Management of Breast Fibroadenomas PDF

Management of Breast Fibroadenomas Ron Greenberg, MD, Yehuda Skornick, MD, Ofer Kaplan, MD OBJECTIVE: To identify from the literature and clinical expe- INCIDENCE AND RISK FACTORS rience a rational approach to management of fibroadenomas of the breast....

Management of Breast Fibroadenomas Ron Greenberg, MD, Yehuda Skornick, MD, Ofer Kaplan, MD OBJECTIVE: To identify from the literature and clinical expe- INCIDENCE AND RISK FACTORS rience a rational approach to management of fibroadenomas of the breast. There are no clear-cut data on the incidence of fi- broadenomas in the general population. In one study, the METHOD: Recent literature on detection, diagnosis, and nat- rate of occurrence of fibroadenomas in women who were ural history of fibroadenomas was reviewed. Experience with examined in breast clinics was 7% to 13%,1 while it was over 4,000 women evaluated in the breast clinic at the Tel- Aviv Medical Center contributed to the management strate- 9% in another study of autopsies.2 Fibroadenomas com- gies suggested by review of the literature. prise about 50% of all breast biopsies, and this rate rises to 75% for biopsies in women under the age of 20 years.3,4 RESULTS: Fibroadenomas of the breast are common, ac- Fibroadenomas are more frequent among women in higher counting for 50% of all breast biopsies performed. Physical socioeconomic classes5–7 and in dark-skinned popula- examination, sonography, and fine needle aspiration are ef- fective in distinguishing fibroadenomas from breast cancer. tions.8 The age of menarche, the age of menopause, and Transformation from fibroadenoma to cancer is rare; regres- hormonal therapy, including oral contraceptives, were sion or resolution is frequent, supporting conservative ap- shown not to alter the risk of these lesions.6,7,9,10 Con- proaches to follow-up and management. versely, body mass index and the number of full-term pregnancies were found to have a negative correlation CONCLUSION: Age-based algorithms that allow for conserva- tive management and that limit excision to patients whose with the risk of fibroadenomas.5–7,9,11 Moreover, consump- fibroadenomas fail to regress are presented. tion of large quantities of vitamin C and cigarette smoking were found to be associated with reduced risk of a fi- KEY WORDS: fibroadenoma; breast neoplasms; women. broadenoma.7,12,13 J GEN INTERN MED 1998;13:640–645. No genetics factors are known to alter the risk of fibro- adenoma. However, a family history of breast cancer in first- degree relatives was reported by some investigators to be F ibroadenomas are common benign lesions of the breast that usually present as a single breast mass in young women. They are assumed to be aberrations of nor- related with increased risk of developing these tumors.14,15 mal breast development or the product of hyperplastic PATHOLOGY processes, rather than true neoplasms. The clinician of- Fibroadenomas usually form during menarche (15–25 ten faces the dilemma whether to remove the mass or to years of age), a time at which lobular structures are added to monitor it by means of periodic follow-up examinations. the ductal system of the breast (Fig. 1). Hyperplastic lobules Although removal of these lesions is a definitive solution, are common at that time, and may be regarded as a normal surgery may involve unnecessary excisions of benign le- phase of breast development.16 Hyperplastic lobules were sions and unbecoming cosmesis. Moreover, a policy of con- shown to be histologically identical with fibroadeno- ducting surgery on all patients with fibroadenomas would mas.10,17 Analyses of the cellular components of fibroadeno- place an enormous burden on health care systems. A bal- mas by means of polymerase chain reaction demonstrated anced and rational approach to the management of a fi- that both the stromal and the epithelial cells are poly- broadenoma of the breast needs to address the crucial clonal,18 supporting the theory that fibroadenomas are hy- questions about its association with breast cancer, espe- perplastic lesions associated with aberration of the normal cially whether or not it is a marker of increased risk of maturation of the breast, rather than true neoplasms.16,18 breast malignancy. Another consideration to be weighed The pattern of stromal growth in a fibroadenoma de- is that a substantial percentage of these lesions undergo