Localized Male Breast Carcinoma and Family HistoryAn Analysis of 142 Patients

Arnold Hill, M.Ch.1

Yusuf Yagmur, M.D.1

Katherine N. Tran, B.A.1

John S. Bolton, M.D.2

Mark Robson, M.D.3

Patrick I. Borgen, M.D.1

1 Breast Service, Department of Surgery, MemorialSloan-Kettering Cancer Center, New York, NewYork.

2 Breast Service, Department of Clinical Genetics,Memorial Sloan-Kettering Cancer Center, NewYork, New York.

3 Department of Surgery, Ochsner Clinic, New Or-leans, Louisiana.

Address for reprints: Patrick I. Borgen, M.D.,Breast Service, Department of Surgery, MemorialSloan-Kettering Cancer Center, 1275 York Avenue,New York, NY 10021.

Received June 15, 1998; revision received Febru-ary 12, 1999; accepted February 12, 1999.

BACKGROUND. Male breast carcinoma is rare; therefore, the effect of family history

on the course of the disease has not been well described. Germ-line mutations in

breast carcinoma susceptibility genes, particularly BRCA2, are associated with an

increased risk of male breast carcinoma. The authors sought to correlate significant

family history with clinical phenotype in males with breast carcinoma.

METHODS. One hundred forty-two men with breast carcinoma were treated at

Memorial Sloan-Kettering Cancer Center or the Ochsner Clinic from 1973 to 1994.

The authors reviewed the effect imparted by a family history of breast carcinoma

on the duration of symptoms, the age at diagnosis, and the survival of men with

this disease.

RESULTS. Fifteen percent of male breast carcinoma patients had a first-degree

relative with the disease. Fifty-eight years was the mean age at diagnosis for those

with a family history, compared with 61 years for those without (P 5 not significant

[NS]). The mean duration of symptoms was 23 months for those with a family

history, compared with 22 months for those without. Three of 22 patients (13.6%)

with a family history, compared with 11 of 90 patients (12%) without a family

history, had Stage III disease (P 5 NS) at presentation. The overall 5-year and

10-year survival rates were 86% and 64%. Survival was not affected by family

history. Lymph node positivity reduced 5-year and 10-year survival rates to 73%

and 50% (P 5 0.0004).

CONCLUSIONS. For men with breast carcinoma, the presence of a family history did

not affect the age at presentation, the duration of symptoms, the stage of disease

at presentation, or the overall survival. In multivariate analysis, the most powerful

predictor of outcome for these men was the status of the axillary lymph nodes. [See

editorial on pages 744 – 6, this issue.] Cancer 1999;86:821–5.

© 1999 American Cancer Society.

KEYWORDS: breast carcinoma, male, axillary lymph nodes, prognosis, survival.

Male breast carcinoma is rare, accounting for less than 1% of allbreast carcinomas.1 The mean age at diagnosis is approximately

65 years, although men of all ages can be affected by the disease.Predisposing risk factors appear to include radiation exposure, estro-gen administration, and diseases associated with hyperestrogenism,such as cirrhosis or Klinefelter syndrome.2–7 There is an increasedincidence of male breast carcinoma among men who have a numberof female relatives with breast carcinoma. An increased risk of malebreast carcinoma has been reported in families who carry a mutationin the breast carcinoma susceptibility gene BRCA2 on chromo-some 13q.8,9

The association between male breast carcinoma and mutations inBRCA2 prompted us to examine the clinical phenotype of male breastcarcinoma patients with a significant family history.

821

© 1999 American Cancer Society

MATERIALS AND METHODSData on 142 men with breast carcinoma treated from1973 to 1994 at Memorial Sloan-Kettering Cancer Cen-ter in New York, New York, or the Ochsner Clinic inNew Orleans, Louisiana, were analyzed in a retrospec-tive review. Patients with clinical evidence of distantmetastases at the time of presentation were excluded.Each patient’s medical record was reviewed retrospec-tively to determine the age at diagnosis, the presentingcomplaint, the duration of symptoms prior to diagno-sis, the family history, the clinical findings, the methodof diagnosis, the treatment given, and the subsequentcourse of the disease.

Follow-up was obtained by direct contact with thepatient or his family, and appropriate informed con-sent was also obtained. The cause of death was ascer-tained from the autopsy report or death certificate,when available, or by correspondence with the hospi-tal and physician caring for the patient at the time ofhis death.

