Cytopathology 1992,3,281-289
Inflammatory lesions of the breast: diagnosis by fine needle aspiration D. K. DAS, P. S O D H A N I , V. KASHYAP, S. PARKASH, J. N. P A N T A N D P. B H A T N A G A R Division of Cytopathology, Institute of Cytology and Preventive Oncology (ICMR), New Delhi, India Accepted for publication 24 April 1992
DAS D. K., SODHANI P., KASHYAP V., PARKASH S., PANT J. N. AND BHATNAGAR P.
(1992)
Cytopathology 3,28 1-289 Inflammatory lesions of the breast: diagnosis by fine needle aspiration
Amongst 1061 breast lesions diagnosed by fine needle aspiration (FNA) over a period of 6 years (1985-1990), 128 were reported to be showing changes consistent with an inflammatory lesion. On review, the cytodiagnosis was found to be inaccurate in 31 cases. The cytological features of the 97 cases that were correctly reported are described in this report. The cytological diagnoses issued in these 97 cases were acute mastitis or breast abscess (57 cases) and tuberculous mastitis (30 cases). Non-specific chronic mastitis and miscellaneous conditions accounted for four and six cases respectively. Acid fast bacilli (AFB) were demonstrated in 28.0% of tuberculous mastitis cases and 10.0% of those diagnosed as acute mastitis or breast abscess. FNA cytology was found to be useful for the diagnosis of inflammatory lesions of breast and their classification, as only five out of 57 cases of acute mastitis/breast abscess and one out of 30 tuberculous mastitis cases were suspected on clinical grounds. Keywords: mastitis, tuberculosis, breast abscess, fat necrosis
INTRODUCTION Inflammatory conditions of the breast may be confused with carcinoma and thus pose a clinical diagnostic problem'+. The importance of inflammatory lesions of the breast as a cause of morbidity is often overshadowed by the magnitude of the problem of breast cancer. This is reflected by the fact that even in large series of fine needle aspiration (FNA) of breast, the inflammatory lesions do not get a separate identity and are included under overall benign
condition^^.^. Between 1985 and 1990, 1061 patients with breast lesions were subjected to FNA or cytological examination of nipple discharge at the Institute of Cytology and Preventive Correspondence: Dr Dilip K Das, MD, PhD, Assistant Professor, Department of Pathology, Faculty of Medicine, Kuwait University, P.O. Box 24923, Safat 13110, Kuwait.
282 D.K. Das et al. Oncology (ICMR), New Delhi. In 128 cases (12.1%) the cytologic features were either suggestive or diagnostic of inflammatory lesions. After review, 3 1 cases were excluded since the cytological diagnosis was found to be inaccurate. In some cases, malignant cells were also present and in others the material was considered to be inadequate for reliable diagnosis. We describe the cytomorphologicial features in the remaining 97 cases. SUBJECTS A N D METHODS The age of the 97 cases included in this study ranged from 16 to 68 years with a median of 30 years. Two of the aspirates were from male patients. In 69 (71.1YO)cases the clinical data provided at the time of FNA was non-specific and no clinical diagnosis was mentioned in nine cases. In the remaining 19 cases a specific clinical diagnosis was offered which included abscess ( n= 7), malignancy (n = 5), fibroadenoma/fibroadenosis ( n = 3), fat necrosis (n= I), antibioma (n = l), cystic disease (n= 1) and tuberculosis ( n = 1). Fine needle aspiration was performed with a 22 or 23 G disposable needle and 20 ml plastic syringe fitted with a Franzen syringe handle. In 95 cases one breast was sampled and both the breasts were aspirated in two cases. Air-dried smears were routinely stained by May-Griinwald-Giemsa (MGG) stain and in a very limited number of cases wet-fixed smears stained by the Papanicolaou stain were also prepared. The Ziehl-Neelsen (Z-N) stain for acid fast bacilli (AFB) was performed initially in 45 cases. Subsequently during the review, MGG-stained smears in 12 cases showing acute inflammation were decolourized and stained by Z-N stain. Depending upon cytomorphological features, the cases were categorized as acute mastitis or breast abscess with and without evidence of organization, chronic non-specific mastitis, tuberculous mastitis and mastitis due to miscellaneous causes. The presence or absence of the following components were noted: ductular cell (DC), neutrophils (N), lymphocytes (L), plasma cells (PC), histiocytes (H), epithelioid cells (Ep), giant cells (GC), necrotic material (Nec), and capillaries (Cap). Based on a subjective impression, the various cellular components were considered to be present in abundant or moderate amounts or scanty. Cases with cytologic features consistent with or suggestive of tuberculous lesions were categorized under three headings’ which are as follows: type I reaction (epithelioid granuloma without necrosis), type I1 reaction (epithelioid granuloma