the prognosis of breast cancer in males

7/30/2019 The Prognosis of Breast Cancer in Males

1/8

154

The Prognosis of Breast Cancerin MalesA Report of 335 CasesVinc e n f F . Guiizee, M.D ., M.P.H.,' Ha kan Ols son , M.D., Ph.D. ,tTorgil Moller, M.D ., Plz.D.,t Rick C. Shallenberger, M.P.H.,*I . W . van de n Blink, M.D.,S Zoltarz Peter, M.D.,fj Miche l Durarzd, M.D.,IIStanley Dische, M.D.,V Franz 1.Cleto n, M.D.,# Roelof Zewuster,** Me i Fang C u i, ttWarren Lane, Ph.D.,$$ and Rudolf Richter, M.D.55

Background. Because of the relative rarity of breastcancer in males, data have not been sufficient to supporta definitive analysis of pertinent prognostic factors. Re-markably, no studies of male patients with breast cancerhave presented survival information based on the num-ber of histologically positive axill ary nodes, the most sen-sitive single indicator of prognosis in women with breastcancer.

Methods. In th is study , the clinical course of breastcancer was documented for 335 male patients registeredfrom 1965 through 1986. For patients to be eligible, diag-noses had to be made within 3 months of registrat ion andthe patients had to have histologic confirmation and re-ceive par t or all of their initial tr eatment at 1 of 11cancercenters participatingin the Inte rnat iona l Patient Data Ex-change System.

Results. The survival rate at 10 years was 84% forpatients with histologically negative nodes, 44% for thosewith one to three positive nodes, and 14% for the group

From the *Department of Patient Studies, University of TexasM. D. Anderson Cancer Center, Houston, Texas; tLund UniversityHospital, Lund, Sweden; $Dr. Daniel den Hoed Cancer Center, Rot-terdam, The Netherlands; 5National Institute of Oncology, Budapest,Hungary; ((FoundationBergonie, Bordeaux, France; TRegional Centerfor Radiotherapy and Oncology, Mt. Vernon Hospital, Northwood,Middlesex, England; #University Hospital, Leiden, The Netherlands;**Netherlands Cancer Institute, Amsterdam, The Netherlands;ItCancer Institute, Chinese Academy of Sciences, Beijing, China;$$Roswell Park Cancer Institute, Buffalo, New York; and West-deutsches Tumorzentrum, Essen, Germany.The authors thank Harold Andersson, Ph.D., Joe Ensor, Ph.D.,and Kenneth Hess, Ph.D., fo r their statistical contributions duringthis study.The International Cancer Patient Data Exchange System is aproject of the Committee on International Collaborative Activities ofthe International Union Against Cancer.Address for reprints: Vincent F. Guinee, M.D., M.P.H., Depart-ment of Patient Studies, University of Texas M. D. Anderson CancerCenter, 1515 Holcombe Boulevard, Box 214, Houston, TX 77030.Accepted for publication May 27, 1992.

with four or more histologically positive nodes. Th e sur-vival rates at 5 years were go%, 73'/0, and 55%, respec-tively. In a multivariable analysis, the risk of death dueto breast cancer for a patient with four or more histologi-cally positive nodes was 6.75 times that of a patient withnegative nodes.

Conclusions. Th e findings of th e aut hor s indicate thefollowing: (1)Th e number of histologically positive axil -lary nodes and, to a lesser degree, tumor diameter aresignificant prognostic factors for br east cancer in malepatients. (2) The prognosis of breast cancer is the same inmale and female patients when compared on the basis ofthe n umber of histologically positive nodes. Cancer 1993;71 :154-61.Key words: b reast cancer male, prognosis, surv ival , axil-lary nodes, multivariable analysis.

Because of the relative rarity of breast cancer in males,the literature has been dominated by reviews, collectedcase reports, and original series with small numbers ofcases. Data available to these authors were not suffi-cient to support a definitive analysis of the prognosis ofmale patients with breast cancer.By the early 1970s, a number of observers believedthat survival of male an d female patients with breastcancer was ~ompa rab le.' -~owever, 10years later6 and20 years later7 investigators still reported that their datashowed a poorer prognosis for male patients. At best,current texts and reviewss-" state that , although maleand female patients with negative axillary nodes mayhave equivalent survival rates, male patients with histo-logically positive axillary nodes have a worse prognosisthan women with positive nodes.In contrast to the detailed pathologic data availableon female patients with breast cancer, prognostic evalu-

ation of men with breast cancer usually has been basedon clinical data.3,"-'4 Only a few case series have re-