pends on its epithelial component: stromal mitotic activity spontaneous regression. Herein, based on our review of was found to be higher near this component.19 Fibroade- the current data on fibroadenomas of the breast and our nomas are stimulated by estrogen and progesterone, and experience, we propose practical algorithms for their by lactation during pregnancy, and they undergo atrophic management. changes in menopause.16 Some fibroadenomas have re- ceptors and respond to growth hormone and epidermal growth factor.20 Received from the Department of Surgery A, Tel-Aviv Medical CLINICAL PRESENTATION AND DIAGNOSIS Center, and the Sackler Faculty of Medicine, Tel-Aviv Univer- sity, Tel-Aviv, Israel. A fibroadenoma is most often detected incidentally Address correspondence and reprint requests to Dr. Kaplan: during a medical examination or during self examination, Dept. of Surgery A, Tel-Aviv Sourasky Medical Center, 6 usually as a discrete solitary breast mass of 1 to 2 cm.15,21 Weizmann St., Tel-Aviv 64239, Israel. Although they can be located anywhere in the breast, the 640 JGIM Volume 13, September 1998 641 Multiple Fibroadenomas From 10% to 16% of patients with multiple fibroade- nomas have two to four in a single breast, which may present initially or be discovered over several years.15,22 Un- like women with a single fibroadenoma, most of the pa- tients with multiple fibroadenomas have a strong family history of these tumors.26 A possible connection between multiple fibroadenomas and oral contraceptives was pro- posed but has not yet been substantiated.27 Giant and Juvenile Fibroadenomas Fibroadenomas larger than 5 cm (about 4% of the to- tal) are commonly defined as being giant fibroadenomas;21 FIGURE 1. Histologic section of a fibroadenoma (hematoxylin- however, this terminology is not universally accepted. Gi- eosin staining, 3 40). The cellular fibroblastic stroma, which re- ant fibroadenomas are usually encountered in pregnant sembles intralobular stroma, encloses glandular and cystic spaces lined by epithelium. Round and oval gland spaces, or lactating women. When found in an adolescent girl, the lined by either single or multiple cell layers, are present in other term juvenile fibroadenoma is more appropriate.15 These areas. The stroma in the connective tissue appears to have lesions in young women constitute 0.5% to 2% of all fi- undergone a more active proliferation with compression on broadenomas, and are rapidly growing masses that cause the gland spaces. asymmetry of the breast, distortion of the overlying skin, and stretching of the nipple. Histologically, they appear to be more cellular and have less lobular components than do simple fibroadenomas. However, giant fibroadenomas majority are situated in the upper outer quadrant.22 A fi- are benign lesions that do not undergo transformation broadenoma is usually smooth, mobile, nontender, and into malignancy.28 rubbery in consistency (Fig. 2). Several other breast le- sions have similar characteristics, and physical examina- tions provided an accurate diagnosis in only one half to two thirds of cases studied.23,24 However, most of the IMAGING TECHNIQUES masses that are erroneously diagnosed by palpation as fi- Sonography broadenomas are found on histologic examination to be another benign form of breast disease,25 such as cystic Breast sonography is often used for the diagnosis of fibrosis. fibroadenomas. The sonographic criteria that support the diagnosis of a fibroadenoma are a round or oval solid mass with a smooth contour and weak internal echoes in a uniform distribution and intermediate acoustic atten- uation29 (Fig. 3). This imaging technique is very useful for differentiating between solid and cystic lesions. However, attempts to correlate between the sonographic features of solid masses compatible with fibroadenomas and patho- logic findings were disappointing.30 There is some overlap in the sonographic criteria for fibroadenomas and for breast cancer,31 and approximately 25% of fibroadenomas appear with irregular margins, which may imply that the lesions are malignant.29 Also, only 82% of biopsy-proven fibroadenomas were visualized by sonography in one study.29 Mammography The yield of mammography in young women is low, and its role in the diagnosis of fibroadenomas is limited. FIGURE 2. Macroscopic appearance of a fibroadenoma. The However, it may disclose features of infiltrative lesions in spherical mass is sharply circumscribed, and could be easily older women. In the mammographic image, fibroadenomas separated from the surrounding breast tissue. The section mar- appear as soft, homogenous, and well-circumscribed nod- gins have a green-white color, and contain slit-like spaces. ules, and inner coarse calcifications are often observed. 