with necrosis), and type I11 reaction (necrosis without epithelioid granuloma). The necrotic material appeared as a pinkish amorphous/granular material. In type TI1 reaction it was either completely acellular or accompanied by varying degrees of neutrophilic infiltration. There were, however, a few cases with acute inflammatory cell reaction and cell debris in which the presence of necrotic material was not definite. These cases were included under acute mastitis/breast abscess. The cytodiagnosis was correlated with AFB positivity. RESULTS Table 1 shows the cytologic features of the lesions. Acute mastitis or breast abscess was diagnosed in 57 (58.8%) cases. In 16 of these cases there was cytologic evidence of organization of the inflammatory response. There were four cases of non-specific chronic mastitis including plasma cell mastitis. In 30 cases (30.9%) the cytologic features were consistent with tuberculous mastitis. The type I, I1 and 111 reactions were observed in 3,
FNA dirtgnosis of injhnmatory lesions ofbreust
283
Table 1. Inflammatory lesions of the breast: frequency of distribution of various lesions and their cellular reaction Cellular reaction Cytodiagnosis Acute mastitis/breast abscess A. Non-organizing B. Organizing Chronic non-specific mastitis Tuberculous mastitis A. Type I reaction B. Type I I reaction C. Type 111 reaction Miscellaneous conditions Total
No. of cases
41 16
4
3 16 11 6
97
DC N
L
H
PC
5 1 3 - 3 - -
3 4 1 - 1 1 6 1 1 4 1 1 4 2
2
-
4 1 1 2 5 1 4 14
78
2
-
6
1
-
1
24
-
7
Ep
2 2 4
29
GC
Nec Cap
7* -
-
-
-
14
3
2
16 1
6
16
-
2
11 -
3 1 1
20
10
34
19
DC, Duct cells (benign); N, neutrophil; L, lymphocyte; PC, plasma cells; H, histiocytes; Ep, epithelioid cells; GC, giant cell; Nec, necrotic material; Cap, capillaries. Type I reaction, epithelioid granuloma without necrosis; Type I1 reaction, epithelioid granuloma with necrosis; Type 111 reaction, necrosis without epithelioid granuloma. *Presence of necrotic material was not definite.
16 and 11 cases respectively. Miscellaneous conditions such as fat necrosis and infected epidermal inclusion cyst were detected in six cases. The cytomorphological features observed in various inflammatory lesions were as follows.
Acute mustitis (breast abscess) The most striking feature in cases of acute mastitis or abscess without evidence of organization was the overwhelming presence of neutrophils which were in varying stages of degeneration (Figure 1). A subjective assessment of the neutrophils showed them to be in abundance in the majority of cases. In four cases neutrophils were scanty. These four cases were included in a group of seven cases where the presence of necrotic material was not definite but the possibility of tuberculosis could not be ruled out altogether. Histiocytes were rarely observed. Benign ductular cells were identified in three cases. In the cases showing evidence of organization, degenerating neutrophils were still the predominant cell type but mononuclear cells consisting of lymphocytes, histiocytes and plasma cells were also seen and the presence of capillaries surrounded by inflammatory cells was a consistent finding (Figure 2a,b).
Chronic non-specijic mastitis Lymphocytes, histiocytes and plasma cells were the predominant cell types. When abundant plasma cells were present, the cytodiagnosis of plasma cell mastitis was given (Figure 3).
284 D. K. Das et al.
Figure 1. FNA smear from breast abscess showing predominance of degenerated neutrophils. A few foam cells can also be seen (MGG x 160).
Figure 2. (a) Smear from organizing breast abscess shows branching capillaries surrounded by inflammatory cells. Numerous degenerated neutrophils can be seen in the background (MGG x 125). (b) Higher magnification of the capillaries surrounded by inflammatory cells which are predominantly lymphomononuclear by nature (MGG x 318).
Tuberculous mastitis
Type I reaction was characterized by presence of epithelioid granuloma. There was no evidence of caseation. Of the three cases studied, benign ductular cells, lymphocytes and multinucleated giant cells were observed in two cases. Type I1 reaction which formed the largest group (16 cases) was characterized by groups of epithelioid cells and necroctic material (Figure 4a). Multinucleated giant cells and lymphocytes were observed in six cases each (Figure 4b). Varying numbers of neutrophils were present in 11 cases (Figures 5a,b). Benign ductular cells were identified in four cases (Figure 6).
FNA diagnosis of injammatory lesions of breast
285
Figure 3. FNA smear from plasma cell mastitis showing large number of plasma cells in addition to lymphocytes and foam cells (MGG x 500).
Figure 4. (a) Tuberculousmastitis. FNA smear shows a group of epithelioid cells and necrotic material (MGG x 200). (b) A huge multinucleated giant cell along with epithelioid cells, lymphocytes and necrotic material (MGG x 160).
