13, 18, 27, 29, 30 Poor prognostic factors for PBL in our study are stage IVE Ann Arbor classification, elevated LDH, and high IPI.

Three reports published in China,24 Singapore,9 and Spain1 reported 3‐year overall survival rate of 59.5%,24 77%,9 and 80%,1 respectively, compared to our study where our patients had a 3‐year overall survival rate of 64.3%. This is, however, keeping in mind that patients involved in the study in China and Spain were either stage IE or stage IIE.1, 24 Therefore, our patients have fairly good prognosis.

6. CONCLUSION

Primary breast lymphoma has been shown to behave similarly, both clinically and radiologically, as invasive carcinoma.5, 6 Therefore, it is prudent that histological confirmation be made10 because the mainstay of treatment for primary breast lymphoma is systemic chemotherapy.8, 17 Aggressive surgical resection in patients with primary breast lymphoma does not benefit them11 and can be associated with possible operative morbidity which may in turn delay their systemic treatment. In this retrospective review, our patients with nondisseminated breast lymphoma have demonstrated a fairly good survival outcome following solely chemotherapy.

7. CLINICAL LEARNING POINTS

1. PBL can present very similarly to breast carcinoma; however, their treatment modalities differ, which emphasizes the importance of obtaining histological diagnosis in patients presenting with breast mass.

2. Surgical intervention in PBL is mainly to obtain histological diagnosis, and aggressive surgical resection does not play a role in improving survival outcomes.

3. Systemic chemotherapy remains the mainstay of treatment for PBL.

CONFLICT OF INTEREST

The authors declare that they have no conflict of interest.

AUTHORS' CONTRIBUTION

All authors had full access to the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Conceptualization, M.Y.F., W.P.L., SM.T.; Methodology, M.Y.F., W.P.L.; Investigation, M.Y.F., W.P.L.; Formal Analysis, M.Y.F., C.K.; Resources, C.M.J.S.; Writing ‐ Original Draft, M.Y.F., W.P.L.; Writing ‐ Review & Editing, M.Y.F., W.P.L.; Visualization, M.Y.F., C.K.; Supervision, SM.T.; Funding Acquisition, NIL

Foo MY, Lee WP, Seah CMJ, Kam C, Tan S‐M. Primary breast lymphoma: A single‐centre experience. Cancer Reports. 2019;2:e1140. 10.1002/cnr2.1140