642 Greenberg et al., Management of Breast Fibroadenomas JGIM ASPIRATION CYTOLOGY retrospective studies,14,35–41 which demonstrated a 1.3 to 2.1 increased risk of breast cancer in women with fi- Fine needle aspiration (FNA) has become a popular broadenomas compared with the general population. The method in the evaluation of breast masses. The charac- elevated risk was persistent, and did not decrease with teristic cytologic features of fibroadenomas are: clusters time. A more recent study designed to delineate the possi- of spindle cells without inflammatory or fat cells, found in ble correlation between the histologic features of the fi- 96% of all fibroadenomas; aggregates of cells with a papil- broadenomas and the risk for subsequent breast cancer lary configuration resembling elk antler (antler horn clus- used the term “complex fibroadenoma.”15 This term ap- ters), found in 93% of all cases; and uniform cells with plies to fibroadenomas having the histologic characteris- well-defined cytoplasm lying in rows and columns (honey- tic of being more than 3 mm in diameter, or with elements comb sheets), found in 95% of all fibroadenomas.32 Taken of sclerosing adenosis, epithelial calcifications, or papil- together with the clinical diagnosis of fibroadenoma, FNA lary apocrine metaplasia, which were associated with a can improve the sensitivity of the diagnosis to 86% with a 3.1 elevated risk of breast cancer. Proliferative changes in specificity of 76%,21,30 while for breast cancer FNA is 96% the parenchyma adjacent to the fibroadenoma were re- sensitive and 98% specific. Thus, while aspiration cytol- lated to a further increase of the risk to 3.88. The relative ogy may confuse fibroadenomas with other benign breast risk for women with a familial history of breast cancer lesions, incorrect diagnosis of a malignant process is rare. and complex fibroadenoma was 3.72, compared with con- The overall diagnostic efficacy of these three modali- trol women with a family history of breast cancer without ties—namely, manual breast examination, imaging and fibroadenoma. In these studies, women with noncomplex cytology is approximately 70% to 80%, but they provide a fibroadenomas and no family history of breast cancer 95% (62% SD) accurate differentiation between a benign were not at a greater risk of breast cancer. and a malignant lesion. A follow-up period of 1 to 3 years Malignant transformations in the epithelial compo- after fibroadenoma is diagnosed and breast cancer is ex- nents of fibroadenomas are generally considered rare. The cluded using the three modalities can enhance the accu- incidence of a carcinoma evolving within a fibroadenoma racy of the diagnosis.33,34 was reported to be 0.002% to 0.0125%.42,43 About 50% of these tumors were lobular carcinoma in situ (LCIS), 20% FIBROADENOMA AND BREAST CANCER were infiltrating lobular carcinoma, 20% were ductal car- cinoma in situ (DCIS), and the remaining 10% were infil- Any analysis of the associations of fibroadenomas trating ductal carcinoma. The clinical, sonographic and with breast cancer must address two main questions: mammographic findings are usually similar to those of whether or not a fibroadenoma is a marker for increased benign fibroadenomas,44,45 and the malignant changes risk of breast cancer, and whether or not breast cancer are often noted only when the fibroadenoma is excised. can evolve from the epithelial component of a fibroade- In a clinicopathologic study of 105 women with carci- noma. The first issue was originally assessed in several noma developing within fibroadenomas, the mean age was higher than in patients with benign fibroadenomas (44 vs 23 years).33,34,46 However, in that study, DCIS and LCIS in equal frequencies comprised 95% of the cases, and carcinoma in situ was also present in the adjacent breast tissue in about 20% of these women. No axillary metastases were found in any of the study patients. NATURAL HISTORY There are inherent obstacles in studying the natural course of breast fibroadenomas, and the data