In all of the 11 cases with type 111 reaction there was necrotic material but no epithelioid cells. Varying numbers of neutrophils were present in five cases and six cases showed acellular necrotic material. Miscellaneous lesions There were five cases of fat necrosis and one case of infected epidermal inclusion cyst in this group. In the cases with fat necrosis, fat cells were seen in varying stages of degeneration. Neutrophils and foamy histiocytes were observed in four cases each. Multinucleated giant cells and epithelioid cells were noticed in two and one case respectively. The lone case
286
D.K . Daset al.
Figure 5 (a) FNA smear from tuberculous mastitis showing a group of epithelioid cells surrounded by scanty inflammatory cells, both lymphocytes and neutrophils (MGG x 200). (b) A small group of epithelioid cells surrounded by intense neutrophilic reaction (MGG x 200).
Figure 6 Tuberculous mastitis. A group of benign ductular cells, epithelioid cells and scanty necrotic material (MGG x 318).
of infected epidermal inclusion cyst contained numerous anucleated squames and many neutrophils. The lymph node aspirates in a case of acute mastitis (breast abscess) and a case of fat necrosis showed features of reactive hyperplasia. FNA of lymph node was also performed in six cases of tuberculous mastitis. Of these, five cases showed features of tuberculous lymphadenitis and one showed reactive hyperplasia. Correlation of cytologic diagnosis with AFB positivity (Table 2 )
Z-N staining was performed on 30 cases of acute mastitis. AFB was demonstrated in three cases (10%). Of these three cases, one was a previously reported negative case which turned out to be positive during review and two were decolourized MGG stained smears restained
FNA diagnosis of inflammatory lesions of breast
287
Table 2 Acid-fast bacilli (AFB) positivity in inflammatory lesions of the breast
Cytological diagnosis Acute mastitis/breast abscess A. Non-organizing B. Organizing Chronic non-specific mastitis Tuberculous mastitis A. Type I reaction B. Type I1 reaction C. Type 111 reaction Miscellaneous conditions Total
No. of cases
No. tested for AFB (Z-N stain)
41 16 3
20 10 0
2* (10.0) l f (10.0) 0 (0.0)
3 16 11 6
2 12 11 2
0 (0.0) 3 (25.0) 4 (36.4) 0 (0.0)
97
57
10 (17.5)
No. positive for AFB (YO)
*One positive case was obtained during rescreening of Z-N stained smear and the other from a decolourized MGG stained smear which was restained by Z-N stain. +Obtained from a decolourized M G G stained smear which was restained by Z-N stain.
by Z-N stain. Z-N staining was performed in 25 cases altogether with cytologic features of tuberculous mastitis. AFB were demonstrated in 7 (28%) cases. Whereas no AFB were demonstrated in cases with type I reaction, cases with type I1 and I11 reaction showed the presence of AFB in three (25%) and four (36.4%) cases respectively. DISCUSSION Infective lesions of the breast constitute 4 . 1 4 0 % of benign lesions in females in western literature' lo. In two Indian studies4,", the frequency has been found to be 11 .O% and 22.6% respectively. The above studies are based on histopathological diagnosis. In this study, which is a cytological one, 97 (9.1"/o) of the 1061cases showed unequivocal evidence of inflammatory disease. Our findings are very similar to those of Franzen and Zajicek'*, who found 41 (7.4%) cases ofmastitis and six (1.1YO)of fat necrosis in 556 cytologically diagnosed and subsequently biopsy proven cases. Acute mastitis or breast abscess which is caused by bacteria such as the staphylococcus*is the most common variety of mastitis'. There were very few cases of chronic non-specific mastitis in this study compared with that reported by Khanna et ~ 1who . ~found it to be the most common variety. Franzen and Zajicek" also found a very large number of chronic mastitis cases (22 out of 41). Tuberculous mastitis has been found to constitute 0.66-1.6% of all mammary lesions in reports from the west between 1940 and 1960'3-'5.In recent years, it has become extremely rare. In one laboratory currently receiving over 1500 breast aspirates per year, only five cases of tuberculosis were seen in the last 15years (personal communications). In the present study, tuberculous mastitis constituted 2.8% of all mammary lesions and 30.9% of inflammatory lesions of the breast. Similar high frequencies have been observed in other Indian studies. Nayar and SaxenaI6 found tuberculous mastitis in 3.4% of all mammary lesions. Khanna found that 25.1 YOof the infective lesions of breast were due to tuberculosis. et
