Clinical epidemiology of bilateral breast cancer

3629

Clinical Epidemiology of Bilateral Breast Cancer Hermann Brenner, M.D., M.P.H.,* Barbara Engelsmann,* Christa Stegmaier, Dipl.Med.lnf.,t and Hartwig Ziegler, Dipl.Volkswirtt

Background. The clinical epidemiology of bilateral breast cancer was assessed in a population-based study from Saarland, Germany.

Methods. Clinical aspects of breast cancer, including age and tumor spread at diagnosis and survival, were compared between 251 patients with bilateral breast cancer and all 9585 patients with a first diagnosis of breast cancer registered with the population-based cancer registry of Saarland, Germany, between 1968 to 1987.

Results. The patients with bilateral breast cancer were on average considerably younger at the time of diagnosis of the first breast cancer than the total group of patients with breast cancer (mean age, 55.7 years versus 60.5 years). Overall, the distribution of tumor spread at diagnosis of the second breast cancers was similar to the corresponding distribution of first breast cancers, but a much less favorable distribution was observed among second breast cancers that developed in women younger than 50 years of age. The survival rates after diagnosis of a second breast cancer were much worse than the survival rates after a first breast cancer diagnosis. These differences persisted after control for important covariates such as age and tumor spread and were more pronounced if the first tumor was diagnosed in an advanced stage and if the time interval between diagnosis of the first and second cancers was short. Nevertheless, tumor spread at diagnosis of the second cancer was by far the most important predictor of survival among patients with bilateral breast cancer.

Conclusions. The differences in clinical aspects of first and second breast cancers should be considered in clinical management of patients with breast cancer. Fur- ther research is required to explain these differences. Cancer 1993; 723629-35.

From the *University of Ulm, Unit of Epidemiology, Ulm; and the ?Cancer Registry of Saarland, Statistical Office, Saarbriicken, Germany.

Supported in part by the GDS Searle Foundation for Preventive Medicine and Epidemiology.

Address for reprints: Hermann Brenner, M.D., University of Ulm, Unit of Epidemiology, Am Hochstrass 8, D-89081 Ulm, Ger- many.

Accepted for publication July 19, 1993.

Key words: bilateral, breast cancer, cancer registries, epidemiology.

Cancer registry-based studies from different parts of the world have consistently shown an increased risk of cancer in the contralateral breast among patients with an initial breast cancer compared with the general population. lU7 Findings on the clinical aspects of second cancers of the breast have been less consistent, which is partly due to the large heterogeneity of the design (ranging from case reports to more rigorous epidemio- logic investigations) and methodologic approaches used to assess this question. Although diagnosis of second breast cancers in earlier stages than diagnosis of first breast cancers has been reported occassionally,s~9 most studies did not find such difference^."-'^ Most authors found the patients’ prognosis from the time of diagnosis of the second cancer to be similar to or worse than the prognosis from the time of diagnosis of patients with unilateral breast cancer.8,10,14,18-22 F rom this finding, it was concluded that the second tumor acts as an independent competing cause of death with the first cancer. A common finding in most studies was that the survival rates of bilateral breast carcinoma increased with increasing time intervals between the diagnosis of the first and second breast cancer^.^',^^ Other investigators compared the prognosis from the time of diagnosis of the first tumor among patients with unilateral and bilateral breast cancer and found the prognosis to be the same15.23,24 or better9,’’ in those patients with non- synchronous bilateral carcinomas as in those patients with single tumors.

This approach is misleading, however, because the patients with bilateral breast cancer constitute a selec- tive sample who live long enough to develop a second cancer.

Predominant among other limitations of many previous studies has been the failure to control for important covariates such as age or tumor spread at diagnosis. Furthermore, most studies were based on case series of

3630 CANCER December 25, 2993, Volume 72, No. 12

university hospitals, which are typically highly selec- tive. Finally, the number of patients with bilateral breast cancer was small in most of these case series (typically fewer than 100).

In the current study, we present a population- based analysis of the clinical epidemiology of bilateral breast cancer among 251 patients with bilateral breast cancer registered with the cancer registry of Saarland, Germany, between 1968 and 1987.

