- Email: [email protected]
Lipoma of the breast: a diagnostic dilemma
ARTICLE IN PRESS The Breast (2004) 13, 408–411
THE
BREAST www.elsevier.com/locate/breast
ORIGINAL ARTICLE
Lipoma of the breast: a diagnostic dilemma C. Lannga,*, B.Ø. Eriksenb, J. Hoffmanna a
Breast Surgery Clinic, Horsholm Hospital, Usserod Kongevej 102, Horsholm 2970, Denmark Department of Radiology, Horsholm Hospital, Horsholm, Denmark
b
Received 21 January 2004; received in revised form 20 April 2004; accepted 27 April 2004
KEYWORDS Breast; Lipoma; Triple diagnosis; Mammography; Ultrasound; Fine-needle aspiration cytology
Summary Lipoma of the breast often causes diagnostic and therapeutic uncertainty. Clinically it may be difficult to distinguish a lipoma from other conditions. Fineneedle aspiration cytology (FNAC) is often not helpful. Both mammography and ultrasound scanning are often negative. For the present study, 108 women with a clinical diagnosis of lipoma were enrolled prospectively. The clinical diagnosis of lipoma was found to be incorrect in 25.0% of these cases. Mammography and ultrasound revealed a lipoma in only 3.0% and 21.0%, respectively. FNAC revealed only fat cells in 74.0% of cases. In all, only 9 patients (11.4%) fulfilled the triple diagnostic criteria, theoretically making tumour excision mandatory in the remaining cases. Our proposal for management is for any clinical diagnosis of lipoma to be confirmed by either FNAC revealing fat cells or a core biopsy consistent with a lipoma. The mammogram and the ultrasound need not necessarily demonstrate a lipoma, but obviously must not show anything to raise the suspicion of malignancy at the site. If these criteria are met it is not necessary to excise the tumour. & 2004 Elsevier Ltd. All rights reserved.
Introduction Lipoma of the breast is a benign condition that provokes little interest in the literature and scant mention in standard texts.1,2 Although it is a banal condition it often causes diagnostic uncertainty, which results in confusion concerning treatment strategies. This uncertainty can arise at any point in the diagnostic process: clinically it may be difficult to distinguish a lipoma from a prominent *Corresponding author. Tel.: þ 45-4829-6965; fax: þ 45-48292667. E-mail address: [email protected] (C. Lanng).
fat lobule or from other benign or malignant processes. Fine-needle aspiration cytology (FNAC) reveals either fat cells only, which by definition renders the test ‘inadequate’,3 or fat cells together with normal epithelial cells, which casts doubt on the diagnosis of lipoma. Both mammography and ultrasound scanning are often negative.4,5 This results in a situation where it is difficult to fulfil the requirements of triple diagnosis, which may culminate in unnecessary tumour excision with all its psychological and economic consequences. A prospective investigation was planned with the object of documenting this problem and designing a suitable treatment strategy.
0960-9776/$ - see front matter & 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.breast.2004.04.011
ARTICLE IN PRESS Lipoma of the breast
Patients and methods The Breast Surgery Clinic, Horsholm Hospital, is a regional breast care centre serving a population of 380,000. The clinic receives over 3000 new referrals per year, about 12% of which are for breast cancer. The clinic is served by five full-time breast surgeons, two full-time diagnostic radiologists and three full-time pathologists. Between 1 January and 31 December 2002 a prospective registration was carried out: All women referred with a breast lump were first assessed by one of the five surgeons. If a clinical diagnosis of lipoma was made the patient was enrolled. The clinical criterion was a well-defined, smooth, possibly slightly lobulated, soft or semi-firm, mobile tumour detected on palpation. The mammographic and ultrasound findings were recorded. The mammographic criteria were: well defined, encapsulated radiotranslucency, often with compression of adjacent breast tissue. The ultrasound criteria were: oval, lobulated, homogeneous, solid mass with an echogenicity similar to that of normal fat. The results of the FNAC and/or core biopsy were recorded. After being given clear information about the presumed diagnosis, the patient was asked to choose between observation or excision. If excision was chosen the histological result was recorded. If observation was preferred a clinical reassessment was carried out 6 months later.