288 D . K . Das et al. According to Macansh et aL3 granulomatous mastitis is characterized histologically by multiple non-caseating epithelioid granulomas with microabscess formation and giant cells involving extensive areas of breast tissue. The differential diagnosis of such lesions includes duct ectasia, fat necrosis, sarcoidosis, foreign body granuloma, and infective conditions like tuberculosis and typhoid. The diagnosis of tuberculous lesions has been based on demonstration of epithelioid granuloma with or without necrosis and necrosis without epithelioid granuloma'. When epithelioid granulomas are observed, tuberculosis should be considered the first possibility in Indian patients until proved otherwise, since it is a very common disease in this subcontinent. The other condition where we observed epithelioid granuloma or giant cells was fat necrosis (n = 2) but the presence of degenerated fat cells distinguished this lesion from tuberculosis. Unlike the cases with epithelioid granuloma, those with necrosis but no epithelioid cells pose a diagnostic problem, especially when associated with acute inflammatory cell infiltration. This may also be appreciated from the finding that 10% of cases categorized as acute mastitis or breast abscess on morphological grounds were positive for AFB. According to Das et al.7and Rajwanshi et al." AFB positivity is highest in lesions showing necrosis but no epithelioid granuloma and this was also observed in the present study. Breast aspirates are much better for demonstration of AFB compared with histologic sections. Nayar and Saxena'' failed to demonstrate AFB in sections from 12 patients in whom the histology revealed caseating epithelioid granulomas. It is suggested that the aspirates may be sent directly to microbiology for culture when Z-N stained samples are negative for AFB. The enormous contribution of FNA to the diagnosis of inflammatory lesions of the breast can be realized when one analyses the pre-aspiration clinical diagnoses of these cases. Only five of the 57 cases diagnosed on cytology as acute mastitis or breast abscess had been correctly diagnosed clinically. Similarly, of the 30 tuberculous mastitis cases, the possibility of tuberculosis was considered clinically in only one case. Thus, fine needle aspiration, besides offering a quick diagnosis in inflammatory lesions of breast and solving the problem of benign vs malignant, can be of use in subtyping the lesions into non-specific and specific inflammations and part of the sample can be used for bacteriological studies. ACKNOWLEDGEMENTS The authors wish to thank Mrs Alice Mathew and Mr James Luke for their secretarial assistance in the preparation of the manuscript. REFERENCES 1 Harging Rains AJ, Ritchie DH. The breast In:
Bailey and Love's Short Practice ofsurgery, 19th edn. London: ELBS/H.K. Lewis, 1984; 656-82. 2 Cotran RS, Kumar V, Robbins SL. Robbins Pathologic Basis of Disease, 4th edn. Philadephia: WB Saunders Co. 1989; 1184-5. 3 Macansh S , Greenberg M, Barraclough B, Pacey F. Fine needle aspiration cytology of granulomatous mastitis. Acta Cytol1990; 3 4 3842.
4 Khanna S , Arya NC, Khanna NN. Spectrum of benign breast disease. Indian J Surg 1988; 5 0 169-75. 5 Smallwood J, Herbert A, Guyer P., Taylor I. Accuracy of aspiration cytology in the diagnosis of breast disease. Br J Surg 1985; 72: 841-3. 6 Hammond S , Keyhani-Rofagha S, O'Toole RV. Statistical analysis of fine needle aspiration cytology of the breast. A review of 678 cases plus 4265 cases from the literature. Acta Cytoll987;31: 276-80.
FNA diagnosis of inflammatory lesions of breast 7 Das DK, Pant JN, Chachra KL, Murthy NS, Satynaryana L, Thankamma TC, Kakkar PK. Tuberculous lymphadenitis: correlation ofcellular components and necrosis in lymph node aspirate with AFB positivity and bacillary count. Indian J Pathol Mirrohiol1990; 3 3 1-10. 8 Sartwell PE, Arthes FG, Tonascia JA. Epidemiology of benign breast lesions; lack of association with oral contraceptive use. New Engl J Med 1973; 288: 5 5 1 4 . 9 Oluwole SF, Freeman HP. Analysis of benign breast lesions in blacks. Am J Surg 1979; 137: 786-9. 10 Fundorburk WW, Rosero E, Lefall LD. Breast lesions in blacks. Surg Gynecol Obstet 1972; 135: 58-60. I 1 Rangabashyam N, Gyanprakashan D, Krishnaraj B, Manohar V, Vijaylakshmi SR. Spectrum of benign breast lesions in Madras. J R Coil Surg Edinb 1983; 28: 369-73.
289
12 Franzen S, Zajicek J. Aspiration biopsy . . in diagnosis of palpable lesions of the breast. Acta Radio1 (Stockh) 1968; 7: 241-62. 3 Sandison AT, Walker JC. Inflammatory mastitis, mammary duct ectasia and mammilary fistula: inflammatory lesions of the breast. Br J Surg 1962; 5 0 57-64. 4 Dawson EK, Harvey WF. Macro and micro diagnosis of cancer: a laboratory survey of routine mammary lesions. Edin Med J 1942; 4 9 401-8. 15 Haagensen CD. Diseases of the Breast. Philadelphia: WB Saunders, 1956: 3 0 4 1 1. 16 Nayar M, Saxena HMK. Tuberculosis of the breast: a cytomorphologicalstudy of needle aspirates and nipple discharges. Acta Cytol 1984; 2 8 325-8. 17 Rajwanshi A, Bhambhani S, Das DK. Fine needle aspiration cytology in diagnosis of tuberculosis. Diagn Cytopatholl987; 3: 13-6.