Material and Methods

Saarland is a small state located in southwestern Ger- many with a population of approximately 1 million in- habitants. The cancer registry of Saarland was estab- lished in 1966. Within the western part of Germany, it is the only state for which reliable and sufficiently com- plete population-based cancer registration has been performed for the last 25 year^.^^,^^ Registration of cancer patients with the registry comes from multiple sources, including general practitioners, clinicians, radi- ologists, and pathologists, who are involved in diagnosis and treatment of the patients. Approximately 90% of all tumors are histologically verified. Multiple notifications on the same patient (including notifications on second primary tumors) are identified through a so- phisticated computer-assisted record linkage that in- cludes search strategies based on various combinations of patients’ names, birth dates, sex, and addresses. Mor- tality follow-up is ensured by an analogous record linkage of registry data with Saarland’s official mortality statistics (migration of cancer patients across state borders is negligible in this population).

Our analyses are made on the basis of patients with notification of a first invasive breast cancer (Interna- tional Classification of Diseases, 9th revision position 174) to the registry in the 20-year period from January 1, 1968, to December 31, 1987. These patients were followed regarding development of a second contralat- era1 breast cancer and vital status until December 31, 1989. Direct estimates of completeness of population- based cancer registration in Saarland are not available. An indirect indicator of a high level of completeness of breast cancer registration, however, is the fact that the registry was notified of fewer than 5% of the cases by death certificate only.27 The breast cancer was histologically verified in more than 90% of the patients included in the current study. All notifications of bilateral breast cancer were reviewed by one of the authors (B.E.). Ut- most care was taken to distinguish primary from secondary contralateral breast tumors based on all available information, particularly the histopathologic findings, such as in situ components of the second tumors. Nevertheless, unequivocal distinction was difficult in

some cases. To minimize erroneous inclusion of secondary cancers at the contralateral breast, the following additional criteria were required for inclusion of patients with bilateral breast cancer in the study: (1) no evidence of local spread in the first breast across the midline of the chest to the second breast and (2) a mini- mum time period of more than 1 year between the diagnosis of the first and second cancers (metachronous cancers). The clinical aspects of the second breast cancers of patients with bilateral breast cancer, including age and tumor spread at diagnosis and survival, were compared with the clinical aspects of all first breast cancers that the registry was notified of between 1968 and 1987. This comparison was preferred over a comparison with first breast cancers of patients with unilateral breast cancer only (patients with a first diagnosis of breast cancer between 1968 and 1987 and no diagnosis of contralateral breast cancer by the end of 1989), because a contralateral breast cancer may develop in many of these patients in later years. The comparison is also clinically more relevant, because clinicians cannot distinguish between patients with unilateral and, eventually, bilateral breast cancer when treating patients with a first breast cancer.

The statistical analyses were performed on a per- sonal computer using the software packages SAS” and EGRET.29 Survival analyses were performed from the time of diagnosis of the second cancer using the actuarial method (univariate and bivariate analyses) and the Cox proportional hazards model (multivariate analy- s ~ s ) . ~ ’ , ~ ~ Survival rates of all patients with breast cancer that the registry was notified of were calculated from the time of their first diagnosis for comparison using the actuarial method. To investigate the influence of the first cancer on survival, subgroups of patients with bilateral cancer were defined by the tumor spread of the first cancer and the time interval between the diagnosis of the first and second cancers. Survival rates after diagnosis of the second cancer among these subgroups were compared with survival rates after a first breast cancer diagnosis among matched comparison groups. Two comparison patients per patient with bilateral breast cancer were selected from the total group of patients with a first breast cancer diagnosis by individual matching on age, tumor spread, and year of diagnosis to control for potential confounding by these factors.

