- Email: [email protected]
Second Primary Cancers in Men With Prostate Cancer: An Increased Risk of Male Breast Cancer
0022-5347/03/1694-1345/0 THE JOURNAL OF UROLOGY® Copyright © 2003 by AMERICAN UROLOGICAL ASSOCIATION
Vol. 169, 1345–1348, April 2003 Printed in U.S.A.
DOI: 10.1097/01.ju.0000056706.88960.7c
SECOND PRIMARY CANCERS IN MEN WITH PROSTATE CANCER: AN INCREASED RISK OF MALE BREAST CANCER ¨ RN TAVELIN CAMILLA THELLENBERG, BEATRICE MALMER, BJO
AND
¨ NBERG HENRIK GRO
From the Department of Radiation Sciences, Oncology, Umeå University, Umeå, Sweden
ABSTRACT
Purpose: Evaluation of second primary cancers provides valuable insight about etiology and shared risk factors. Studies of second primary cancers following prostate cancer conclude that overall risk of second primary cancers decreases. However, risk of bladder cancer and kidney cancer increases. We examine the risk of common and rare second primary cancers following prostate cancer in a large population based cohort to identify possible common etiological factors. Materials and Methods: All prostate cancer cases in the Swedish Cancer Registry (135, 713) from 1958 to the end of 1996 constituted the study base. Risk (standardized incidence ratio) of second primary cancers was calculated as the ratio between observed and expected number of cancers. We used 2-tailed 95% confidence intervals (CI) to test significance. Results: An overall increased risk (standardized incidence ratio 1.17, 95% CI 1.15–1.19) of second primary cancers was found but was only seen in the first 6 months of followup (ratio 3.45, 3.32–3.57). The most interesting finding was an increased risk (ratio 2.01, 95% CI 1.44 –2.74) of male breast cancer. Other tumor sites with increased risk were the small intestine (standardized incidence ratio 1.39, 95% CI 1.09 –1.51), skin melanoma (ratio 1.33, 95% CI 1.16 –1.51) and endocrine tumors (ratio 1.41, 95% CI 1.13–1.74). Conclusions: A small but increased risk of second primary cancers following prostate cancer was found, most likely due to increased surveillance during the first 6 months after diagnosis. However, following prostate cancer there is an increased risk of endocrine related second primary cancers such as male breast cancer and carcinoids in the small intestine. To our knowledge these associations have not been reported previously, and they warrant more study. KEY WORDS: prostatic neoplasms; neoplasms; second primary; breast neoplasms; male; registries
Prostate cancer is the most common cancer in Sweden with approximately 7,400 new cases every year. It is mostly diagnosed in elderly men but the impact of prostate cancer mortality is substantial as about 2,500 men die of this disease each year. Despite the high incidence in the population, few etiological factors have been identified. To date the only well-known and widely accepted risk factors are age, ethnicity and positive family history.1 Studying the incidence of second primary cancers provides insight into shared etiological factors of the initial and subsequent cancer. Several different explanations for the occurrence of second primary cancer have been identified, including common genetic predisposition (for example the LiFraumeni syndrome and hereditary nonpolyposis colon cancer), shared environmental factors such as tobacco use, or treatment effects of chemotherapy, radiotherapy or hormonal therapy used to treat the initial cancer. Previous studies conclude that generally there is a decreased overall risk of second primary cancer after prostate cancer.2– 4 However, some studies have shown an increased risk of bladder and kidney cancer but they involved a small number of patients.4 – 6 We attempted to estimate accurately the risk of subsequent cancers after prostate cancer in a large population based cohort. By performing a larger study it is possible to study unusual tumors that are difficult to study in small cohorts. High quality data from the Swedish Cancer Registry from 1958 to 1996 were used, which allowed for nearly complete followup of incidence data from the same well-defined population as the study cohort.