are not un- equivocal. Some investigators believe that most fibroade- nomas grow over a 12-month period to gain a size of 2 to 3 cm, after which they remain unchanged for several years.15 As definite diagnosis can be obtained only from histologic sections, solitary solid masses usually have been excised, and long term follow-up surveys are limited in number. These studies followed young women for up to 29 years, and regression or complete resolution of the fibroade- FIGURE 3. Sonographic appearance of a fibroadenoma. The nomas were noted in 16% to 59% of all cases.21,23,46,47 It mass is homogenous, with sharp and smooth margins. Slight was extrapolated that the probability that a fibroadenoma posterior and edge enhancements are visible. Neither com- would resolve after 5 years is approximately 50%, and the pression effects nor internal echoes are present. “lifetime” of a fibroadenoma is about 15 years.34 Among JGIM Volume 13, September 1998 643 the 50% of fibroadenomas that did not regress spontane- Not all women can be candidates for conservative treat- ously, about half did not change, and the remaining 25% ment: the patient’s age, a family history of malignancy, enlarged in size during the follow-up.21 and any data on proliferative changes in the breasts from From their incidence in mastectomy specimens, it previous biopsies must be taken into consideration. has been assumed that fibroadenomas tend to regress The risk of missing breast cancer in women under 25 and loss their cellularity with age. The rare finding of fi- years of age who have fibroadenomas as diagnosed by broadenomas in the older age groups also supports the physical examination, sonography, and FNA is 1 in 229 to hypothesis of regression of fibroadenomas.48 The mecha- 1 in 700.21,24 This risk remains very low in women under nisms offered to explain the regression of fibroadenomas the age of 35 years. Therefore, it has been recommended are infarction, calcification, and hyalinization.15,49 that young patients should be observed with frequent clinical evaluations, and the lesions excised in women over the age of 35 years.22,23,30 Other investigators sug- TREATMENT gested that the cutoff age should be 25 years.33 As fibroadenomas are benign breast lesions, it could The preferred management of multiple fibroadeno- be argued that they should not be excised and can be ex- mas is complete excision. However, this approach can pected to regress spontaneously. Moreover, 30% of breast lead to undesirable scarring or to extensive ductal dam- tumors that are diagnosed as fibroadenomas are found age if all the fibroadenomas are excised through one inci- postsurgically to be other types of benign lesions. In Cant sion.26,50 Giant fibroadenomas tend to shrink after cessa- et al.’s follow-up studies on clinically diagnosed fibroade- tion of lactation, so their removal should be delayed until nomas, persistent lesions were excised after 3 years: fi- the patient’s hormonal status returns to normal, and a broadenomas were found in the histologic examinations smaller excision can be performed.15,21 It may be very dis- of 97% of these cases.33,34 These findings suggest that the figuring to excise juvenile fibroadenomas because of their other benign lesions had resolved spontaneously during large sizes; nevertheless, no recurrences were reported af- 1 to 3 years, that the remaining masses were true fibroad- ter complete excision, and normal and symmetrical devel- enomas, and that conservative management is warranted. opment of the breasts can be anticipated.28,51 FIGURE 4. Management of a fibroadenoma (FA) in women younger than 35 years of age. 644 Greenberg et al., Management of Breast Fibroadenomas JGIM MANAGEMENT recommendation is to treat a woman with multiple fi- broadenomas in the same manner as a woman with a sin- In our breast clinic at the Tel-Aviv Medical Center, gle lesion. Physical examination, ultrasound, and FNA more than 4,000 women are examined each year. From ex- should be done, and if a diagnosis of multiple fibroade- perience, we believe that if conservative management of noma can be made with confidence, conservative treat- fibroadenoma is to be advocated, physical examination, ment with follow-up every 6 months should be recom- sonography, and FNA should all be performed, and their mended. Excisional biopsy is advised for any mass for results should be compatible with fibroadenoma. In women which the diagnosis is not clear-cut. older than 35 years, mammography should also be carried out. Because of the increased incidence of carcinoma in older women, we recommend two different treatment ap- REFERENCES proaches for fibroadenomas in women younger and older 1. Dent DM, Hacking EA, Wilkie W. Benign breast disease clinical than 35. classification and disease distribution. Br J Clin Pract. 