Results

Between January 1,1968, and December 31,1987,9585 patients with a first diagnosis of breast cancer were registered with the cancer registry among whom 251 patients experienced development of a contralateral

Bilateral Breast CancerlBrenner et al . 3631

Table 1. kge and Tumor Spread at Diagnosis of Breast Cancer

Table 2. Tumor Spread by Age Among Patients With Breast Cancer

Bilateral breast cancert


All First Second Categories vatients* cancer cancer

n = 9585 n = 251 n = 251 Age at diagnosis I 49 yr 23.1% 33.1% 20.3% 50-64 yr 36.1% 40.2% 34.7% 2 65 yr 40.6% 26.7% 45.0% Mean (yr) 60.5 55.7 61.1

n = 6548 n = 186 n = 142 Spread

Local 36.2% 40.9% 33.1% Regional 5 1.6% 54.3% 52.8% Distant 12.2% 4.8% 14.1%

* First diagnosis between 1968 and 1987 t First cancer diagnosed between 1968 and 1987, second cancer diagnosed between 1969 and 1989, with >12 mo between diagnosis of first and second cancers.

breast cancer before December 31, 1989 that met our rigorous inclusion criteria.

Table 1 shows the distribution of age and tumor spread at diagnosis of the first breast cancer among all patients and the first and second breast cancer in the subgroup with bilateral cancer. The age distribution among patients with bilateral breast cancer at the time of diagnosis of the second cancer was similar to the age distribution of all patients with a first diagnosis of breast cancer (see Table 1). The patients with bilateral breast cancer were considerably younger than the comparison group of all patients with breast cancer, however, at the time of detection of their first cancer (mean age, 55.7 versus 60.5 years; P < 0.01 for the difference in age distribution between patients with unilateral and bilateral breast cancer).