Inflammatory lesions of the breast: diagnosis by fine needle aspiration D. K. DAS, P. S O D H A N I , V. KASHYAP, S. PARKASH, J. N. P A N T A N D P. B H A T N A G A R Division of Cytopathology, Institute of Cytology and Preventive Oncology (ICMR), New Delhi, India Accepted for publication 24 April 1992
DAS D. K., SODHANI P., KASHYAP V., PARKASH S., PANT J. N. AND BHATNAGAR P.
(1992)
Cytopathology 3,28 1-289 Inflammatory lesions of the breast: diagnosis by fine needle aspiration
Amongst 1061 breast lesions diagnosed by fine needle aspiration (FNA) over a period of 6 years (1985-1990), 128 were reported to be showing changes consistent with an inflammatory lesion. On review, the cytodiagnosis was found to be inaccurate in 31 cases. The cytological features of the 97 cases that were correctly reported are described in this report. The cytological diagnoses issued in these 97 cases were acute mastitis or breast abscess (57 cases) and tuberculous mastitis (30 cases). Non-specific chronic mastitis and miscellaneous conditions accounted for four and six cases respectively. Acid fast bacilli (AFB) were demonstrated in 28.0% of tuberculous mastitis cases and 10.0% of those diagnosed as acute mastitis or breast abscess. FNA cytology was found to be useful for the diagnosis of inflammatory lesions of breast and their classification, as only five out of 57 cases of acute mastitis/breast abscess and one out of 30 tuberculous mastitis cases were suspected on clinical grounds. Keywords: mastitis, tuberculosis, breast abscess, fat necrosis
INTRODUCTION Inflammatory conditions of the breast may be confused with carcinoma and thus pose a clinical diagnostic problem'+. The importance of inflammatory lesions of the breast as a cause of morbidity is often overshadowed by the magnitude of the problem of breast cancer. This is reflected by the fact that even in large series of fine needle aspiration (FNA) of breast, the inflammatory lesions do not get a separate identity and are included under overall benign
condition^^.^. Between 1985 and 1990, 1061 patients with breast lesions were subjected to FNA or cytological examination of nipple discharge at the Institute of Cytology and Preventive Correspondence: Dr Dilip K Das, MD, PhD, Assistant Professor, Department of Pathology, Faculty of Medicine, Kuwait University, P.O. Box 24923, Safat 13110, Kuwait.
282 D.K. Das et al. Oncology (ICMR), New Delhi. In 128 cases (12.1%) the cytologic features were either suggestive or diagnostic of inflammatory lesions. After review, 3 1 cases were excluded since the cytological diagnosis was found to be inaccurate. In some cases, malignant cells were also present and in others the material was considered to be inadequate for reliable diagnosis. We describe the cytomorphologicial features in the remaining 97 cases. SUBJECTS A N D METHODS The age of the 97 cases included in this study ranged from 16 to 68 years with a median of 30 years. Two of the aspirates were from male patients. In 69 (71.1YO)cases the clinical data provided at the time of FNA was non-specific and no clinical diagnosis was mentioned in nine cases. In the remaining 19 cases a specific clinical diagnosis was offered which included abscess ( n= 7), malignancy (n = 5), fibroadenoma/fibroadenosis ( n = 3), fat necrosis (n= I), antibioma (n = l), cystic disease (n= 1) and tuberculosis ( n = 1). Fine needle aspiration was performed with a 22 or 23 G disposable needle and 20 ml plastic syringe fitted with a Franzen syringe handle. In 95 cases one breast was sampled and both the breasts were aspirated in two cases. Air-dried smears were routinely stained by May-Griinwald-Giemsa (MGG) stain and in a very limited number of cases wet-fixed smears stained by the Papanicolaou stain were also prepared. The Ziehl-Neelsen (Z-N) stain for acid fast bacilli (AFB) was performed initially in 45 cases. Subsequently during the review, MGG-stained smears in 12 cases showing acute inflammation were decolourized and stained by Z-N stain. Depending upon cytomorphological features, the cases were categorized as acute mastitis or breast abscess with and without evidence of organization, chronic non-specific mastitis, tuberculous mastitis and mastitis due to miscellaneous causes. The presence or absence of the following components were noted: ductular cell (DC), neutrophils (N), lymphocytes (L), plasma cells (PC), histiocytes (H), epithelioid cells (Ep), giant cells (GC), necrotic material (Nec), and capillaries (Cap). Based on a subjective impression, the various cellular components were considered to be present in abundant or moderate amounts or scanty. Cases with cytologic features consistent with or suggestive of tuberculous lesions were categorized under three headings’ which are as follows: type I reaction (epithelioid granuloma without necrosis), type I1 reaction (epithelioid granuloma with necrosis), and