1988;42 For women in which a fibroadenoma is diagnosed be- (suppl 56):69–71. fore the age of 35, we recommend conservative manage- 2. Franyz VK, Pickern JW, Melcher GW, Auchincoloss JR. Incidence of chronic cystic disease in so-called normal breast: a study based ment with a protocol of follow-up every 6 months in order on 225 post mortem examinations. Cancer. 1951;4:762–7. to detect any changes of the lesion (Fig. 4). In cases of re- 3. Schuerch C, Rosen PP, Hirota T, Itabashi M. A pathologic study of gression, the follow-up should continue until complete benign breast disease in Tokyo and New York. Cancer. 1982;50: regression. Fibroadenomas that either do not completely 1899–902. regress, or remain unchanged by the age of 35, should 4. Onuigb WIB. Adolescent mass in Nigerian igbos. Am J Surg. 1979; 137:367–71. be excised surgically. Fibroadenomas that become larger 5. Brinton LA, Vajsey MP, Flavel R. Risk factors for benign breast should be excised without delay. In patients with a family disease. Am J Epidemiol. 1981;113:203–14. history of breast cancer, or known changes of complex fi- 6. Soini I, Aine R, Lauslthti K. Independent risk factor of benign and broadenoma, we recommend excisional biopsy shortly af- malignant breast lesion. Am J Epidemiol. 1981;114:507–14. ter diagnosis has been established. 7. Yu H, Rohan TE, Cook MG, Howe GR. Risk factor for fibroade- noma: a case control study in Australia. Am J Epidemiol. 1992; When a fibroadenoma is detected in a woman older 135:247–58. than 35 years of age, and all the findings of the above- 8. Funder Burk WW, Rosero E, Leffall LD. Breast lesions in blacks. mentioned diagnostic modalities (including mammogra- Surg Gynecol Obstet. 1972;135:58–61. phy) support this diagnosis, a short follow-up period of 6-12 9. Ravnihar B, Segel DG, Lindther J. An epidemiologic study of months is justified because other benign changes may re- breast cancer and benign breast neoplasm in relation to the oral contraceptive and estrogen use. Eur J Cancer. 1979;15:395–405. solve and an operation can be obviated (Fig. 5). After 12 10. Canny PF, Berkowitz GS, Kelsey JL. Fibroadenoma and the use of months, any persistent fibroadenoma should be excised. exogenous hormones: a case control study. Am J Epidemiol. Our experience, and that reported by others,52 is that 1988;127:454–61. when offered the option of conservative management, most 11. Parazzini F, La Vecchia C, Franceshi S. Risk factors for pathologi- women will eventually prefer excisional biopsy. Our cally confirmed benign breast disease. Am J Epidemiol. 1984;120: 115–22. 12. Berkowitz GS, Canny PF, Vivolsy VA. Cigarette smoking and benign breast disease. J Epidemiol Community Health. 1985;39:308–13. 13. Rohan TE, Cook MG, Potter JD. A case control study of diet and benign proliferative epithelial disorder of the breast. Cancer Res. 1990;50:3176–81. 14. Dupont WD, Page DL, Park FF. Long-term risk for breast cancer in women with fibroadenoma. N Engl J Med. 1994;331:10–15. 15. Haagensen CD. Disease of the breast. 3rd ed. Philadelphia, Pa; W.B. Saunders; 1996:267–83. 16. Hughes LE, Mansel RE, Webster DJT. Aberration of normal devel- opment and involution: a new perspective on pathogenesis and nomenclature of benign breast disorders. Lancet. 1987;11:1316–9. 17. Parks AG. The micro anatomy of the breast. Ann R Coll Surg Engl. 1959;25:235–51. 18. Noquchi S, Motomura K, Inaji H, Imorka S. Clonal analysis of fi- broadenoma and phylloides tumor of the breast. Cancer Res. 1993;53(17):4071–4. 19. Sawhney N, Garrahan N, Douglas Jones AG, Williams ED. Epithe- lial stromal interaction in tumors: a morphologic study of fibroepi- thelial tumors of the breast. Cancer. 1992;70(8):2115–20. 