StatisticsSurvival curves were estimated using the Kaplan–Meier method.10 Univariate survival analyses wereperformed with the log rank test. The Cox regressionmethod11 was used in the multivariate survival analy-ses to identify combinations of factors that had asignificant impact on survival. Statistical analysis wasperformed using the SPSS advanced statistics 7.5 soft-ware package (SPSS Inc., Chicago, IL).

RESULTSCharacteristics of All Male Patients with BreastCarcinomaThe median age at diagnosis of the men in this serieswas age 61 years (mean, 61 years; range, 25– 86 years).The median length of follow-up was 78 months (mean,87 months; range, 1–245 months). The median dura-tion of symptoms, which was available for 103 pa-tients, was 9 weeks (mean, 21 weeks; range, 1–312weeks). Clinically significant gynecomastia waspresent in 29 of 142 patients (20%). No patient hadKlinefelter syndrome or had undergone orchiectomy.Thirteen patients (9%) gave a history of trauma to theinvolved breast within 2 years of the diagnosis ofbreast carcinoma. Fourteen patients (10%) had a his-tory of a prior nonbreast malignancy. The diagnosis ofcarcinoma of the male breast was made by open bi-opsy in 129 patients (91%) and by cytologic examina-tion of fine-needle aspirate specimens alone in 8 pa-tients. The type of biopsy was not specified in 5patients.

The average number of axillary lymph nodes re-

moved was 21 (range, 1– 42). Fifty-seven patients hadpathologically positive lymph nodes. Negative Level Ilymph nodes with involved Level II nodes (skip me-tastasis) were seen in 8 patients (6%). Twelve patients(15%) had involved Level III axillary lymph nodes. Twopatients had positive Rotter lymph nodes. Histologicevidence of lymphatic invasion was present in 25 of108 patients (23%), blood vessel invasion in 18 of 107patients (17%), muscle invasion in 7 of 110 (6%), skininvasion in 9 of 111 (8%), subdermal lymphatic inva-sion in 8 of 107 (7%), and nipple invasion in 39 of 125(31%). The breast carcinoma was reported as multi-centric in 8 of 107 patients (7%).

Details concerning adjuvant therapy were avail-able for 131 patients. Thirty-one patients received cy-totoxic chemotherapy (30 received cyclophospha-mide, methotrexate, and 5-fluorouracil, and 1 receiveda doxorubicin-based regimen). Twenty-four patientsreceived tamoxifen. Seventy-six patients did not re-ceive either adjuvant chemotherapy or hormonal ther-apy. Adjuvant radiation therapy was administered to12 patients.

Eleven patients (8%) had local recurrence and 44patients (31%) suffered a distant relapse as the site oftheir first recurrence. The most frequent site of distantrelapse was bone, occurring in 38 (86%). The secondmost common site of distant failure was lung (11patients, 25%). Among the subset of 57 axillary lymphnode positive patients, 27 suffered a distant relapse(47%). Subsequent contralateral breast carcinomasoccurred in 5 of the 142 patients.

The actuarial 5- and 10-year disease free survivalrates for the entire group were 83% and 53%, respec-tively. The actuarial 5- and 10-year overall survivalrates were 86% and 64%, respectively. Disease freesurvival rates at 5 and 10 years for axillary lymph nodenegative patients were 98% and 73% and for lymphnode positive patients were 70% and 38% (P 5 0.0001).Overall survival rates for the same subsets at 5 and 10years for axillary lymph node negative patients were99% and 76% and for lymph node positive patientswere 73% and 50%, respectively (P 5 0.0003). In mul-tivariate analysis, the most powerful predictor of out-come for men with breast carcinoma was the status ofthe axillary lymph nodes (Table 1).

Characteristics of Men with a Family History of BreastCarcinomaDetails concerning a family history of breast carci-noma were available for 123 patients. Nineteen pa-tients (15%) had a first-degree relative with a history ofbreast carcinoma. Four had two first-degree relativeswith breast carcinoma, two patients had three first-degree relatives with the disease, and one patient had

822 CANCER September 1, 1999 / Volume 86 / Number 5

four first-degree relatives (a mother and three sisters)with the disease. Six patients had a history of a sec-ond-degree relative with breast carcinoma.

The mean duration of symptoms for the groupthat had a family history of breast carcinoma was 23months (range, 1–156 months), compared with 22months (range, 1–312 months) for those without afamily history of breast carcinoma. The mean age atdiagnosis of breast carcinoma was 58.7 years for thosewith a family history, compared with 61.8 years forthose without a family history (P 5 not significant[NS]).

Among those patients for whom the stage of dis-ease and family history were known, 3 of 22 (13.6%)with a family history compared with 12 of 90 (13%)without a family history of breast carcinoma had StageIII disease (P 5 NS) at presentation.