MATERIALS AND METHODS
The study cohort was defined as all men diagnosed with malignant adenocarcinoma of the prostate gland (ICD7 177) who were reported to the Swedish Cancer Registry from 1958 to 1996. Initially all cases were included in the study (including cancers discovered during autopsy) for a total of 135,713. Followup is almost complete because the official records of population data in Sweden are accurate. A second primary cancer was defined as a malignant tumor diagnosed on the same date or after prostate cancer was diagnosed. Accuracy in reporting second primary cancers in the Swedish Cancer Registry has been reviewed extensively and is good, with 94% concordance between data in the registry, and medical and pathological records.7 Person-years were calculated from the date of prostate cancer diagnosis to the date of the second primary cancer, date of death, loss to followup (that is, migration out of Sweden) or December 31, 1996 (whichever came first). The program PYRS (IARC, Lyon, France) was used. The expected number of cancer cases was calculated by multiplying the calendar interval and age specific cancer incidence rate for Sweden by person-years in 5-year intervals. Cancer incidence rates for Sweden for 1958 to 1996 were obtained from the Swedish Cancer Registry and used in all calculations. The standardized incidence ratio was defined as the ratio between the observed and expected number of cases. A 2-tailed 95% confidence interval (CI) of the standardized incidence ratio was calculated according to Byar’s formula. To evaluate if the observed number of cases differed significantly from the expected number, the aforementioned CI was used. A difference was considered significant if the CI did not overlap unity. We used a time
Accepted for publication November 27, 2002. Supported by grants from the Lion’s Cancer Research Foundation, Umeå University. 1345
1346
MALE BREAST CANCER RISK IN PATIENTS WITH PROSTATE CANCER
cutoff of 6 months because we did not primarily sort out autopsy cases, which would exclude second primary cancers diagnosed at autopsy with the prostate cancer and in the diagnostic evaluation at the first cancer diagnosis. A cutoff of 6 months also would possibly identify clinically significant second primary cancers.
tinal tract cancers (except small intestine) and smoking related cancers displayed decreased or no difference in risk. Brain tumors expressed no increased risk after the first 6 months of followup but when analyzing brain tumors (gliomas and meningiomas) separately a small increase in risk of meningiomas was noted.
RESULTS
DISCUSSION
There were 10,526 second primary cancers observed among 135,713 patients with prostate cancer, of which 8,984 were expected, for a modest overall increased risk (standardized incidence ratio 1.17 95% CI 1.15–1.19) of second primary cancers. However, when looking at second primary cancers diagnosed after the initial 6 months following prostate cancer diagnosis, there was a decreased risk (standardized incidence ratio 0.93 95% CI 0.91– 0.94), which was more pronounced at older ages (table 1). For the 2 youngest cohorts (age less than 60 and 60 to 64 years) risk was not significantly decreased. A significantly increased risk (standardized incidence ratio 2.01 95% CI 1.44 –2.74) of male breast cancer after prostate cancer was observed in 40 cases (table 2). The increased risk was higher for the first 6 months after prostate cancer diagnosis (nonsignificant) but persisted later when most of the tumors were diagnosed. A younger age at prostate cancer diagnosis is (less than 60 years) increased the risk (standardized incidence ratio 5.29, 95% CI 1.06 –15.47) of subsequent breast cancer, and the risk continued to be increased significantly for patients older than 70 (1.85, 1.18 – 2.75) (table 3). From 1958 to 1980 the risk of male breast cancer was somewhat higher than that in the following period studied (standardized incidence ratio 2.12, 95% CI 1.35–3.19 versus 1.48, 0.77–2.59). An increased risk of tumors developing in the small intestine (standardized incidence ratio 2.20, 95% CI 1.83–2.61), mostly carcinoids, was observed. Carcinoids make up