The tumor spread at diagnosis was documented in approximately two thirds of all breast cancer notifications, and only those cases are included in the following analyses in which tumor spread is included as a covar- iate. Three levels of tumor spread are distinguished: local (neither lymph node nor distant metastases), regional (lymph node metastases but no distant metastases), and distant. Approximately one third of all first breast cancers with known tumor spread were localized at the time of their diagnosis, whereas approximately one half and one sixth of the tumors showed regional and distant spread, respectively (see Table 1). A similar pattern was observed for the second cancers in patients with bilateral breast cancer, whereas the distribution of tumor spread was slightly more favorable for the first cancers in this group of patients. Simultaneous stratifi-

~~~ ~ ~

All First Second (vr) Svread uatients* cancer cancer Age

I 4 9 Local Regional Distant

Local Regional Distant


50-64

t 65

n = 1701 41.2% 52.7%

6.2% n = 2539

35.5% 53.1% 11.4%

n = 2308 33.4% 49.1 % 17.6%

n = 67 44.8% 53.7%

n = 77 1.5%

36.4% 57.1%

6.5% n = 42

50.0% 42.9%

7.1%

n = 26 19.2% 61.5% 19.2% n = 53 34.0% 49.1% 17.0% n = 63 38.1% 52.4% 9.5%

* First diagnosis between 1968 and 1987 t First cancer diagnosed between 1968 and 1987, second cancer diagnosed between 1969 and 1989, with > 12 mo between diagnosis of first and second cancers.

cation by age and cancer spread at diagnosis revealed the following pattern (see Table 2): whereas the distribution of cancer spread was most favorable in the youngest age group (5 49 years) and worst in the oldest age group among all patients with a first breast cancer, an opposite trend was observed for the second tumors of the patients with bilateral breast cancer. Distribution of tumor spread was much less favorable in second cancers of patients with bilateral breast cancer than among all first breast cancers in the youngest age group ( P < 0.01), whereas only minor, nonsignificant differences were found for the other age groups.

Table 3. Tumor Spread by Year of Diagnosis Among Patients With Breast Cancer


Year of diagnosis Spread

1968-1977 Local Regional Distant


1978-1989

All patients*

n = 2932

54.4%

n = 3616

49.3%

34.1%

11.5%

37.9%

12.8%

First cancer

n = 108

59.3%

n = 78 48.7% 47.4%

3.9%

35.2%

5.6%

Second cancer

n = 28 21.4% 50.0% 28.6%

n = 114 36.0% 53.5% 10.5%

* First diagnosis between 1968 and 1987. t First cancer diagnosed between 1968 and 1987, second cancer diagnosed between 1969 and 1989, with > 12 mo between diagnosis of first and second cancers.

3632 CANCER December 15, 1993, Volume 72, No. 12

Table 4. Two and One-Half Year and Five-Year Survival Probability by Selected Covariates Among Patients With Breast Cancer

All patients: Bilateral breast cancert

n 2.5 yr (%) 5.0 yr (%) n 2.5 yr (%) 5.0 yr (%)

Age (yr) 5 49 2207 83.8 72.2 53 51.5 30.5 50-64 3376 77.3 64.2 87 64.3 40.3 2 65 3574 69.0 53.4 109 67.7 45.4

Local 2372 89.8 79.6 47 91.2 65.0 Regional 3356 72.6 55.5 75 65.6 41.3 Distant 753 28.7 18.4 20 30.0 10.0

1968-1977 4173 73.4 58.8 54 51.9 33.3 1978-1987 4984 77.8 65.6 155 67.0 42.7

5 60 mo NA 135 55.1 36.0 > 60 mo NA 114 72.5 45.6

Spread

Period

Time interva1

NA: not applicable. * First diagnosed between 1968 and 1987. t First cancer diagnosed between 1968 and 1987, second cancer diagnosed between 1969 and 1989, with a time interval of > 12 mo between diagnoses.

Changes in the distribution of tumor spread in all patients with breast cancer between 1968 and 1977 and 1978 and 1987 were modest (see Table 3). In contrast, major progress from an unfavorable distribution of tumor spread between 1968 and 1977 to a distribution that is comparable to the distribution of tumor spread in all patients with breast cancer was achieved for second tumors of patients with bilateral breast cancer (P =

0.035 for the difference between both time intervals). Table 4 summarizes the results of the actuarial sur-

vival analyses. Overall, the prognosis was much worse among patients with a second contralateral breast cancer than among all patients with a first diagnosis of breast cancer, In contrast to the latter, the survival rates increased with increasing age at diagnosis among the patients with bilateral cancer. A more unfavorable prognosis of patients with bilateral breast cancer was found for each category of tumor spread. The prognosis improved over time in both comparison groups, particularly in the patients with bilateral breast cancer who also had a better prognosis if the time interval between diagnosis of the first and second breast cancers was more than 60 months.

In multivariate survival analyses among the patients with bilateral breast cancer, the spread of the second tumor was found to be by far the most important prognostic factor (see Table 5). After simultaneous ad- justment for all other variables listed in Table 5 , the relation between age at diagnosis and prognosis was in the opposite direction (i.e., better prognosis at younger