The most frequent histologic type of breast carci-noma in men with a family history was infiltratingductal carcinoma with ductal carcinoma in situ, whichwas found in 10 patients (42%) (Table 2). Eight men(36%) with a family history of breast carcinoma hadpathologically positive lymph nodes, compared with39 men (41%) without a family history.

Estrogen receptor status was available for 11 menwith a family history of breast carcinoma, 10 (91%) ofwhom were positive, compared with 35 men (77%)without a family history. Progesterone receptor statuswas available for 10 men with a family history; it waspositive in 8 patients (80%), compared with 28 men(72%) without a family history.

The actuarial 5- and 10-year relapse free survivalrates for the men with a family history were 79% and44%, respectively, compared with 84% and 63%, re-spectively, for those without a family history (log rankstatistic 5 1.21; degrees of freedom 5 1; P 50.271)(Fig. 1). The actuarial 5- and 10-year overall survivalrates were 79% and 45%, respectively, in those with a

family history compared with 86% and 70%, respec-tively, in those without (log rank statistic 5 1.95; de-grees of freedom 5 1; P value 50.162) (Fig. 2).

DISCUSSIONMale breast carcinoma presents at an older age thanits counterpart in females.12 In our series the medianage was 61 years (range, 25– 86 years). It is very rare inyoung men. Three men were younger than 30 years at

TABLE 1Predictors of Outcome for Male Patients with Breast Carcinoma

FactorP value(univariate)

P value(multivariate)a

Axillary lymph nodes 0.0004 0.0115Duration of symptoms (#6 mos, .6 mos) 0.0035 NSTumor size (Tis, T1–2, T3–4) 0.0254 NSFamily history NS NSAge at diagnosisb NS NSEstrogen receptor NS NSProgesterone receptor NS NS

NS: not significant. Significance was defined as P , 0.05.a Cox regression method.b Analyzed as a continuous variable.

TABLE 2Characteristics of Male Breast Carcinoma Patients with or without aFamily History

Characteristicsa

No. of patients (%)

Positivefamilyhistory

No familyhistory

Symptoms at presentation 25 98Mass 20 (80) 79 (81)Bloody nipple discharge 5 (20) 12 (12)Nipple retraction 7 (28) 13 (13)Skin ulceration 2 (8) 2 (2)

Duration of symptoms 19 69#6 mos 9 (47) 23 (33).6 mos 10 (53) 46 (66)

Operation performed 25 98Modified radical mastectomy 16 (64) 77 (78)Radical mastectomy 4 (16) 14 (14)Total mastectomy 4 (16) 6 (6)Extended radical mastectomy 0 (0) 1 (1)Wide local excision only 1 (4) 0 (0)

Histologic type 24 98Infiltrating ductal carcinoma 9 (37) 52 (53)Infiltrating ductal 1 ductal carcinoma in situ 10 (42) 28 (28)Ductal carcinoma in situ 5 (20) 16 (16)Medullary or tubular carcinoma 0 (0) 2 (2)

Tumor size 19 610–2 cm 15 (79) 49 (80)2–5 cm 4 (21) 12 (20)

Axillary lymph node status 22 94Negative 14 (64) 55 (59)All positive 8 (36) 39 (41)1–3 positive 6 (27) 27 (29)$4 positive 2 (9) 12 (12)

Stage 22 900 5 (23) 14 (16)I 6 (27) 30 (33)II 8 (36) 34 (37)III 3 (14) 12 (13)

Status at last follow-up 25 98No evidence of disease 11 (44) 46 (47)Alive with disease 1 (4) 7 (7)Dead of disease 12 (48) 24 (24)Dead of unknown causes 0 (0) 4 (4)Alive, status unknown 0 (0) 13 (13)Dead of other causes 1 (4) 4 (4)

a These were the characteristics for which data on patients were available.

Male Breast Carcinoma and Family History/Hill et al. 823

diagnosis in our study. The presence of a family his-tory of breast carcinoma did not significantly changethe mean age of presentation in our series, which was58 years for those with a family history compared with61 years for those without a family history of breastcarcinoma.

The mean delay between the onset of symptomsand the commencement of treatment was 21 weeks inour series. There was no significant difference in themean duration of symptoms with (23 months) or with-out (22 months) a family history of breast carcinoma.Delay in diagnosis did have an impact on survival. The80 patients with a delay of 6 months or less had 5- and10-year survival rates of 90% and 70%, respectively.The 23 patients whose symptoms were present forlonger than 6 months had 5- and 10-year survival ratesof 71% and 56%, respectively (P50.035) (Table 1).