about 25% to 50% of all tumors in the small bowel. In this study the second primary cancer in the small bowel consisted of 70.5% carcinoids, 18.6% adenocarcinomas and 10.9% various histopathologies. The risk was higher the earlier in life the patient was diagnosed with prostate cancer but persisted across the age cohorts (table 3). After 6 months of followup there was an increased risk of malignant melanomas and endocrine tumors (parathyroid, thymus, pituitary, adrenal gland and endocrine pancreas). Among endocrine tumors parathyroid tumors were the most common (46.8% of all observed cases). Melanoma showed only a slight change in risk with respect to early or late diagnosis after prostate cancer, and its risk decreased 10 years after prostate cancer diagnosis regardless of age at diagnosis (table 4). A high increased risk was present for kidney (standardized incidence ratio 7.06, 95% CI 6.71–7.90) and bladder cancer (7.29, 6.71–7.90) for the first 6 months of followup and then disappeared in the longer followup. The exception to this pattern was in young men (age less than 60 at prostate cancer diagnosis) having an increased risk of bladder cancer even after the first 6 months. Gastrointes-
Our study revealed a new and interesting finding of an increased risk for male breast cancer. A possible explanation might be that these tumors metastasize from prostate cancer and are mistakenly diagnosed as mammary gland tumors. Recently it has been shown that prostate specific antigen is present in breast tissue and, if this is not recognized earlier, it can lead to misclassifications and fewer diagnoses of breast cancer.8 A more plausible explanation may be that hormonal treatment of the prostate cancer with estrogens induced the breast cancer.9 This explanation might be supported by the fact that the risk of breast cancer was higher in the early period of our study from 1958 to 1980 when estrogens were used more often in the treatment of prostate cancer. Similarly case reports of breast cancer and the Reifenstein syndrome have been reported.10 The Reifenstein syndrome is characterized by infertility, testicular feminization, frequent hypospadias and other defects. These men have germline mutations in the androgen receptor that cause androgen insufficiency and, therefore, estrogen excess. Another genetic condition is Klinefelter’s syndrome, associated with an increased risk of male breast cancer but not with prostate cancer.11 Genetic factors may explain the association of breast and prostate cancer because several studies have reported an increased risk of male and female breast cancer in families with multiple cases of prostate cancer.12 However, Olsson et al reported a decreased risk of prostate cancer in relatives of men with breast cancer.13 Mutations in the BRCA2 gene have been associated with prostate cancer and male breast cancer. In a population based study of 34 male breast cancer cases from southern Sweden mutations in the BRCA2 gene were found in approximately 20%.14 In families with breast cancer from Iceland with the BRCA2 founder mutation 995 del5 a clustering of prostate cancer and male breast cancer is present. Men carrying the mutation seem to have aggressive disease.15 The Breast Cancer Linkage Consortium found an RR of 4.65 (95% CI 3.48 – 6.22) for prostate cancer in BRCA2 mutation carriers, with lifetime risk estimations of 19.8% for prostate cancer and 6% for male breast cancer by carriers age 70 years.16 The incidence of male breast cancer in Sweden is similar to that of the rest of Europe but lower than that of black and Jewish populations. The reported high risk of kidney and bladder cancer shown in some studies4, 5 could not be verified in our study except in the first 6 months. Most bladder cancers were diagnosed on the same day or the day after prostate cancer diagnosis (data not shown). The diagnosis of prostate and bladder cancer is likely due to autopsy cases or cases diagnosed in the routine diagnostic evaluation, cystoscopy or urography. A few studies
TABLE 1. Differences in risk of second primary cancer depending on age at prostate cancer diagnosis 0–5 Mos. After Diagnosis Age (yrs.)