2/8

Breast Can cer in M a l e s l Gu in e e e t a l . 1 55ported the histologic status of axillary node^.^,'^-'^ Re-markably, no studies of men with breast cancer havepresented survival information based on the number ofhistologically positive axillary nodes. And yet, the num-ber of histologically involved axillary nodes is consid-ered to be the most sensitive single indicator of progno-sis in women. 21 This correlation was reported byFisher et a1.22n 1968 and confirmed in subsequent stud-i e s , 2 3 2 4 As long as information on the survival of menwith breast cancer did not include the number of histo-logically positive nodes, valid comparisons could not bemade with survival of women with breast cancer.This investigation was performed to produce amore reliable assessment of the prognosis of breastcancer in men. We analyzed clinical and pathologicdata with particular attention to axillary nodes and tu-mor diameter.Patients and MethodsCancer centers participating in the International PatientData Exchange System collect a common clinical dataset on patients with cancer. Representatives from eachcenter meet annually to maintain uniform data collec-tion and quality control and to plan new studies. Theirdata bases were searched for male patients with breastcancer registered from January 1, 1965, through De-cember 31, 1986. For patients to be eligible for thestudy, diagnoses had to be made within 3 months ofregistration and the patients had to have histologicallyconfirmed carcinoma of the breast an d receive part orall of their initial treatment a t one of the participatinginstitutions. There were 380 patients who met these cri-teria: Rotterdam, 69; Mt. Vernon, 55; Lund, 50; Hous-ton, 49; Budapest, 43; Bordeaux, 39 ; Amsterdam, 17;Beijing, 17; Leiden, 15; Essen, 14; and Buffalo, 12 .

It should be mentioned that the number of patientsfrom each participating center does not necessarily re-flect its total case load of breast cancer cases in men. Atsome of the participating centers, most of the patientpopulation will have completed initial cancer treatmentbefore registration. Men with breast cancer who hadprior treatment were not included in this study to avoidthe biases introduced by analyzing patients who mighthave failed to respond to treatment at other hospitals.At each center, charts were reviewed during No-vember and December 1988, with the use of a codesheet and corresponding coding manual. The cause ofdeath was based on statements written in the medicalrecord of the patient by the treating physician. If adeath was not specified as resulting from breast canceror another specified cause i t was recorded that the causewas unknown. The study was coordinated by theHouston data center.Not all data items were available for each patient.

As is common clinical practice, when a more serious orsignificant physical sign is recorded, a sign consideredless serious often is unrecorded. Also, in instances inwhich patient status was reported by TNM classifica-tion, actual measurements might not be available. Eachgraph and table presented includes all patients withknown information on the assessed parameters.Analy t ic M ethodsOn univariable analysis, age, skin ulceration, clinicaland histologic tumor diameter, and axillary node statuswere found to be significant prognostic factors. Otherfactors considered, but not found to be significant, wereas follows: perinodal extension of tumor, skin and chestwall fixation, and histologic type.Survival analysis was performed with the Berkson-Gage method, as implemented in the Statistical Pack-age for the Social Sciences X, version 9 (Evanston, IL).Differences between survival curves were tested by theLee-Desu statistic.25Simultaneous assessment of multi-ple variables was pursued with the Cox proportionalhazards regression.26 The proportional hazards as-sumption was assessed by stratification on each covar-iate in turn while including the other variables as fac-tors in the Cox model. The log(-log) plots of the ad -justed baseline survival function estimates for eachstratum were roughly parallel. The linearity assumptionfor interval-scaled covariates was verified with the useof cubic spline functions.27W e used the implementa-tion of the Cox model as coded by the BMDP StatisticalSoftware Package Program 2L (1990, Los Angeles, CA).Multivariable analysis was performed on factorsthat were significant on univariable analysis. Interac-tion terms between pairs of these variables were notsignificant. These factors were entered into three Coxproportional hazards models with clinical data and dif-ferent end points for cause of death. A fourth modelwas used to calculate the relative risks based on patho-logic data. Tumor size was coded as a continuous vari-able and is presented in the models as the effect of a30-mm difference in diameter. The age of 65 years wasthe point that reflected the greatest difference in sur-vival; therefore, age was dichotomized at 65 years. Thenumber of positive nodes was grouped into three sets:none, one to three, an d four or more. All other variableswere analyzed as binary variables.ResultsPatient CharacteristicsEleven patients with a diagnosis of in situ carcinoma ofthe breast and 26 patients with distant metastases wereexcluded from the analysis of prognostic factors. Eight


3/8

156 CANCER fanuary 2, 2993, Volume 71, No. 1others were excluded because pertinent clinical datawere not recorded. The median age of the 335 patientsincluded in the analysis was 63.0 years, with a range of23-97 years (mean, 62.8 years). The most frequent his-tologic designations were adenocarcinoma (44.2%),ductal carcinoma (33.7%), and epithelial carcinoma,not otherwise specified (16.4%).