Results A clinical diagnosis of lipoma was recorded in 108 women. Their ages ranged from 26 to 88 years (median 51 years). The tumour size ranged from 5 to 35 mm (median 15 mm). Mammography revealed a lipoma in only 3 of 102 cases (2.8%). (Three women were not examined mammographically because of their young age and 3, because they had a mammography showing a normal result within the previous 12 months.) In 1 case the mammogram showed that the palpable tumour was a cancer. Further details are shown in Table 1. Ultrasound examination showed a lipoma in 23 out of 107 cases (21.3%). In 2 cases ultrasound revealed that the tumour was a cancer. Further details are shown in Table 1. FNAC was done in 106 cases and showed fat cells only in 80 cases (74.0%), fat cells and epithelial cells in 11 and malignant cells in 1; it
409
was judged inadequate in 3. Further details are shown in Table 2. Core needle biopsy was performed in addition to FNAC in 7 cases and showed a lipoma in 3, fibroadenosis in 1, fat tissue in 4, and cancer in 1. Another 2 patients underwent core biopsy instead of FNAC, revealing a cancer in 1 and a lipoma in the other. Twenty-four patients chose to undergo excision of their tumour. The 2 patients who had a preoperative diagnosis of cancer underwent definitive surgery. In 1 patient, whose FNAC had revealed atypical cells preoperatively, a histological diagnosis of a ruptured cyst with an inflammatory reaction was recorded. The remaining 23 patients had a presumptive preoperative diagnosis of lipoma. Histological examination confirmed the diagnosis of lipoma in 16 of these 23 cases (69.6%) (Table 3). Of the 82 women who were not operated upon, 2 (2.4%) declined to attend for follow-up. The other 80 were seen at clinical follow-up 6 months later. In 63 of these the lipoma was largely unchanged, and in 17 the tumour had disappeared.
Table 1 Mammography (MG) and ultrasound (US) findings.
Lipoma Fat only at tumour site Normal glandular tissue Fibroadenosis Fibroadenoma Cancer No examination Total
MG no. (%)
US no (%)
3 76 10 12
23 36 33 12 1 2 1
108
108
(2.8) (70.4) (9.3) (11.8) F 1 (0.9) 6 (5.6)
(21.3) (33.3) (30.6) (11.1) (0.9) (1.9) (0.9)
Table 2 Fine-needle aspiration cytology. No (%) No examination Inadequate material Fat cells only Benign epithelial cells Fat þ epithelial cells Possible fat necrosis Atypia þ fat cells Malignant cells Total
2 3 80 8 11 2 1 1 108
(1.9) (2.8) (74.0) (7.4) (10.2) (1.9) (0.9) (0.9)
ARTICLE IN PRESS 410
C. Lanng et al.
Analysis of triple diagnosis for lipoma
Discussion
When the 2 cases that were lost to follow-up were excluded the clinical diagnosis of lipoma was incorrect in 27 out of 106 cases (25.5%): 10 of 26 in the operated group and 17 of the 80 not operated on. There were 16 cases with a histological diagnosis of lipoma and we assume that the 63 women who were not operated and whose tumour was largely unchanged after 6 months’ observation also had lipomas. There were thus 79 patients with a lipoma available for this analysis (Fig. 1). Mammography had been done in 74 of these cases and showed a lipoma in only 2 (3.0%). Ultrasound had been carried out in 78 cases and revealed a lipoma in 16 (21.0%). FNAC revealed only fat cells in 63 of the 78 (80.8%) patients in whom it was done. In the remaining 15 cases FNAC was either inadequate or not consistent with the diagnosis of lipoma. In all, only 9 patients (11.4%) fulfilled the triple diagnostic criteria (lipoma on clinical examination and on ultrasound or mammography, and FNAC showing only fat cells).