No. Cases
Standardized Incidence Ratio (95% CI)
Less than 60 60–64 65–69 70–79 Greater than 80
67 146 354 1,335 1,104
All ages
3,006
6 or More Mos. After Diagnosis
All Cases
No. Cases
Standardized Incidence Ratio (95% CI)
No. Cases
Standardized Incidence Ratio (95% CI)
6.12 (4.74–7.78) 4.12 (3.48–4.85) 3.77 (3.39–4.18) 3.05 (2.89–3.22) 3.75 (3.53–3.98)
295 691 1,328 3,810 1,396
1.07 (0.94–1.20) 1.01 (0.94–1.09) 0.95 (0.91–1.01) 0.92 (0.89–0.95) 0.85 (0.81–0.90)
362 837 1,682 5,145 2,500
1.26 (1.13–1.40) 1.21 (1.13–1.29) 1.14 (1.08–1.19) 1.12 (1.10–1.16) 1.30 (1.25–1.35)
3.45 (3.32–3.57)
7,520
0.93 (0.91–0.95)
10,526
1.17 (1.15–1.19)
1347
MALE BREAST CANCER RISK IN PATIENTS WITH PROSTATE CANCER TABLE 2. Sites of second primary cancers diagnosed after prostate cancer diagnosis 0–5 Mos. After Diagnosis ICD7 Ca Site (No. cases)
6 or More Mos. After Diagnosis
All Cases
Standardized Incidence Ratio Standardized Incidence Ratio Standardized Incidence Ratio No. Cases No. Cases No. Cases (95% CI) (95% CI) (95% CI)
Head/neck (140/148) Esophagus (150) Stomach (151) Small intestine (152) Colon (153) Rectum (154) Liver (155) Pancreas (157) Lung (162) Male breast (170) Kidney (180) Bladder (181) Melanoma (190) Brain (193) Endocrine (195) Hematological mal. (200–5)
34 29 223 55 308 133 149 152 371 5 279 592 24 114 38 250
1.26 (0.87–1.76) 2.01 (1.35–2.89) 2.54 (2.22–2.89) 9.76 (7.35–12.71) 3.22 (2.87–3.60) 2.18 (1.83–2.59) 4.42 (3.74–5.19) 3.72 (3.16–4.37) 3.49 (3.14–3.86) 2.63 (0.85–6.14) 7.06 (6.71–7.90) 7.29 (6.71–7.90) 1.32 (0.85–1.97) 6.20 (5.11–7.44) 5.37 (3.80–7.37) 3.15 (2.77–3.57)
207 98 667 74 945 517 279 321 730 35 316 768 225 175 88 671
0.86 (0.74–0.98) 0.76 (0.62–0.93) 0.87 (0.80–0.94) 1.39 (1.09–1.75) 1.03 (0.97–1.10) 0.92 (0.84–1.00) 0.86 (0.77–0.97) 0.87 (0.77–0.97) 0.78 (0.72–0.84) 1.95 (1.36–2.71) 0.90 (0.80–1.00) 1.00 (0.93–1.07) 1.33 (1.16–1.51) 1.09 (0.94–1.26) 1.41 (1.13–1.74) 0.91 (0.84–0.98)
241 127 890 129 1,253 650 428 473 1,101 40 595 1,360 249 289 126 921
0.90 (0.79–1.02) 0.89 (0.74–1.06) 1.04 (0.97–1.11) 2.20 (1.83–2.61) 1.24 (1.17–1.31) 1.04 (0.96–1.12) 1.20 (1.09–1.32) 1.15 (1.05–1.26) 1.05 (0.99–1.12) 2.01 (1.44–2.74) 1.52 (1.40–1.65) 1.59 (1.51–1.68) 1.33 (1.17–1.50) 1.62 (1.44–1.81) 1.81 (1.51–2.16) 1.13 (1.05–1.20)
TABLE 3. Prostate cancer diagnoses at different patient ages and influence on risk of second primary cancers Age Less Than 60 ICD7 Ca Site (No. cases)
Age 60–69
Age Greater Than 70
Standardized Incidence Ratio Standardized Incidence Ratio Standardized Incidence Ratio No. Cases No. Cases No. Cases (95% CI) (95% CI) (95% CI)
Small intestine (152) 5 2.62 (0.84–6.11) Colon (153) 18 0.72 (0.43–1.14) Rectum (154) 14 0.77 (0.42–1.29) Male breast (170) 3 5.29 (1.06–15.47) Kidney (180) 14 0.94 (0.51–1.58) Bladder (181) 47 1.77 (1.30–2.36) Melanoma (190) 12 1.26 (0.65–2.20) Endocrine (195) 3 0.80 (0.16–2.33) Second primary cancer diagnosed after the first 6 months only.