Male breast carcinoma patients typically pre-sent with a more advanced stage of disease than dowomen.13,14 However, breast carcinoma in men is notbiologically more aggressive than in women.15 In thisseries, presentation with Stage III disease was no morecommon in those with a family history of breast car-cinoma (13.6%) compared with those without a familyhistory of breast carcinoma (12%). Much evidence hasbeen provided in the literature to suggest that malebreast carcinoma is similar to its female counterpart,including data from our institution noting a similarincidence of p53 mutations in both male and femalebreast carcinoma.16

Axillary lymph node involvement remains themost important prognostic factor for male breast car-cinoma (Table 1). In both univariate and multivariateanalyses, positive axillary lymph nodes significantly

affected survival in male breast carcinoma patients.Tumor size was significant only in univariate analysis.The presence of a family history of breast carcinomadid not significantly affect survival.

Historically, most reports indicate that men withbreast carcinoma have a poorer prognosis.12,17,18

However, the authors have previously reported thesurvival rates of 58 male breast carcinoma patientscompared with 174 female breast carcinoma patientswho were matched by disease stage and age at diag-nosis.15 The survival rates at 10 years were similar formale and female patients. The authors concluded thatthe prognosis of male breast carcinoma patients wascomparable to that of female patients of the same ageand stage of disease at diagnosis. The 5- and 10-yearoverall survival rates of 86% and 64% and the diseasefree survival rates of 83% and 56% presented in thisstudy support the belief that the prognosis for maleand female breast carcinoma patients is similar stagefor stage. We found that the presence of a familyhistory of breast carcinoma did not affect the survivalof male breast carcinoma patients.

In 1994, the BRCA2 gene, which predisposes het-erozygotes to breast carcinoma, was mapped to chro-mosome 13q12–13.19 In 1995, the BRCA2 gene wascloned.8 BRCA2 was located by performing a genomiclinkage search using 15 families with multiple cases ofearly-onset breast carcinoma that were not linked toBRCA1.20 Like BRCA1, BRCA2 appears to be a tumorsuppressor gene.21

Rare instances of families have been reportedin which several men have developed breast carci-noma.12,22–25 A higher incidence of breast carcinomahas been noted in Jewish males.26,27 This observation

FIGURE 2. This Kaplan–Meier plot shows overall survival after treatment of

carcinoma of the male breast for those with and without a family history (P 5

0.16).

FIGURE 1. This Kaplan–Meier plot shows disease free survival after treat-

ment of carcinoma of the male breast for those with and without a family

history (P 5 0.27).

824 CANCER September 1, 1999 / Volume 86 / Number 5

may be related to the frequency of founder mutationsin the latter group. The presence of a BRCA2 mutationis thought to be a factor in the development of certaincases of male breast carcinoma.20,28 Two recent stud-ies have examined the proportion of male breast car-cinoma cases that may be related to mutations in theBRCA2 gene. Friedman et al. examined 54 cases ofmale breast carcinoma from the U.S. and found that4% had mutations in the BRCA2 gene.29 Haraldsson etal. examined 34 patients from Sweden and found 5different truncating BRCA2 mutations in 7 (21%) of the34 cases.30 This suggests that BRCA2 mutation as acausative factor for male breast carcinoma varies con-siderably in different population groups.

In conclusion, 15% of male breast carcinoma pa-tients had a first-degree relative with breast carci-noma. The presence of a family history did not affectthe age at presentation, the duration of symptoms, thestage of disease at presentation, or the overall survival.This suggests that hereditary male breast carcinomascarry a prognosis similar to that of their sporadiccounterparts. Such a suggestion would imply that pre-treatment BRCA2 testing in males with breast carci-noma is not warranted, as it would not change thera-peutic approaches, and that treatment should not bechanged based on family history.

REFERENCES1. Parker SL, Tong T, Bolden S, Wingo PA. Cancer statistics. CA

Cancer J Clin 1997;47:5–27.2. Erlichman C, Murphy KC, Elhakim T. Male breast cancer: a

13-year review of 89 patients. J Clin Oncol 1984;8:903–9.3. Jaiyesimi IA, Budzar AU, Sahin AA, Ross MA. Carcinoma of

the male breast. Ann Intern Med 1992;117:771–7.4. Jackson AW, Muldal S, Ockey CH, O’Connor PJ. Carcinoma

of the male breast in association with the Klinefelter syn-drome. Br Med J 1965;1:223–5.