The median follow-up of the entire study cohort(335 patients) was 52 months, with a mean of 65.4months. There were 95 deaths resulting from breastcancer. Disease-specific survival analyses were basedon deaths resulting from breast cancer, leading us tocensor the survival time of 83 patients because of"death due to other causes." The patients in whomcancer was diagnosed before 1984 and who were stillalive at last follow-up had a median follow-up of 89months (mean, 99 months). Of the 42 living patientswith recent diagnoses (1984-1986), 39 had current fol-low-up at the time of the study.

There were 2 24 patients with reports of both clini-cal and histologic axillary findings. The number ofnodes examined histologically ranged from 1 to 39,with a median of 8. Patients with negative nodes onhistologic examination had a median of 9 nodes exam-ined; patients with 1-3 positive nodes had a median of6 examined; and patients with 4 or more positive lymphnodes had a median of 13examined. Patients with nega-tive clinical axillary findings had a median of ninenodes examined histologically. Patients with positiveclinical axillary findings had a median of eight nodesexamined.

Three hundred eight of the 335 patients had sur-gery. Most patients (220) had a mastectomy with nodedissection, 58 had mastectomy without node dissection,and 30 had a local excision. Most of these patients alsohad radiation therapy (245 of 308). Twenty-two of the27 men who did not have surgery received radiationtherapy. Three of the remaining five patients had che-motherapy or antiestrogen therapy alone or in combina-tion. Two patients had no treatment after biopsy. Che-motherapy was used for 38 patients, and antiestrogentherapy was used in 25 .Chest W all and Skin lnvolvementThe patients who had tumor fixed to the chest wall (40)or skin (73) had essentially the same survival curves(Fig. 1) as those in whom fixation was absent ( P =0.86and P = 0.19, respectively). If ulceration was noted,there was an immediate decrease in survival (P


4/8

Breast Cancer in MaleslGuinee et al . 157100

80

L5 4 02t t2o t 11

0 2 4 5 6 8 10Years From Dlagnosls

Figure 1. Disease-specific survival rate in 33 5 men with breastcancer according to skin involvement and chest wall fixation. -:N o fixation (n = 178); - - : skin fixation (n = 73); - - - -: chestfixation (n = 40); -- - -: skin ulceration (n = 44) .

nodes. Eight of these had clinically positive axillae.Within 48 months, seven died of breast cancer and oneof another cause. One patient, late on study, was aliveon last contact at 26 months.Five-Year Survival Rate by Combined Axillary NodeStatusThe 94 patients with histologically negative nodes hada 90% 5-year survival rate (Table 2). The 5-year sur-vival rate in those with one to three histologically posi-tive nodes was 7383, and in those with four or morepositive nodes it was only 5 5 % .At each level of histo-logic node involvem,ent, patients with a clinically nega-tive axillary status had a significantly better survivalrate than patients who had clinically positive findings.Among the patients with four or more histologicallypositive nodes, those who had clinically negative axillaehad a 5-year survival rate of 78%, in contrast to the

45% survival rate of those with clinically positiveaxillae.Comparabi l i ty of Patie nts Among CentersTo assess possible variation in the study populationamong institutions, data from the five centers withfewer than 20 patients were compared with the datafrom the six centers with more than 35 patients. Thetwo groups of patients had essentially the same presen-tation of clinical and pathologic node status and patho-logic tumor size. For example, 40% of patients fromcenters with fewer than 20 study subjects had palpablenodes versus 45% of patients from centers with morethan 35 study subjects. Similarly, the percentages ofpatients with histologically negative nodes were 23%and 28%, respectively. The mean pathologic tumor di-ameter was 24 mm for each group. The two groups hadoverlapping survival curves, based on all deaths ( P=0.36) and breast cancer-specific deaths ( P = 0.33).Multiva riable AnaZysis of Prognostic FactorsA series of Cox proportional hazards models was usedto delineate the relative contribution of factors that arepredictive of risk of death (Fig. 4).Both clinical axillarynode status and clinical tumor diameter were predictiveof death due to breast cancer (model 1).A patient withpalpable axillary nodes had 2.14 times the risk of deathdue to breast cancer as a patient with a clinically nega-tive axilla. An increase in tumor diameter of 30 mmcarried a similar increase in risk (2.08).