type I11 reaction (necrosis without epithelioid granuloma). The necrotic material appeared as a pinkish amorphous/granular material. In type TI1 reaction it was either completely acellular or accompanied by varying degrees of neutrophilic infiltration. There were, however, a few cases with acute inflammatory cell reaction and cell debris in which the presence of necrotic material was not definite. These cases were included under acute mastitis/breast abscess. The cytodiagnosis was correlated with AFB positivity. RESULTS Table 1 shows the cytologic features of the lesions. Acute mastitis or breast abscess was diagnosed in 57 (58.8%) cases. In 16 of these cases there was cytologic evidence of organization of the inflammatory response. There were four cases of non-specific chronic mastitis including plasma cell mastitis. In 30 cases (30.9%) the cytologic features were consistent with tuberculous mastitis. The type I, I1 and 111 reactions were observed in 3,
FNA dirtgnosis of injhnmatory lesions ofbreust
283
Table 1. Inflammatory lesions of the breast: frequency of distribution of various lesions and their cellular reaction Cellular reaction Cytodiagnosis Acute mastitis/breast abscess A. Non-organizing B. Organizing Chronic non-specific mastitis Tuberculous mastitis A. Type I reaction B. Type I I reaction C. Type 111 reaction Miscellaneous conditions Total
No. of cases
41 16
4
3 16 11 6
97
DC N
L
H
PC
5 1 3 - 3 - -
3 4 1 - 1 1 6 1 1 4 1 1 4 2
2
-
4 1 1 2 5 1 4 14
78
2
-
6
1
-
1
24
-
7
Ep
2 2 4
29
GC
Nec Cap
7* -
-
-
-
14
3
2
16 1
6
16
-
2
11 -
3 1 1
20
10
34
19
DC, Duct cells (benign); N, neutrophil; L, lymphocyte; PC, plasma cells; H, histiocytes; Ep, epithelioid cells; GC, giant cell; Nec, necrotic material; Cap, capillaries. Type I reaction, epithelioid granuloma without necrosis; Type I1 reaction, epithelioid granuloma with necrosis; Type 111 reaction, necrosis without epithelioid granuloma. *Presence of necrotic material was not definite.
16 and 11 cases respectively. Miscellaneous conditions such as fat necrosis and infected epidermal inclusion cyst were detected in six cases. The cytomorphological features observed in various inflammatory lesions were as follows.
Acute mustitis (breast abscess) The most striking feature in cases of acute mastitis or abscess without evidence of organization was the overwhelming presence of neutrophils which were in varying stages of degeneration (Figure 1). A subjective assessment of the neutrophils showed them to be in abundance in the majority of cases. In four cases neutrophils were scanty. These four cases were included in a group of seven cases where the presence of necrotic material was not definite but the possibility of tuberculosis could not be ruled out altogether. Histiocytes were rarely observed. Benign ductular cells were identified in three cases. In the cases showing evidence of organization, degenerating neutrophils were still the predominant cell type but mononuclear cells consisting of lymphocytes, histiocytes and plasma cells were also seen and the presence of capillaries surrounded by inflammatory cells was a consistent finding (Figure 2a,b).
Chronic non-specijic mastitis Lymphocytes, histiocytes and plasma cells were the predominant cell types. When abundant plasma cells were present, the cytodiagnosis of plasma cell mastitis was given (Figure 3).
284 D. K. Das et al.
Figure 1. FNA smear from breast abscess showing predominance of degenerated neutrophils. A few foam cells can also be seen (MGG x 160).
Figure 2. (a) Smear from organizing breast abscess shows branching capillaries surrounded by inflammatory cells. Numerous degenerated neutrophils can be seen in the background (MGG x 125). (b) Higher magnification of the capillaries surrounded by inflammatory cells which are predominantly lymphomononuclear by nature (MGG x 318).
Tuberculous mastitis
Type I reaction was characterized by presence of epithelioid granuloma. There was no evidence of caseation. Of the three cases studied, benign ductular cells, lymphocytes and multinucleated giant cells were observed in two cases. Type I1 reaction which formed the largest group (16 cases) was characterized by groups of epithelioid cells and necroctic material (Figure 4a). Multinucleated giant cells and lymphocytes were observed in six cases each (Figure 4b). Varying numbers of neutrophils were present in 11 cases (Figures 5a,b). Benign ductular cells were identified in four cases (Figure 6).
FNA diagnosis of injammatory lesions of breast
285
Figure 3. FNA smear from plasma cell mastitis showing large number of plasma cells in addition to lymphocytes and foam cells (MGG x 500).
Figure 4. (a) Tuberculousmastitis. FNA smear shows a group of epithelioid cells and necrotic material (MGG x 200). (b) A huge multinucleated giant cell along with epithelioid cells, lymphocytes and necrotic material (MGG x 160).