24 203 126 8 90 233 71 28
1.74 (1.11–2.58) 0.97 (0.84–1.11) 0.88 (0.74–1.05) 1.81 (0.78–3.57) 0.91 (0.75–1.11) 1.15 (1.00–1.30) 1.38 (1.08–1.74) 1.39 (0.92–2.01)
45 724 377 24 121 488 142 57
1.21 (0.88–1.61) 1.06 (0.99–1.14) 0.94 (0.84–1.04) 1.85 (1.18–2.75) 0.89 (0.78–1.12) 0.90 (0.82–0.98) 1.31 (1.10–1.54) 1.48 (1.12–1.92)
TABLE 4. Risk of second primary cancer after prostate cancer according to time after prostate cancer diagnosis 0.5–5 Yrs. Ca. Site
Small intestine Colon Rectum Male breast Kidney Bladder Melanoma Brain Endocrine
No. Cases
Standardized Incidence Ratio (95% CI)
48 579 307 25 198 498 142 103 64
1.51 (1.11–2.00) 1.07 (0.98–1.16) 0.91 (0.81–1.01) 2.18 (1.41–3.22) 0.91 (0.79–1.05) 1.08 (0.99–1.18) 1.39 (1.17–1.64) 1.03 (0.84–1.25) 1.64 (1.26–2.09)
5–10 Yrs.
10 or More Yrs.
No. Cases
Standardized Incidence Ratio (95% CI)
No. Cases
Standardized Incidence Ratio (95% CI)
19 248 155 7 85 178 63 44 13
1.39 (0.84–2.17) 1.05 (0.93–1.19) 1.08 (0.92–1.27) 1.47 (0.59–3.03) 0.96 (0.77–1.19) 0.90 (0.77–1.04) 1.47 (1.13–1.88) 1.10 (0.80–1.48) 0.82 (0.44–1.41)
7 118 55 3 33 92 20 28 11
0.90 (0.36–1.85) 0.84 (0.70–1.01) 0.67 (0.51–0.88) 1.11 (0.22–3.25) 0.70 (0.48–0.98) 0.81 (0.65–0.99) 0.83 (0.51–1.28) 1.34 (0.89–1.94) 1.38 (0.69–2.46)
address the question of whether full dose radiotherapy of the prostate gland and pelvic lymphatic nodules for prostate cancer induces second primary cancer in adjacent tissues. Brenner et al showed that in a large number of patients with prostate cancer treated with surgery or radiotherapy there was a significant increase in risk of solid tumors and infield sarcomas in the radiotherapy group.17 In addition, there was an increased risk of bladder cancer among men treated with radiation. In our study the increased risk of bladder cancer occurred in the early period after prostate cancer diagnosis and, thus, could not be due to radiation treatment. No increased risk was present for sarcomas. However, no information on primary treatment is available in the Swedish Cancer Registry. Malignant skin melanomas show an increased risk and persist for years after prostate cancer diagnosis. Greenberg et al also reported an increased risk of melanoma following prostate cancer.5 This finding may be due to hormonal influence as suggested by Levi et al in their study of second primary cancers following cutaneous malignant melanoma.18 They found a standardized incidence ratio of 2.1 (95% CI
1.2–3.4) for prostate cancer as second primary but their study was based on few cases. The finding of an increased risk of small bowel cancer after prostate cancer is intriguing but has been reported previously. A survey of multiple primaries following cancer of the digestive system in Connecticut revealed an increase in risk of 3.1 for prostate cancer after small intestine tumor diagnosis.19 However, no histology was reported in this study. An increased risk of prostate cancer with high levels of insulinlike growth factor-1 has been reported,20 and the strong influence of insulin-like growth factor-1 on the growth of carcinoids could explain this association. A third of the entire small bowel tumors in our study were diagnosed at autopsy, and more than half of the cases had carcinoid histology. These numbers are reflected by the high standardized incidence ratio of 9.76 (95% CI 7.35–12.71) in the first 6 months of followup. According to the Data Swedish Cancer Registry approximately 25% of cancers of the small intestine were generally diagnosed at an autopsy. As in other studies2– 4 there was a decreased risk of subsequent malignancies that were more pronounced in older pa-
1348
MALE BREAST CANCER RISK IN PATIENTS WITH PROSTATE CANCER TABLE 5. Cohort studies of second primary cancers following prostate cancer References/Study Base
No. Pts.