Eleven deaths were classified as resulting from anunknown cause. In the second model, these deathswere included as though they resulted from breastcancer. The relative risks remained approximately thesame. In these models, which included information onthe clinical axillary node status and clinical tumor diam-eter, skin ulceration did not represent significant addi-tional risk for death due to breast cancer.

Table 1. Median Age, Skin Ulceration, Clinical Axillary Status, and 5-Year Disease-SpecificSurvival Rate According to Clinical Tumor DiameterClinical Median Skin Positive 5-yeardiameter (mm) No. age (yr) ulceration nodes survival rate

0-10 23 55 4% 30 % 94%11-201 87 62 5% 39% 80%21-30 55 65 15% 53% 78%31-40 38 62 13% 45% 8 101041-50 26 66 23% 65% 41%5 1 + 28 66 43% 71yo 39 %

Unknown 78 61 11% 34% 80%Total 335 63 13% 45 % 76%


5/8

158 CANCER January 2, 2993, Volume 71, No. 1

*O00 2 4 5 6 8 10

Years From DiagnosisFigure 2. Disease-specific survival rate in 335 men with breastcancer according to clinical axillary status. -: Clinically negative (n= 185);- - -: clinically positive ( n = 150) (Pi .001).

An analysis also was performed on the effect of theprognostic factors of breast cancer on all causes ofdeath, including breast cancer. In this model, nodes andtumor size remained significant. In addition, age ap-peared to be a significant prognostic factor.

Finally, in a fourth model, the risk of death due tobreast cancer was measured for pathologic tumor sizeand number of histologically positive nodes. The riskfor a patient with four or more positive nodes was 6.75times that of a patient with histologically negativenodes. Similarly, a patient with one to three positivenodes had more than three times the risk of death. His-tologic nodal status carried more risk than pathologic

100

80

2o t0 1 1

0 2 4 5 6 8 10Years From Dlagnosls

Figure 3. Disease-specific survival rate in 224 men with breastcancer according to the number of histologically positive axillarylymph nodes (P


6/8

Breast Cancer in MiileslGuinee et a l . 159

Figure 4 . Relative risk of death formale patients with breast canceraccording to the Cox proportionalhazards analysis of four alternativemodels. Bold boxes in


7/8

160 CANCER January 2 , 2993, Volume 71, No. 1viva1 rates of our patients were 90% and 65% respec-tively. Our 84% 10-year survival rate for patients withhistologically negative nodes is essentially the sameas found by Hultborn et al.15 (81%) and Heller etal." (79%).

Where our study differs is the further definition ofhistologically positive axillary node status into catego-ries of one to three and four or more positive nodes. Ourcomposite 5-year survival rate of 65% for those withpositive nodes resulted from the combined experienceof those with one to three positive nodes (a 73% sur-vival rate) and four or more positive nodes (a 55% sur-vival rate). The 10-year survival rate of male patientswith histologically positive nodes has been reported inthe range of 11% o 2 5 y 0 . ' ~ ~ ' ~n our study, patients withfour or more positive nodes had only a 14% 10-yearsurvival rate, but those with one to three positive nodeshad a 44% 10-year survival rate. The dominant role ofhistologic node status in determining prognosis clearlywas supported in the multivariable analysis.N o reason has been given in other studies why his-tologically positive nodes were not subdivided in calcu-lating survival. In most instances, the small numbers ofpatients studied probably would not have supportedadditional subdivision. Nevertheless, the fact that posi-tive pathologic node findings were not subdivided cer-tainly contributed to interstudy variations.

Comparison with survival rates reported for femalepatients with breast cancer must allow for variation intime period, age, and study eligibility criteria. In addi-tion, survival calculations have been reported as abso-!Ute, relative, disease-specific, or not specified. Withconsideration of these factors, the survival rates at 5and 10 years of the men with breast cancer in this studywere comparable to those of the larger published seriesof female patients with breast ~ a n c e r . ~ ~ - ~ *or example,among premenopausal and postmenopausal women,Fisher et aL40 reported 5-year survival rates of 62% and32% after radical mastectomy for patients with one tothree and four or more positive nodes, respectively. At10 years, the survival rates decreased to 37% and 13%,respectively, in the same groups.