In all of the 11 cases with type 111 reaction there was necrotic material but no epithelioid cells. Varying numbers of neutrophils were present in five cases and six cases showed acellular necrotic material. Miscellaneous lesions There were five cases of fat necrosis and one case of infected epidermal inclusion cyst in this group. In the cases with fat necrosis, fat cells were seen in varying stages of degeneration. Neutrophils and foamy histiocytes were observed in four cases each. Multinucleated giant cells and epithelioid cells were noticed in two and one case respectively. The lone case
286
D.K . Daset al.
Figure 5 (a) FNA smear from tuberculous mastitis showing a group of epithelioid cells surrounded by scanty inflammatory cells, both lymphocytes and neutrophils (MGG x 200). (b) A small group of epithelioid cells surrounded by intense neutrophilic reaction (MGG x 200).
Figure 6 Tuberculous mastitis. A group of benign ductular cells, epithelioid cells and scanty necrotic material (MGG x 318).
of infected epidermal inclusion cyst contained numerous anucleated squames and many neutrophils. The lymph node aspirates in a case of acute mastitis (breast abscess) and a case of fat necrosis showed features of reactive hyperplasia. FNA of lymph node was also performed in six cases of tuberculous mastitis. Of these, five cases showed features of tuberculous lymphadenitis and one showed reactive hyperplasia. Correlation of cytologic diagnosis with AFB positivity (Table 2 )
Z-N staining was performed on 30 cases of acute mastitis. AFB was demonstrated in three cases (10%). Of these three cases, one was a previously reported negative case which turned out to be positive during review and two were decolourized MGG stained smears restained
FNA diagnosis of inflammatory lesions of breast
287
Table 2 Acid-fast bacilli (AFB) positivity in inflammatory lesions of the breast
Cytological diagnosis Acute mastitis/breast abscess A. Non-organizing B. Organizing Chronic non-specific mastitis Tuberculous mastitis A. Type I reaction B. Type I1 reaction C. Type 111 reaction Miscellaneous conditions Total
No. of cases
No. tested for AFB (Z-N stain)
41 16 3
20 10 0
2* (10.0) l f (10.0) 0 (0.0)
3 16 11 6
2 12 11 2
0 (0.0) 3 (25.0) 4 (36.4) 0 (0.0)
97
57
10 (17.5)
No. positive for AFB (YO)
*One positive case was obtained during rescreening of Z-N stained smear and the other from a decolourized MGG stained smear which was restained by Z-N stain. +Obtained from a decolourized M G G stained smear which was restained by Z-N stain.
by Z-N stain. Z-N staining was performed in 25 cases altogether with cytologic features of tuberculous mastitis. AFB were demonstrated in 7 (28%) cases. Whereas no AFB were demonstrated in cases with type I reaction, cases with type I1 and I11 reaction showed the presence of AFB in three (25%) and four (36.4%) cases respectively. DISCUSSION Infective lesions of the breast constitute 4 . 1 4 0 % of benign lesions in females in western literature' lo. In two Indian studies4,", the frequency has been found to be 11 .O% and 22.6% respectively. The above studies are based on histopathological diagnosis. In this study, which is a cytological one, 97 (9.1"/o) of the 1061cases showed unequivocal evidence of inflammatory disease. Our findings are very similar to those of Franzen and Zajicek'*, who found 41 (7.4%) cases ofmastitis and six (1.1YO)of fat necrosis in 556 cytologically diagnosed and subsequently biopsy proven cases. Acute mastitis or breast abscess which is caused by bacteria such as the staphylococcus*is the most common variety of mastitis'. There were very few cases of chronic non-specific mastitis in this study compared with that reported by Khanna et ~ 1who . ~found it to be the most common variety. Franzen and Zajicek" also found a very large number of chronic mastitis cases (22 out of 41). Tuberculous mastitis has been found to constitute 0.66-1.6% of all mammary lesions in reports from the west between 1940 and 1960'3-'5.In recent years, it has become extremely rare. In one laboratory currently receiving over 1500 breast aspirates per year, only five cases of tuberculosis were seen in the last 15years (personal communications). In the present study, tuberculous mastitis constituted 2.8% of all mammary lesions and 30.9% of inflammatory lesions of the breast. Similar high frequencies have been observed in other Indian studies. Nayar and SaxenaI6 found tuberculous mastitis in 3.4% of all mammary lesions. Khanna found that 25.1 YOof the infective lesions of breast were due to tuberculosis. et