We have been unable to determine the exact incidence of breast lipoma in the literature. The lesion is paradoxically described as both common5,6 and uncommon.7 One publication8 reports ‘‘only fat, predominantly fat or possible lipoma’’ in 18 out of 550 (3.2%) excision biopsies performed to exclude malignancy. Donegan4 reports an incidence of 2.2%. In our series, there were 16 histologically verified lipomas in a period in which we undertook 343 excision biopsies (4.6%). In the same period there were 108 women with a clinical diagnosis of lipoma among 3073 (3.5%) referred to the clinic with breast symptoms. The problem is thus not frequent, but it is common enough to require responsible treatment guidelines. It is interesting to note that even in our highly specialised environment the clinical diagnosis of lipoma by specialist breast surgeons was wrong in 25.5% of cases. Two cases regarded as lipomas on clinical examination by these specialists were revealed as cancers by supplementary tests. Mammography failed to show a lipoma in 97.2% of cases, and ultrasound carried out by specialist breast diagnosticians failed to demonstrate a lipoma in 78.7% of cases. What is the explanation for this poor clinical standard? We believe that many cases diagnosed as a lipoma, even by our experienced breast clinicians, were in fact prominent fat lobules which on excision would not be regarded as lipomas when examined under the microscope, because they lacked a capsule.2,8 This was so in 8% of our cases. We have no explanation for why these fat lobules tend to become less prominent after a period of observation, as they did in 20% of the cases in our series. These innocent fat lobules would have been unnecessarily excised had we strictly followed the rules of triple diagnosis. We have also shown that even experts can mistake cancers, fibroadenomas, phyllodes tumours and duct papillomas for lipomas. It should be mentioned that ‘‘pseudo’’-lipomas, harmatomas (adenolipomas), angiolipomas, angiomyolipomas and liposarcomas can mimic lipomas. It is also of interest that in 2 of our cases diagnosed clinically as lipomas were in fact localised areas of fat necrosis, a finding that has also been described by others.9 How do we explain our low detection rate at mammography? Although there are impressive descriptions of what lipomas look like on mammography,9 they may not be seen if they are small,4 if they are very large,5 or if the breast tissue is largely replaced by fat.5 In our series the median tumour size was 15 mm, which may partly explain why so many were missed on mammography.
Table 3
Histology after excision. No (%)
Lipoma Fibroadenoma Papilloma þ duct ectasia Fat necrosis Fatty tissue–no capsule Cancer Ruptured cyst þ inflammation Benign phyllodes tumour
16 1 1 2 2 2 1 1
(61.7) (3.8) (3.8) (7.7) (7.7) (7.7) (3.8) (3.8)
Total
26 (100)
Patients: 115
Follow-up: 82
Excised: 26
No follow-up:2 Not lipoma: 10 Disappeared: 17
Lipoma at follow-up: 63*
Lipoma at histology: 16*
Figure 1 Patients recruited for the study. *Available for assessment of triple diagnosis.