ages) than in the unadjusted actuarial analyses. Never- theless, this trend must be interpreted with caution due to the wide confidence intervals around the estimates of the relative risk. As before, a short time interval be-

Table 5. Relative Risk of Death After Diagnosis of a Second Primary Neoplasm of the Breast by Selected Variables*

Predictor variable 95% confidence

RR interval

Time between first and second breast cancers

> 60 mo 5 60 mo

Age at diagnosis of second breast cancer (yr)

5 49 50-64 z 65

Spread of second breast cancer


1 .o 1.61

1.0 1.13 2.04

1 .o 2.65 9.66

- 0.98-2.66

- 0.57-2.23 0.99-4.19

- 1.37-5.50 4.35-21.46

R R relative risk of death. * Results of a multivariate analysis (simultaneously including all predictor variables enlisted in the table) among patients with bilateral breast cancer using the Cox proportional hazards model. Inclusion criteria: diagnosis of the first breast cancer between 1968-1987, diagnosis of the second breast cancer between 1969-1989, and a time interval of more than 12 mo between diagnosis of the first and second breast cancers.

Bilateral Breast CancerlBrenner et al. 3633

Table 6. Two and One-Half-Year- and Five-Year Survival Probability Among Patients With Bilateral Breast Cancer and Matched Comparison Groups With a First Diagnosis of Breast Cancer

Bilateral breast cancer* Matched comparison groupt

n 2.5 yr (YO) 5.0 yr (Yo) n 2.5 yr (Yo) 5.0 yr (YO)

Spread of first cancer Local 54 74.3 51.3 105 82.4 69.1 Regional 98 46.8 23.4 194 74.5 62.1

Time interval between first and second cancer (mo)

13-30 57 46.7 32.5 111 76.9 66.4 31-60 54 61.9 34.9 107 70.4 58.7 61-90 44 71.0 38.7 86 71.4 63.2 > 90 45 78.7 51.4 91 83.9 67.9

* Inclusion criteria: diagnosis of the first breast cancer between 1968-1987, diagnosis of the second breast cancer between 1969-1989, and a time interval of more than 12 mo between diagnosis of the h s t and second breast cancers. t Matched on age, tumor spread, and year of diagnosis.

tween diagnosis of the first and second cancers was associated with an increased risk of death.

Table 6 shows the results of the matched survival comparisons. The prognosis of patients with bilateral breast cancer whose first cancer showed regional or distant spread at the time of diagnosis was much worse than the prognosis of the matched comparison group of patients with a first diagnosis of breast cancer. Al- though differences in prognosis were also found between patients with bilateral breast cancer whose first cancer was localized at the time of diagnosis and their comparison group, these differences were much less pronounced. Similarly, the differences in the prognosis between patients with bilateral breast cancer and their matched comparison groups decreased with the time interval between diagnosis of the first and second breast cancers, although differences persisted even for a time interval of more than 90 months. All of the differences in survival rates between the subgroups of patients with bilateral breast cancer and their matched comparison groups of patients with a first diagnosis of breast cancer shown in Table 6 are highly statistically significant ( P < 0.02 in the log-rank test), with the ex- ception of the subgroup, whose time interval was more than 90 months.

Discussion

A limitation of our study is that the extent of histopathologic and clinical information varied widely among registered patients due to the heterogeneity of the notify- ing sources, and overly restrictive data protection rules in Germany hindered access to original files for a stan- dardized completion of the registry data. Thus, we were unable to assess the role of therapeutic factors or to identify potential histopathologic characteristics that

may account for the observed differences in prognosis from our data. Similarly, distinction of secondary and primary contralateral breast cancer was hindered by the varying degree of detail of the histopathologic information. This should not be of too much concern, however, because previous studies have concluded that an intra vitam apparent breast metastasis from a contralateral cancer must be exceedingly rare.15

Despite these limitations, use of population-based cancer registry data is a major strength of our study in that a relatively large number of unselected patients with bilateral breast cancer could be assessed and compared with unselected patients with a first diagnosis of breast cancer. In contrast to many previous investigations, known prognostic factors were considered as potential confounders by appropriate stratification and matching, which allows for more valid comparisons.