Comparison of Clinical an d Pathologic Node Sta tusThe survival advantage at 5 years for patients with clin-ically negative axillae was statistically significant for pa-tients with one to three and four or more histologicallypositive nodes. The presence of histologically positivenodes with clinically negative axillae probably reflects alower nodal tumor burden. Indeed, of the patients withone to three histologically positive nodes, approxi-

mately half (35 of 74) had clinically negative axillae,whereas, of the patients with four or more histologi-cally positive nodes, only one-fourth had clinically neg-ative axillary findings (1 6 of 56) .ConclusionsIn male patients with breast cancer, the prognostic im-portance of the number of histologically positive nodeswas established. It is seen in the comparison of survivalcurves and supported by proportional hazards analysis.In multivariable analysis models, tumor size retainedprognostic significance in the presence of either clinicalor histologic axillary node findings. Finally, survivalrates for breast cancer are the same for men and womenwhen the number of histologically positive axillarynodes is considered.References

1 . Holleb AI, Freeman H',Farrow JH. Cancer of male breast. NYState ]Me d 1968; 68:544-53.

2 . Classen JN , Montague ACW, Wilgis 5. Cancer in the malebreast. Arch Surg 1970; 100:66-7.3 . Scheike 0.Male breast cancer: factors influencing prognosis. BrJ Cancer 1974; 30:261-71.4 . Donegan WL, Perez-Mesa CM. Carcinoma of the male breast.

Arch Surg 1973; 106:273-9.5 . Mausner IS, Shimkin MB, Moss N H , Rosemond GP. Cancer ofthe breast in Philadelphia hospitals 1951-1964. Cancer 1969;6 . Satiani MB, Powell RW. Cancer of the male breast: a thirty-yearexperience. Am Surg 1978; 44236-93.7 . Ciatto 5, lossa A, Bonardi R, Pacini P. Male breast carcinoma:review of a multicenter series of 150 cases. Tumori 1990;8. Ouriel K. Carcinoma of the male breast. NY State ]Me d 1988;9 . Henderson IC, Harris JR, Kinne DW, Hellman S. Cancer of thebreast. In: DeVita VT, Hellman S, Rosenberg SA , editors.Cancer: principles and practice of oncology. 3rd ed. Philadel-phia: JB Lippincott, 1989:1245.

10. Crichlow RW, Galt SW . Male breast cancer. Su r g Clin North Am11. Spence RAJ, Mackenzie G , Anderson JR, Lyons AR, Bell M.Long-term survival following cancer of the male breast inNorthern Ireland: a report of 81 cases. Cancer 1985; 55:648-52.12 . Ribeiro G . Male breast carcinoma: a review of 301 cases fromChristie Hospital and Holt Radium Institute, Manchester. Br ]

Cancer 1985; 51:115-9.13. van Gee1 AN, van Slooten EA, Mavrunac M, Hart AAM.A retro-spective study of male breast cancer in Holland. Br] u rg 1985;14 . Ramantanis G, Besbeas 5, Garas JG. Breast cancer in the male: areport of 13 8 cases. World J Surg 1980; 4:621-4.15. Hultborn R, Friberg S, Hultborn KA, Peterson LE, Ragnhult I.Male breast carcinoma: 11.A study of the total material reportedto the Swedish Cancer Registry 1958-1967 with respect to treat-ment, prognostic factors and survival. Acta Oncol 1987; 26~ 327 -

41 .

23:260-74.

761555-8.881291-2.

1990; 70~1165-77.

72:724-7.