ARTICLE IN PRESS Lipoma of the breast
Our detection rate for lipoma was slightly better with ultrasound (21.3%), but this again is lower than would be expected in view of the convincing ultrasound features described in the literature.6 A possible explanation is that lipomas have an echogenicity similar to that of the surrounding fat on ultrasound.4 FNAC of a lipoma might be expected to reveal only fat cells.4 This was confirmed in almost 75% of our cases. If we accept an FNAC result with predominantly fat cells together with normal epithelial cells, confirmation of the lipoma diagnosis rises to nearly 86%. There is a simple explanation for the presence of epithelial cells in the FNAC: the lipoma offers no resistance to the inserted needle,4 so that the needle probably passes through the lipoma and also samples cells from the surrounding glandular tissue. In our opinion, the presence of normal epithelial cells in an FNAC revealing predominantly fat cells should not negate the diagnosis of lipoma. It is very important that the cytologist be informed of the clinical suspicion of lipoma so that the finding of fat cells does not lead to a cytological assessment of ‘‘inadequate’’. It is noteworthy that ultrasound detected the 2 cases of cancer in our study. Only 1 of the 2 was seen on mammography. Both cancers were detected at the initial assessment, and no further cancers emerged in the observation period. If the principle of triple diagnosis were followed religiously, all 106 of the noncancer patients would have had their tumours excised because either the mammography or the ultrasound had failed to confirm the diagnosis of lipoma or because the FNAC revealed only fat cells (thus, by definition, rendering the test inadequate) or normal epithelial cells (thus contradicting the diagnosis of lipoma). This would be a psychological burden for many patients and economically wasteful for the health services. In our series, 80 women were spared a redundant procedure by the choice of a conservative approach. Surgeons fearing litigation might be tempted to excise all lipomas because they seldom fulfil the rules of triple diagnosis. We feel that this is an example of surgical overkill and should be advised against. At the other extreme is the opinion of Pui and Movson,9 who suggest that a definitive diagnosis of lipoma is possible without histological confirmation if the mammographic appearance is characteristic. We dispute this claim.
411
Our proposal for a rational approach to the diagnosis and management of lipoma of the breast is as follows. A clinical diagnosis of lipoma must be confirmed by either FNAC revealing fat cells or fat cells plus normal epithelial cells, or a core biopsy consistent with a lipoma. The mammogram and the ultrasound need not necessarily demonstrate a lipoma, but obviously must not show anything arousing the suspicion of malignancy at the site. Mammography and ultrasound will usually show normal fatty tissue in the area concerned. If these four criteria are met, namely clinical diagnosis of lipoma, FNAC showing fat cells with or without normal epithelial cells or core biopsy consistent with a lipoma, and mammogram and ultrasound showing nothing suspicious at the site, then it is completely justifiable to offer the patient nonoperative management in the form of follow-up palpation after 6 months. (Since both the cancers in the survey were identified at the initial assessment, it might be argued that the 6-month followup is superfluous.) In all other situations, or when the diagnosis is in doubt, the tumour must be excised.1,4 Other indications for removal of a lipoma are patient’s choice, cosmetic deformity from large tumours, and rapid growth obvious on observation.1,4
References 1. Hughes LE, Mansel RE, Webster DJT. Miscellaneous conditions. In: Webster DJT, Hughes Le, authors. Benign disorders and diseases of the breast, 2nd ed. London: Saunders; 2000. p. 234. 2. Schnitt SJ, Connolly JL. Benign disorders. In: Harris JR, Lippmann ME, Morrow M, Hellmann S, editors. Diseases of the breast. 1st ed. Philadelphia: Lippincott-Raven; 1996. p. 37. 3. Perry NM. Quality assurance in the diagnosis of breast disease. Eur J Cancer 2001;37:159–72. 4. Donegan WL. Common benign conditions of the breast. In: Donegan WL, Spratt JS, editors. Cancer of the breast. 5th ed. Philadelphia: Saunders; 2002. p. 67–74. 5. Rodriquez LF, Shuster BA, Milliken RG. Giant lipoma of the breast. Br J Plast Surg 1997;50:263–5. 6. Steyaert L. The breast. In: Meire HB, Cosgrove DO, Dewbury KC, Farrant P, editors. Clinical ultrasound, 2nd ed. London: Churchill Livingstone; 2001. p. 777. 7. Brebner DM, Cosmann B, Shapiro J. Lipoma of the breast diagnosed by film and xeromammography. S Afr Med J 1976;50:685–8. 8. Hall FM, Connolly JL, Love SM. Lipomatous pseudomass of the breast. Radiology 1987;164:463–4. 9. Pui M, Movson IJ. Fatty tissue breast lesions. Clin Imaging 2003;27:150–5.