288 D . K . Das et al. According to Macansh et aL3 granulomatous mastitis is characterized histologically by multiple non-caseating epithelioid granulomas with microabscess formation and giant cells involving extensive areas of breast tissue. The differential diagnosis of such lesions includes duct ectasia, fat necrosis, sarcoidosis, foreign body granuloma, and infective conditions like tuberculosis and typhoid. The diagnosis of tuberculous lesions has been based on demonstration of epithelioid granuloma with or without necrosis and necrosis without epithelioid granuloma'. When epithelioid granulomas are observed, tuberculosis should be considered the first possibility in Indian patients until proved otherwise, since it is a very common disease in this subcontinent. The other condition where we observed epithelioid granuloma or giant cells was fat necrosis (n = 2) but the presence of degenerated fat cells distinguished this lesion from tuberculosis. Unlike the cases with epithelioid granuloma, those with necrosis but no epithelioid cells pose a diagnostic problem, especially when associated with acute inflammatory cell infiltration. This may also be appreciated from the finding that 10% of cases categorized as acute mastitis or breast abscess on morphological grounds were positive for AFB. According to Das et al.7and Rajwanshi et al." AFB positivity is highest in lesions showing necrosis but no epithelioid granuloma and this was also observed in the present study. Breast aspirates are much better for demonstration of AFB compared with histologic sections. Nayar and Saxena'' failed to demonstrate AFB in sections from 12 patients in whom the histology revealed caseating epithelioid granulomas. It is suggested that the aspirates may be sent directly to microbiology for culture when Z-N stained samples are negative for AFB. The enormous contribution of FNA to the diagnosis of inflammatory lesions of the breast can be realized when one analyses the pre-aspiration clinical diagnoses of these cases. Only five of the 57 cases diagnosed on cytology as acute mastitis or breast abscess had been correctly diagnosed clinically. Similarly, of the 30 tuberculous mastitis cases, the possibility of tuberculosis was considered clinically in only one case. Thus, fine needle aspiration, besides offering a quick diagnosis in inflammatory lesions of breast and solving the problem of benign vs malignant, can be of use in subtyping the lesions into non-specific and specific inflammations and part of the sample can be used for bacteriological studies. ACKNOWLEDGEMENTS The authors wish to thank Mrs Alice Mathew and Mr James Luke for their secretarial assistance in the preparation of the manuscript. REFERENCES 1 Harging Rains AJ, Ritchie DH. The breast In:
Bailey and Love's Short Practice ofsurgery, 19th edn. London: ELBS/H.K. Lewis, 1984; 656-82. 2 Cotran RS, Kumar V, Robbins SL. Robbins Pathologic Basis of Disease, 4th edn. Philadephia: WB Saunders Co. 1989; 1184-5. 3 Macansh S , Greenberg M, Barraclough B, Pacey F. Fine needle aspiration cytology of granulomatous mastitis. Acta Cytol1990; 3 4 3842.
4 Khanna S , Arya NC, Khanna NN. Spectrum of benign breast disease. Indian J Surg 1988; 5 0 169-75. 5 Smallwood J, Herbert A, Guyer P., Taylor I. Accuracy of aspiration cytology in the diagnosis of breast disease. Br J Surg 1985; 72: 841-3. 6 Hammond S , Keyhani-Rofagha S, O'Toole RV. Statistical analysis of fine needle aspiration cytology of the breast. A review of 678 cases plus 4265 cases from the literature. Acta Cytoll987;31: 276-80.
FNA diagnosis of inflammatory lesions of breast 7 Das DK, Pant JN, Chachra KL, Murthy NS, Satynaryana L, Thankamma TC, Kakkar PK. Tuberculous lymphadenitis: correlation ofcellular components and necrosis in lymph node aspirate with AFB positivity and bacillary count. Indian J Pathol Mirrohiol1990; 3 3 1-10. 8 Sartwell PE, Arthes FG, Tonascia JA. Epidemiology of benign breast lesions; lack of association with oral contraceptive use. New Engl J Med 1973; 288: 5 5 1 4 . 9 Oluwole SF, Freeman HP. Analysis of benign breast lesions in blacks. Am J Surg 1979; 137: 786-9. 10 Fundorburk WW, Rosero E, Lefall LD. Breast lesions in blacks. Surg Gynecol Obstet 1972; 135: 58-60. I 1 Rangabashyam N, Gyanprakashan D, Krishnaraj B, Manohar V, Vijaylakshmi SR. Spectrum of benign breast lesions in Madras. J R Coil Surg Edinb 1983; 28: 369-73.
289
12 Franzen S, Zajicek J. Aspiration biopsy . . in diagnosis of palpable lesions of the breast. Acta Radio1 (Stockh) 1968; 7: 241-62. 3 Sandison AT, Walker JC. Inflammatory mastitis, mammary duct ectasia and mammilary fistula: inflammatory lesions of the breast. Br J Surg 1962; 5 0 57-64. 4 Dawson EK, Harvey WF. Macro and micro diagnosis of cancer: a laboratory survey of routine mammary lesions. Edin Med J 1942; 4 9 401-8. 15 Haagensen CD. Diseases of the Breast. Philadelphia: WB Saunders, 1956: 3 0 4 1 1. 16 Nayar M, Saxena HMK. Tuberculosis of the breast: a cytomorphologicalstudy of needle aspirates and nipple discharges. Acta Cytol 1984; 2 8 325-8. 17 Rajwanshi A, Bhambhani S, Das DK. Fine needle aspiration cytology in diagnosis of tuberculosis. Diagn Cytopatholl987; 3: 13-6.