5. Lynch HT, Kaplan AR, Lynch JF. Klinefelter syndrome andcancer: a family study. JAMA 1974;229:809 –11.

6. Giammarini A, Rocchi M, Zennaro W, Filippi G. XX malewith breast cancer. Clin Genet 1980;18:103– 8.

7. Decker JP, Lerner HJ, Schwartz I. Breast carcinoma in a 46XXtrue hermaphrodite. Cancer 1982;49:1481– 4.

8. Wooster R, Bignell G, Lancaster J, Swift S, Seal S, Mangion J,et al. Identification of the breast cancer susceptibility geneBRCA2. Nature 1995;378:789 –92.

9. Thorlacius S, Tryggvadottir L, Olafsdottir GH, Jonasson JG,Ogmundsdottir HM, Tulinius H, et al. Linkage to BRCA2region in hereditary male breast cancer. Lancet 1995;346:544 –5.

10. Kaplan EL, Meier P. Nonparametric estimation from incom-plete observations. J Am Stat Assoc 1958;53:457– 81.

11. Cox DR. Regression models and life tables. J R Stat Soc B1972;34:187–202.

12. Crichlow RW. Carcinoma of the male breast. Surg GynecolObstet 1972;134:1001–19.

13. Siddiqui T, Weiner R, Moreb J, Marsh PD. Cancer of themale breast with prolonged survival. Cancer 1988;62:1632– 6.

14. Lefor AT, Numann PJ. Carcinoma of the breast in men. N YState J Med 1988;88:293– 6.

15. Borgen PI, Senie RT, McKinnon WMP, Rosen PPR. Carci-noma of the male breast: analysis of prognosis comparedwith matched female patients. Ann Surg Oncol 1997;4:385– 8.

16. Anelli A, Anelli TF, Youngson B, Rosen PP, Borgen PI. Mu-tations in the p53 gene in male breast cancer. Cancer 1995;75:2233– 8.

17. Hudson MJR, Smart CJ. Carcinoma of the male breast. Br JSurg 1974;61:440 – 4.

18. Scheike O. Male breast cancer. 6. Factors influencing prog-nosis. Br J Cancer 1974;30:261–71.

19. Wooster R, Neuhausen S, Mangion J, Quirk Y, Ford D, Col-lins N, et al. Localization of a breast cancer susceptibilitygene, BRCA2, to chromosome 13q12– q13. Science 1994;265:2088 –90.

20. Easton DF, Bishop DT, Ford D, Crockford GP. Breast cancerlinkage consortium genetic linkage analysis in familialbreast and ovarian cancer: results from 214 families. Am JHum Genet 1993;52:678 –701.

21. Davies K. Breast cancer genes: further enigmatic variations.Nature 1995;378:762–3.

22. Everson RB, Li FP, Fraumeni JF Jr., Fishman J, Wilson RE,Stout D, et al. Familial male breast cancer. Lancet 1976;7949:9 –12.

23. Marger D, Urdaneta N, Fisher JJ. Breast cancer in brothers:case reports and a review of 30 cases of male breast cancer.Cancer 1975;36:458 – 61.

24. Teasdale C, Forbes JF, Baum M. Familial breast cancer.Lancet 1976;7955:360 –1.

25. LaRaja RD, Pagnozzi JA, Rotherberg RE, Georgiou J, SabatiniMT, Hirschman RJ. Carcinoma of the breast in three sib-lings. Cancer 1985;55:2709 –11.

26. Lin RS, Kessler II. Epidemiologic findings in male breastcancer. Proc Am Assoc Cancer Res 1980;21:72.

27. Mabuchi K, Bross DS, Kessler II. Risk factors for male breastcancer. J Natl Cancer Inst 1985;74:371–5.

28. Narod SA, Ford D, Devilee P, Barkadottir RB, Lynch HT,Smith SA, et al. An evaluation of genetic heterogeneity in 145breast ovarian cancer families. Breast cancer linkage con-sortium. Am J Hum Genet 1995;56:254 – 64.

29. Friedman LS, Gayther SA, Kurosaki T, Gordon D, Noble B,Casey G, et al. Mutation analysis of BRCA1 and BRCA2 in amale breast cancer population. Am J Hum Genet 1997;60:313–9.

30. Haraldsson K, Loman L, Zhang Q, Johannsson O, Olsonn H,Borg A, et al. BRCA2 germline mutations are frequent inmale breast cancer patients without a family history of thedisease. Cancer Res 1998;58:1367–71.

Male Breast Carcinoma and Family History/Hill et al. 825