THE
BREAST www.elsevier.com/locate/breast
ORIGINAL ARTICLE
Lipoma of the breast: a diagnostic dilemma C. Lannga,*, B.Ø. Eriksenb, J. Hoffmanna a
Breast Surgery Clinic, Horsholm Hospital, Usserod Kongevej 102, Horsholm 2970, Denmark Department of Radiology, Horsholm Hospital, Horsholm, Denmark
b
Received 21 January 2004; received in revised form 20 April 2004; accepted 27 April 2004
KEYWORDS Breast; Lipoma; Triple diagnosis; Mammography; Ultrasound; Fine-needle aspiration cytology
Summary Lipoma of the breast often causes diagnostic and therapeutic uncertainty. Clinically it may be difficult to distinguish a lipoma from other conditions. Fineneedle aspiration cytology (FNAC) is often not helpful. Both mammography and ultrasound scanning are often negative. For the present study, 108 women with a clinical diagnosis of lipoma were enrolled prospectively. The clinical diagnosis of lipoma was found to be incorrect in 25.0% of these cases. Mammography and ultrasound revealed a lipoma in only 3.0% and 21.0%, respectively. FNAC revealed only fat cells in 74.0% of cases. In all, only 9 patients (11.4%) fulfilled the triple diagnostic criteria, theoretically making tumour excision mandatory in the remaining cases. Our proposal for management is for any clinical diagnosis of lipoma to be confirmed by either FNAC revealing fat cells or a core biopsy consistent with a lipoma. The mammogram and the ultrasound need not necessarily demonstrate a lipoma, but obviously must not show anything to raise the suspicion of malignancy at the site. If these criteria are met it is not necessary to excise the tumour. & 2004 Elsevier Ltd. All rights reserved.
Introduction Lipoma of the breast is a benign condition that provokes little interest in the literature and scant mention in standard texts.1,2 Although it is a banal condition it often causes diagnostic uncertainty, which results in confusion concerning treatment strategies. This uncertainty can arise at any point in the diagnostic process: clinically it may be difficult to distinguish a lipoma from a prominent *Corresponding author. Tel.: þ 45-4829-6965; fax: þ 45-48292667. E-mail address: [email protected] (C. Lanng).
fat lobule or from other benign or malignant processes. Fine-needle aspiration cytology (FNAC) reveals either fat cells only, which by definition renders the test ‘inadequate’,3 or fat cells together with normal epithelial cells, which casts doubt on the diagnosis of lipoma. Both mammography and ultrasound scanning are often negative.4,5 This results in a situation where it is difficult to fulfil the requirements of triple diagnosis, which may culminate in unnecessary tumour excision with all its psychological and economic consequences. A prospective investigation was planned with the object of documenting this problem and designing a suitable treatment strategy.
0960-9776/$ - see front matter & 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.breast.2004.04.011
ARTICLE IN PRESS Lipoma of the breast
Patients and methods The Breast Surgery Clinic, Horsholm Hospital, is a regional breast care centre serving a population of 380,000. The clinic receives over 3000 new referrals per year, about 12% of which are for breast cancer. The clinic is served by five full-time breast surgeons, two full-time diagnostic radiologists and three full-time pathologists. Between 1 January and 31 December 2002 a prospective registration was carried out: All women referred with a breast lump were first assessed by one of the five surgeons. If a clinical diagnosis of lipoma was made the patient was enrolled. The clinical criterion was a well-defined, smooth, possibly slightly lobulated, soft or semi-firm, mobile tumour detected on palpation. The mammographic and ultrasound findings were recorded. The mammographic criteria were: well defined, encapsulated radiotranslucency, often with compression of adjacent breast tissue. The ultrasound criteria were: oval, lobulated, homogeneous, solid mass with an echogenicity similar to that of normal fat. The results of the FNAC and/or core biopsy were recorded. After being given clear information about the presumed diagnosis, the patient was asked to choose between observation or excision. If excision was chosen the histological result was recorded. If observation was preferred a clinical reassessment was carried out 6 months later.