20. Van Agthoven T, Timmerans M, Foekens JA, Dorssers LC. Differ- ential expression of estrogen, progesterone and epidermal growth factor receptors in normal, benign and malignant human breast tissues using dual staining immunohistochemistry. Am J Pathol. FIGURE 5. Management of a fibroadenoma (FA) in women 1994;144(6):1238–46. older than 35 years of age. 21. Dent DM, Cant PJ. Fibroadenoma. World J Surg. 1989;13:706–10. JGIM Volume 13, September 1998 645 22. Foster ME, Garrahan N, Williams S. Fibroadenoma of the breast: a cancer in women with benign breast disease. J Natl Cancer Inst. clinical and pathological study. J R Coll Surg Edinb. 1988;33:13–6. 1980;65:13–20. 23. Wilkinson S, Anderson TJ, Rifkind E, Chetty U. Fibroadenoma of 38. Moskowits M, Gartside P, Wirman JA, McLaughlin C. Proliferative the breast: a follow up of conservative management. Br J Surg. disorder of the breast as risk factors for breast cancer in a self- 1989;76:390–1. selected screened population. Radiology. 1980;134:289–91. 24. Cant PJ, Learmonth GM, Dent DM. When can fibroadenoma be 39. Carter CL, Corle DK, Micozzi MS, Schatzkin A, Tailor PR. A pro- managed conservatively. Br J Clin Pract. 1988;42(suppl 56):62–6. spective study of the development of breast cancer in 16,692 wo- 25. Carty NS, Carter C, Rubin C, Ravichandran D. Management of fi- men with benign breast disease. Am J Epidemiol. 1988;128:467–77. broadenoma of the breast. Ann R Coll Surg Engl. 1995;77:127–30. 40. McDivitt RW, Stevens JA, Lee MC, Wingo PA. Histologic types of 26. Williamson ME, Lyons K, Hyghes LE. Multiple fibroadenoma of the benign breast disease and the risk for breast cancer. Cancer. breast a problem of uncertain incidence and management. Ann R 1992;69:1408–14. Coll Surg Engl. 1993;75:161–3. 41. Krieger N, Hiatt RA. Risk of breast cancer after benign breast dis- 27. Gregg WI. Gallactorrhea after contraceptive hormones. N Engl J eases: variation by histologic type, degree of atypia, age at biopsy, Med. 1966;274:1432–4. and length of follow up. Am J Epidemiol. 1992;135:619–31. 28. Pike AM, Oberman HA. Juvenile cellular adenofibromas. Am J 42. Deschenes L, Jacob S, Fobia J, Christen A. Beware of breast fi- Surg Pathol. 1985;9:730–5. broadenoma in middle aged women. Can J Surg. 1985;28:372–3 29. Cale Beuglet C, Soriano RZ, Kurtz AB, Goldberg BB. Fibroade- 43. Bazanowski Konarky K, Harrison EG, Payne WS. Lobular carci- noma of the breast sonomammography correlated with pathology noma arising fibroadenoma of the breast. Cancer. 1975;35:450–6. in 122 patients. AJR. 1983;140:369–75. 44. Pick PW, Iossifide IA. Occurrence of breast carcinomas within a fi- 30. Wilkinson S, Forrest APN. Fibroadenoma of the breast. Br J Surg. broadenoma: a review. Arch Pathol Lab Med. 1984;108:590–3. 1985;72:835–8. 45. Oyyello L, Gump PE. The management of patients with carcino- 31. Jackson VP, Rothschild PA, Kreipke DL, Mal TJ. The spectrum of mas in fibroadenomatous tumors of the breast. Surg Gynecol Ob- sonographic finding of fibroadenoma of the breast. Invest Radiol. stet. 1985;160:99–104. 1986;21:34–9. 46. Diaz NM, Palmer JO, McDivitt RW. Carcinoma arising within fi- 32. Bottels K, Chan JS, Holly EA. Cytologic criteria for fibroadenoma: broadenoma of the breast: a clinicopathological study of 105 pa- a step wise logistic regression analysis. Am J Clin Pathol. 1988; tients. Am J Clin Pathol. 1991;95:614–22. 89:707–13. 47. Smallwood JA, Roberts A, Guyer DP, Taylor I. The natural history 33. Cant PJ, Madden MV, Close PM, Learmonth GM. Case for conser- of fibroadenoma. Br J Clin Pract. 1988;56(suppl):86–7. vative management of selected fibroadenoma of the breast. Br J 48. Sainsbury JRL, Nicholson S, Needham GK, Wadehra V. Natural Surg. 1987;74:857–9. history of benign breast lumps. Br J Surg. 1988;75:1080–2. 34. Cant PJ, Madden MV, Coleman MG, Dent DM. Nonoperative man- 49. Kerm WH, Clark RW. Retrogression of fibroadenoma of the breast. agement of breast mass diagnosed as fibroadenoma. Br J Surg. Am J Surg. 1973;126:59–62. 1995;82:792–4. 50. Naraynsingh V, Raju GC. Familial bilateral fibroadenoma of the 35. Dupont W, Parl FF, Hartman WH. Breast cancer associated with breast. Postgrad Med J. 1985;64:430–40. proliferative breast disease and atypical hyperplasia. Cancer. 1993; 51. Mies C, Rosen PP. Juvenile fibroadenoma with atypical epithe- 71:125–30. lium. Am J Surg Pathol. 1987;11:184–7. 36. Dupont WD, Page DL. Risk factors for breast cancer in women 52. Dixon JM, Dobie V, Lamb J, et al. Assessment of the acceptability with proliferative breast disease. N Engl J Med. 1985;312:146–51. of conservative management fibroadenoma of the breast. Br J 37. Hutchinson WB, Thomas DB, Hamlin WB, Roth GJ. Risk of breast Surg. 1996;83:264–5.