Our results clearly demonstrate that patients with bilateral breast cancer have a much worse prognosis at the time of diagnosis of the second breast cancer than patients with a diagnosis of a first breast cancer. This disadvantage in prognosis persists after control for age and tumor spread. The disadvantage is worse if the time interval between the first and second cancers is relatively short and if the first tumor is diagnosed in an advanced stage, which supports the hypothesis that both cancers independently act as competing causes of death. However, a major disadvantage in prognosis persists even for prolonged time intervals between the first and second cancers, which cannot be fully ac- counted for by long-term sequelae of the first tumor. A possible explanation is that patients with bilateral breast tumors differ from patients with unilateral breast tumors in other important prognostic factors, such as histopathologic characteristics of their tumors or immunologic factors. For example, multicentric cancer in the

3634 CANCER December 25,1993, Volume 72, No. 12

specimen of the first primary, lobular carcinoma and anaplastic cancers of the first breast have been found to be related to the development of a second cancer in the contralateral brea~t.~, ' , '~

Similarly, patients who experience development of a first breast cancer at younger ages are at higher risk of experiencing development of a second breast cancer.1.3-5,8,14.32,33 I n our data, this is reflected in the varying age distribution at the time of the first cancer between all patients with breast cancer and patients with bilateral breast cancer (see Table 1). Patients who experience development of a first breast cancer at younger ages may differ from other patients with breast cancer in many respects, such as familial predisposition, which is also considered to be a risk factor for bilateral breast cancer.5,7,33-35 The more unfavorable prognosis after a second breast cancer found in the current study contra- dicts findings of higher survival rates of patients with breast cancer with genetic susceptibility. These findings had focused on first breast cancers, however. Future research is required to address the role of genetic susceptibility for prognosis among women with a second breast cancer.

Progress toward a more favorable distribution of tumor spread of second tumors in recent years may reflect more careful clinical surveillance of patients with breast cancer. Nevertheless, it remains disappointing that second breast cancers are still not diagnosed in earlier stages than first breast cancers. Again, this could reflect different characteristics of tumor growth, espe- cially in young patients with bilateral breast cancer, although our data did not allow us to assess this hypothesis. In young patients with bilateral breast cancer, the stage distribution is particularly unfavorable, which ex- plains the surprising fact that the survival rates of patients with bilateral breast cancer are lowest in the youngest age group. If tumor spread is controlled for in multivariate analyses, survival rates are inversely related to age at diagnosis, as is the case for the total group of all patients with a first diagnosis of breast cancer.

The results of the current study have important im- plications for the clinical management of patients with bilateral breast cancer. Clinicians must be aware of the more unfavorable prognosis of patients with a second breast cancer. Because, as with a first breast cancer, tumor spread at diagnosis is the most important predictor of prognosis, all possible efforts should be made to en- hance early detection of second breast cancers, for example, by frequent clinical examination, mammogra- phy, and selected contralateral biopsy among the high- risk group of patients who already experienced a first breast ~ a n c e r . ~ ~ - ~ ' Further research should address potential histopathologic, immunologic, or other factors that may explain the apparent differences in the clinical

aspects of first and second breast cancers and that may suggest different courses of treatment for patients with a second breast cancer.

References

1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

Harvey EB, Brinton LA. Second cancer following cancer of the breast in Connecticut, 1935-1982. Natl Cancer Znst Monogr

Ewertz M, Mouridsen HT. Second cancer following cancer of the female breast in Denmark, 1943-1980. Natl Cancer Znst Monogr 1985; 68:325-9. Storm HH, Jensen OM. Risk of contralateral breast cancer in Denmark 1943-1980. Br ] Cancer 1986; 54:483-92. Murakami R, Hiyama T, Hanai A, Fujimoto I. Second primary cancers following female breast cancer in Osaka, Japan: a population-based cohort study. ]pi ] CIin Oncol 1987; 17:293-302. Schwartz AG, Ragheb NE, Swanson GM, Satariano WA. Racial and age differences in multiple primary cancers after breast cancer: a population-based analysis. Breast Cancer Res Tr 1989;