Results A clinical diagnosis of lipoma was recorded in 108 women. Their ages ranged from 26 to 88 years (median 51 years). The tumour size ranged from 5 to 35 mm (median 15 mm). Mammography revealed a lipoma in only 3 of 102 cases (2.8%). (Three women were not examined mammographically because of their young age and 3, because they had a mammography showing a normal result within the previous 12 months.) In 1 case the mammogram showed that the palpable tumour was a cancer. Further details are shown in Table 1. Ultrasound examination showed a lipoma in 23 out of 107 cases (21.3%). In 2 cases ultrasound revealed that the tumour was a cancer. Further details are shown in Table 1. FNAC was done in 106 cases and showed fat cells only in 80 cases (74.0%), fat cells and epithelial cells in 11 and malignant cells in 1; it
409
was judged inadequate in 3. Further details are shown in Table 2. Core needle biopsy was performed in addition to FNAC in 7 cases and showed a lipoma in 3, fibroadenosis in 1, fat tissue in 4, and cancer in 1. Another 2 patients underwent core biopsy instead of FNAC, revealing a cancer in 1 and a lipoma in the other. Twenty-four patients chose to undergo excision of their tumour. The 2 patients who had a preoperative diagnosis of cancer underwent definitive surgery. In 1 patient, whose FNAC had revealed atypical cells preoperatively, a histological diagnosis of a ruptured cyst with an inflammatory reaction was recorded. The remaining 23 patients had a presumptive preoperative diagnosis of lipoma. Histological examination confirmed the diagnosis of lipoma in 16 of these 23 cases (69.6%) (Table 3). Of the 82 women who were not operated upon, 2 (2.4%) declined to attend for follow-up. The other 80 were seen at clinical follow-up 6 months later. In 63 of these the lipoma was largely unchanged, and in 17 the tumour had disappeared.
Table 1 Mammography (MG) and ultrasound (US) findings.
Lipoma Fat only at tumour site Normal glandular tissue Fibroadenosis Fibroadenoma Cancer No examination Total
MG no. (%)
US no (%)
3 76 10 12
23 36 33 12 1 2 1
108
108
(2.8) (70.4) (9.3) (11.8) F 1 (0.9) 6 (5.6)
(21.3) (33.3) (30.6) (11.1) (0.9) (1.9) (0.9)
Table 2 Fine-needle aspiration cytology. No (%) No examination Inadequate material Fat cells only Benign epithelial cells Fat þ epithelial cells Possible fat necrosis Atypia þ fat cells Malignant cells Total
2 3 80 8 11 2 1 1 108
(1.9) (2.8) (74.0) (7.4) (10.2) (1.9) (0.9) (0.9)
ARTICLE IN PRESS 410
C. Lanng et al.
Analysis of triple diagnosis for lipoma
Discussion
When the 2 cases that were lost to follow-up were excluded the clinical diagnosis of lipoma was incorrect in 27 out of 106 cases (25.5%): 10 of 26 in the operated group and 17 of the 80 not operated on. There were 16 cases with a histological diagnosis of lipoma and we assume that the 63 women who were not operated and whose tumour was largely unchanged after 6 months’ observation also had lipomas. There were thus 79 patients with a lipoma available for this analysis (Fig. 1). Mammography had been done in 74 of these cases and showed a lipoma in only 2 (3.0%). Ultrasound had been carried out in 78 cases and revealed a lipoma in 16 (21.0%). FNAC revealed only fat cells in 63 of the 78 (80.8%) patients in whom it was done. In the remaining 15 cases FNAC was either inadequate or not consistent with the diagnosis of lipoma. In all, only 9 patients (11.4%) fulfilled the triple diagnostic criteria (lipoma on clinical examination and on ultrasound or mammography, and FNAC showing only fat cells).