8/8

Breast Cancer in Males/Guinee ef a l . 16116

17181920

2122

23

24

25

2627282930

Heller KS, Rosen PP, Schottenfeld D, Ashikari R, Kinne DW.Male breast cancer: a clinicopathologic study of 97 cases. A n nSurg 1978; 188:60-5.Yap HY, Tashima CK, Blumenschein GR, Eckles NE. Malebreast cancer: a natural history study. Cancer 1979; 44:748-54.Erlichman C,.Murplhy KC, Elhakim T. Male breast cancer: a 1 3-year review of 89 patients. 1Clin Oncol 1984; 2903-9.Donegan Wt.. Cancer of the breast in men . C A 1991; 41:339-54.Harris JR , Hc4lman ' 5. atural history of breast cancer. In: HarrisJR, Hellman S, Heinderson IC, Kinne DW, editors. Breast dis-eases. 2nd ed. Philadelphia: JB Lippincott, 1991:171.Kinne DW. The surgical m anagemen t of primary breast cancer.Fisher B, Ravdin RG, Ausm an RK, Slack NH, Moore GE, RudolfJN . Surgical adjuvant chemo therapy in cancer of the breast: re-sults of a decade of cooperative investigation. Ann Surg 1968;Fisher B, Bauer M, Wickerham L, Redm ond CK, Fisher ER. Rela-tion of number of positive axillary nodes to the prognosis ofpatients with primary breast cancer: an NSABP update. CancerFisher ER , Palekar A, Rockette H, Redm ond C, Fisher B. Patho -logic findings from the National Surgical Ad juvan t Breast Proj-ect (protocol no. 4): V. Significance of axillary nodal micro- an dmacrometastases. Cancer 1978; 422032-8 .Lee E, Desu M. A computer program for comparing k sampleswith right-censored data. Comput Programs Biomed 1972; 231 5-21 .Cox D. Regression models an d life tables. 1R S t a t Soc [ B ] 1972;34: 187-202.Durrleman S, Sim.on R. Flexible regression mod els with cubicsplines. Sfat Med 1989; 8:551-61.Scheike 0.Male breast cancer. Acta Pathol Microbiol Scand 1975;Somerville P. Carcinoma of the m ale breast: a report of 1 9 casesand a review of the literature. Br 1Surg 1952; 39296-30 3.Crichlow RW. Carcinoma of the male breast. Surg Gynecol Ob-s t e t 1972; 1.34:1011-9.

CA 1991; 41~71-84 .

168~337-56.

1983; 52~155 1-7.

25 l(Suppl).l1-35.

31.32 .33 .34 .35 .

36 .

37 .

38 .39 .

40 .

41 .

42 .

Carlsson G, Hafstrom L, Jonsson P-E. Male breast cancer. ClinOncol 1981; 7:149-55.Norris HJ, Taylor HB. Carcinoma of the male breast. CancerOuriel K, Lotze MT, Hinsh aw JR. Prognostic factors of carci-noma of the male breast. Surg Gynecol Obstet 1984; 159:373-6.Carter CL, Allen C, Henson DE. Relation of tumor size, lymphnode status, and survival in 24,740 breast cancer cases. CancerHultborn R, Friberg S, Hultborn KA. Male breast carcinoma: astudy of the total material reported to the Swedish Cancer Re-gistry 1958-1967 with respect to clinical and histopathologicparameters. Acta Oncol 1987; 26241-56 .Fisher B, Wolmark N, Bauer M, Redmond C, Gebhardt M. Theaccuracy of clinical nodal staging an d of limited axillary dissec-tion as a determ inant of histologic nodal sta tus in carcinoma ofthe breast. Surg Gynecol Obstet 1981; 152765-72.Valagussa P, Bonadonna G , Veronesi U. attern s of relapse andsurvival following radical m astectomy: analysis of 716 consecu-tive patients. Cancer 1978; 41:1170-8.Haagensen CD . Trea tme nt of curable carcinoma of the breast.ln t J Radiat Oncol Biol Phys 1977; 2975-80.Ferguson DJ, Meier P, Kamson T, Dawson PJ, Straus FH, Lo-wenstein FE. Staging of breast cancer an d survival rates: an as-sessment based on 50 years of experience with radical mastec-tomy. ] A M 1982; 248:1337-41.Fisher B, Slack N, Katrych D, Wolmark N. Ten year follow-upresults of patien ts with carcinoma of the breast in a co-operativeclinical trial evaluating surgical adjuvant chemotherapy. SurgGynecol Obstet 1975; 140:528-34.Nemoto T, Vana J, Bedwani R, Baker HW, McGregor FH,Murphy GP . Management an d survival of female breast cancer:results of a national survey by the American College of Sur-geons. Cancer 1980; 459917-2 4.Early Breast Cancer Trialists' Collaborative Group. Systemictreatment of early breast cancer by hormonal, cytotoxic, or im-mu ne therapy: 1 33 rand om ised trials involving 31,000 recur-rences and 24,000 deaths among 75,000 women. Lancet 1992;

1969; 2311428-35.

1989; 63:181-7.

339:71-85.

the prognosis of breast cancer in males

Documents