Horn PL, Thompson WD. Risk of contralateral breast cancer: associations with histologic, clinical and therapeutic factors. Canccr 1988; 62:412-24. Horn PL, Thompson WD. Risk of contralateral breast cancer: associations with factors related to initial breast cancer. Am ] Epideiniol 1988; 128:309-23. Robbins GF, Berg JW. Bilateral primary breast cancers: a prospective clinicopathologic study. Cancer 1964; 17:1501-27. Burns PE, Dabbs K, May C, Lees AW, Birkett LR, Jenkins HJ, et al. Bilateral breast cancer in northern Alberta: risk factors and survival patterns. Can Med Assoc I 1984; 130:881-6. Leis HI', Mersheimer WL, Black MM, de Chabon A. The second breast. New York Med 1965; 65:2460-8. Buls JG, Bennett RC, Chan DPS. Bilateral carcinoma of the breast. Aust N Z ] Surg 1976; 46:336-40. MuelIer CB, Ames F. Bilateral carcinoma of the breast: fre- quency and mortality. Can I Surg 1978; 21:459-65. Adami HO, Hansen J, Jung B, Lindgren A, Rimsten A. Bilateral carcinoma of the breast: epidemiology and histopathology. Acta Radiol Oncol 1981; 20:305-9. Del Turco MR, Ciatto S , Perigli G, Borrelli D, Carcangiu ML, Santucci M. Bilateral cancer of the breast. Tumor; 1982; 68:155- 60. Fisher ER, Fisher B, Sass R, Wickerham L, and Collaborating NSABP Investigators. Pathologic findings from the National Surgical Adjuvant Breast Project (Protocol No. 4). XI. Bilateral breast cancer. Cancer 1984; 54:3002-11. Rosen PR, Groshen S , Kinne DW. Contralateral breast carcinoma: an assessment of risk and prognosis in stage I (TINOM,,) and stage I1 (T,NIMo) patients with 20-year follow up. Surgery 1989; 106:904-10. Giilay H, Hamaloglu E, Bulut 0, Giiksel HA. Bilateral breast carcinoma: 28 years' experience. World ] Surg 1990; 14:529-34. Khafagy MM, Schottenfeld D, Robbins GF. Prognosis of the second breast cancer. Cancer 1975; 35:596-9. Fracchia AA, Robinson D, Legaspi A, Greenall MJ, Kinne DW, Groshen S . Survival in bilateral breast cancer. Cancer 1985; 55: 1414-21.

1985; 68~99-112.

14~245-54.

Bilateral Breast CancerlBrenner et al . 3635

20.

21.

22.

23.

24.

25.

26.

27.

28

Bodian C, Haagensen CD. Bilateral carcinoma of the breast. In: Haagensen CD, editor. Diseases of the breast. 3rd ed. Philadel- phia: WB Saunders, 1986:440-55. Bert6 E, Buzdar A, Smith TL, Hortobagyi GN. Bilateral primary breast cancer in patients treated with adjuvant therapy. Am ] Clin Oncol 1988; 11:114-8. Holmberg L, Adami HO, Ekbom A, Bergstrom R, Sandstrom A, Lindgren A. Prognosis in bilateral breast cancer: effects of time interval between first and second primary tumours. Br ] Cancer

Slack NH, Bross IDJ, Nemoto T, Fisher B. Experiences with bilateral primary carcinoma of the breast. Surg Gynec Obstet 1973;

McCredie JA, Inch WR, Alderson M. Consecutive primary carcinomas of the breast. Cancer 1975; 35:1472-7. Bailey MJ, Royce C, Sloane JP, Ford HT, Powles TJ, Gazet J-C. Bilateral carcinoma of the breast. Br ] Surg 1980; 67:514-6. Muir C, Waterhouse J, Mack T, Powell J, Whelan S, editors. Cancer incidence in five continents, Vol. V. IARC Scientific Publications No. 88. Lyon: International Agency for Research on Cancer, 1987. Schon D, Bertz J, Hoffmeister H, editors. Bevolkerungsbezogene Krebsregister in der Bundesrepublik Deutschland, Vol. 2. Miinchen: MMV Medizin Verlag, 1989. SAS Institute, Inc. SAS user’s guide, 1982 ed. Cary, NC: SAS Institute, lnc., 1982.

1988; 58:191-4.

136:433-40.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

Statistics and Epidemiology Research Corporation: EGRET. First Draft Manual. Seattle: Statistics and Epidemiology Research Corp., 1990. Cutler SJ, Ederer F. Maximum utilization of the life-table method in analyzing survival. ] Chronic Dis 1958; 8:699-712. Cox DR. Regression models and life tables. ] Roy Stat Soc 1972;

Adami H-0, Bergstrom R, Hansen J. Age at first primary as a determinant of the incidence of bilateral breast cancer. Cancer

Anderson DE, Badzioch MD. Bilaterality in familial breast cancer patients. Cancer 1985; 56:2092-8. Harris RE, Lynch HT, Guirgis HA. Familial breast cancer: risk to the contralateral breast. ] Natl Cancer lnst 1978; 60:955-60. Hislop TG, Elwood JM, Coldman AJ, Spinelli JJ, Worth AJ, Elli- son LG. Second primary cancers of the breast: incidence and risk factors. Br ] Cancer 1984; 49:79-85. Wanebo HJ, Senofsky GM, Fechner RE, Kaiser D, Lynn S, Para- dies ]. Bilateral breast cancer: risk reduction by contralateral biopsy (with discussion). Ann Surg 1985; 201:667-77. Senofsky GM, Wanebo HJ, Wilhelm MC, Pope TL, Fechner RE, Broaddus W, et al. Has monitoring of the contralateral breast improved the prognosis in patients treated for primary breast cancer? Cancer 1986; 57:597-602. Sears HF, Janus C, McDermott A, Grotzinger P. Bilateral breast carcinoma: prospective evaluation of factors assisting diagnosis. ] Surg Oncol 1986; 32:203-7.

34: 187-202.

1985; 55~643-7.

Documents

Clinical epidemiology of bilateral breast cancer