We have been unable to determine the exact incidence of breast lipoma in the literature. The lesion is paradoxically described as both common5,6 and uncommon.7 One publication8 reports ‘‘only fat, predominantly fat or possible lipoma’’ in 18 out of 550 (3.2%) excision biopsies performed to exclude malignancy. Donegan4 reports an incidence of 2.2%. In our series, there were 16 histologically verified lipomas in a period in which we undertook 343 excision biopsies (4.6%). In the same period there were 108 women with a clinical diagnosis of lipoma among 3073 (3.5%) referred to the clinic with breast symptoms. The problem is thus not frequent, but it is common enough to require responsible treatment guidelines. It is interesting to note that even in our highly specialised environment the clinical diagnosis of lipoma by specialist breast surgeons was wrong in 25.5% of cases. Two cases regarded as lipomas on clinical examination by these specialists were revealed as cancers by supplementary tests. Mammography failed to show a lipoma in 97.2% of cases, and ultrasound carried out by specialist breast diagnosticians failed to demonstrate a lipoma in 78.7% of cases. What is the explanation for this poor clinical standard? We believe that many cases diagnosed as a lipoma, even by our experienced breast clinicians, were in fact prominent fat lobules which on excision would not be regarded as lipomas when examined under the microscope, because they lacked a capsule.2,8 This was so in 8% of our cases. We have no explanation for why these fat lobules tend to become less prominent after a period of observation, as they did in 20% of the cases in our series. These innocent fat lobules would have been unnecessarily excised had we strictly followed the rules of triple diagnosis. We have also shown that even experts can mistake cancers, fibroadenomas, phyllodes tumours and duct papillomas for lipomas. It should be mentioned that ‘‘pseudo’’-lipomas, harmatomas (adenolipomas), angiolipomas, angiomyolipomas and liposarcomas can mimic lipomas. It is also of interest that in 2 of our cases diagnosed clinically as lipomas were in fact localised areas of fat necrosis, a finding that has also been described by others.9 How do we explain our low detection rate at mammography? Although there are impressive descriptions of what lipomas look like on mammography,9 they may not be seen if they are small,4 if they are very large,5 or if the breast tissue is largely replaced by fat.5 In our series the median tumour size was 15 mm, which may partly explain why so many were missed on mammography.
Table 3
Histology after excision. No (%)
Lipoma Fibroadenoma Papilloma þ duct ectasia Fat necrosis Fatty tissue–no capsule Cancer Ruptured cyst þ inflammation Benign phyllodes tumour
16 1 1 2 2 2 1 1
(61.7) (3.8) (3.8) (7.7) (7.7) (7.7) (3.8) (3.8)
Total
26 (100)
Patients: 115
Follow-up: 82
Excised: 26
No follow-up:2 Not lipoma: 10 Disappeared: 17
Lipoma at follow-up: 63*
Lipoma at histology: 16*
Figure 1 Patients recruited for the study. *Available for assessment of triple diagnosis.
ARTICLE IN PRESS Lipoma of the breast
Our detection rate for lipoma was slightly better with ultrasound (21.3%), but this again is lower than would be expected in view of the convincing ultrasound features described in the literature.6 A possible explanation is that lipomas have an echogenicity similar to that of the surrounding fat on ultrasound.4 FNAC of a lipoma might be expected to reveal only fat cells.4 This was confirmed in almost 75% of our cases. If we accept an FNAC result with predominantly fat cells together with normal epithelial cells, confirmation of the lipoma diagnosis rises to nearly 86%. There is a simple explanation for the presence of epithelial cells in the FNAC: the lipoma offers no resistance to the inserted needle,4 so that the needle probably passes through the lipoma and also samples cells from the surrounding glandular tissue. In our opinion, the presence of normal epithelial cells in an FNAC revealing predominantly fat cells should not negate the diagnosis of lipoma. It is very important that the cytologist be informed of the clinical suspicion of lipoma so that the finding of fat cells does not lead to a cytological assessment of ‘‘inadequate’’. It is noteworthy that ultrasound detected the 2 cases of cancer in our study. Only 1 of the 2 was seen on mammography. Both cancers were detected at the initial assessment, and no further cancers emerged in the observation period. If the principle of triple diagnosis were followed religiously, all 106 of the noncancer patients would have had their tumours excised because either the mammography or the ultrasound had failed to confirm the diagnosis of lipoma or because the FNAC revealed only fat cells (thus, by definition, rendering the test inadequate) or normal epithelial cells (thus contradicting the diagnosis of lipoma). This would be a psychological burden for many patients and economically wasteful for the health services. In our series, 80 women were spared a redundant procedure by the choice of a conservative approach. Surgeons fearing litigation might be tempted to excise all lipomas because they seldom fulfil the rules of triple diagnosis. We feel that this is an example of surgical overkill and should be advised against. At the other extreme is the opinion of Pui and Movson,9 who suggest that a definitive diagnosis of lipoma is possible without histological confirmation if the mammographic appearance is characteristic. We dispute this claim.
411
Our proposal for a rational approach to the diagnosis and management of lipoma of the breast is as follows. A clinical diagnosis of lipoma must be confirmed by either FNAC revealing fat cells or fat cells plus normal epithelial cells, or a core biopsy consistent with a lipoma. The mammogram and the ultrasound need not necessarily demonstrate a lipoma, but obviously must not show anything arousing the suspicion of malignancy at the site. Mammography and ultrasound will usually show normal fatty tissue in the area concerned. If these four criteria are met, namely clinical diagnosis of lipoma, FNAC showing fat cells with or without normal epithelial cells or core biopsy consistent with a lipoma, and mammogram and ultrasound showing nothing suspicious at the site, then it is completely justifiable to offer the patient nonoperative management in the form of follow-up palpation after 6 months. (Since both the cancers in the survey were identified at the initial assessment, it might be argued that the 6-month followup is superfluous.) In all other situations, or when the diagnosis is in doubt, the tumour must be excised.1,4 Other indications for removal of a lipoma are patient’s choice, cosmetic deformity from large tumours, and rapid growth obvious on observation.1,4
References 1. Hughes LE, Mansel RE, Webster DJT. Miscellaneous conditions. In: Webster DJT, Hughes Le, authors. Benign disorders and diseases of the breast, 2nd ed. London: Saunders; 2000. p. 234. 2. Schnitt SJ, Connolly JL. Benign disorders. In: Harris JR, Lippmann ME, Morrow M, Hellmann S, editors. Diseases of the breast. 1st ed. Philadelphia: Lippincott-Raven; 1996. p. 37. 3. Perry NM. Quality assurance in the diagnosis of breast disease. Eur J Cancer 2001;37:159–72. 4. Donegan WL. Common benign conditions of the breast. In: Donegan WL, Spratt JS, editors. Cancer of the breast. 5th ed. Philadelphia: Saunders; 2002. p. 67–74. 5. Rodriquez LF, Shuster BA, Milliken RG. Giant lipoma of the breast. Br J Plast Surg 1997;50:263–5. 6. Steyaert L. The breast. In: Meire HB, Cosgrove DO, Dewbury KC, Farrant P, editors. Clinical ultrasound, 2nd ed. London: Churchill Livingstone; 2001. p. 777. 7. Brebner DM, Cosmann B, Shapiro J. Lipoma of the breast diagnosed by film and xeromammography. S Afr Med J 1976;50:685–8. 8. Hall FM, Connolly JL, Love SM. Lipomatous pseudomass of the breast. Radiology 1987;164:463–4. 9. Pui M, Movson IJ. Fatty tissue breast lesions. Clin Imaging 2003;27:150–5.