Standardized Incidence Ratio (95% CI)
Kleinerman et al/Connecticut Tumor Registry2 Osterlind et al/Danish Tumour Registry3 Greenberg et al/Georgia Center for Cancer Statistics5
18,135
0.85 (0.80–0.90)
Lung Ca (0.7), digestive system (0.8)
19,886
0.51 (0.47–0.55)
McCredie et al/New South Wales Central Cancer Registry, Australia4 Pawlish et al/Metropolitan Detroit Cancer Surveillance System6 Present study/Swedish Cancer Registry
23,067
0.79 (0.75–0.84)
Respiratory Ca (0.5, 0.4–0.6), digestive system (0.5, 0.5–0.6), skin melanoma (1.8, nonsignificant) Bladder (2.2, p ⬍0.01) risk confined to first 3 yrs., kidney (3.3, p ⬍0.01) for year 1 only, skin melanoma (2.5, p ⬍0.05) Kidney (1.3, 1.0–1.7), lung (0.6, 0.5–0.7), gastric (0.7, 0.6– 0.9), pancreas (0.6, 0.4–0.9)
9,794
1.01 (0.95–1.07)
135,713
1.16 (1.12–1.19) 0.92 (0.90–0.94)
3,675
1.2 (1.1–1.4)
tients. This decrease suggests a tendency to overlook symptoms that might be due to a new malignancy in the elderly, and others have suggested such surveillance bias as an explanation. Most of the previous studies, including ours, have revealed decreased risk of smoking related cancers such as lung, and head and neck tumors.2– 4 In our study several uncommon cancers (male breast cancer, small bowel tumors and endocrine tumors) were found at a higher frequency than expected. A reason that these uncommon cancers have not been reported may relate to the small populations of the previous studies. At least 6 times as many patients as the largest of the previous studies were included in our study (table 5). Despite that fact, however, the number of male breast cancer cases was small. Using the Swedish Cancer Registry, which consists of a well-defined population, made it possible to produce high quality data on second primary cancers. In addition, almost complete followup should provide a clear indication of the risks, and there is a low percentage of misclassifications in the registry as demonstrated by Frodin et al.7 A problem associated with our study is the fact that no data regarding primary treatment or disease stage are reported to the registry. In addition, performing a study this large may generate random associations by multiple testing. Coincidental discovery of a second primary cancer at autopsy might not have clinical importance but could be noteworthy when explaining common etiologies. Different time cutoffs have been used in other studies when calculating risks, and there are no data that indicate 1 or 6 months as superior. However, we believe that 6 months is sufficient to complete diagnostic evaluation of prostate cancer.
Bladder (1.57, 1.34–1.83) significant for first 2 months only, lung (0.83, 0.74–0.93), male breast (1.66, 0.54–3.88) All second primary Ca together Followup from 6 months after diagnosis of prostate Ca male breast cancer (1.95, 1.36–2.71), small intestine (1.39, 1.09– 1.75), endocrine tumors (1.41, 1.13–1.74)
5. 6.
7. 8.
9. 10.
11. 12.
13.
14.
CONCLUSIONS
We report significantly increased risk of male breast cancer following prostate cancer and a decreased overall risk for second primary cancer. Several uncommon endocrine cancers and skin melanomas also showed significantly increased risk of occurrence after prostate cancer. These findings warrant further detailed investigations.
15.
REFERENCES
18.
1. Steinberg, G. D., Carter, B. S., Beaty, T. H., Childs, B. and Walsh, P. C.: Family history and the risk of prostate cancer. Prostate, 17: 337, 1990 2. Kleinerman, R. A., Liebermann, J. V. and Li, F. P.: Second cancer following cancer of the male genital system in Connecticut, 1935– 82. Natl Cancer Inst Monogr, 68: 139, 1985 3. Osterlind, A., Rorth, M. and Prener, A.: Second cancer following cancer of the male genital system in Denmark, 1943– 80. Natl Cancer Inst Monogr, 68: 341, 1985 4. McCredie, M., Macfarlane, G. J., Stewart, J. and Coates, M.: Second primary cancers following cancers of the kidney and
Comments (Standardized Incidence Ratio, 95% CI)
16. 17.
19.
20.
prostate in New South Wales (Australia), 1972–91. Cancer Causes Control, 7: 337, 1996 Greenberg, R. S., Rustin, E. D. and Clark, W. S.: Risk of genitourinary malignancies after cancer of the prostate. Cancer, 61: 396, 1988 Pawlish, K. S., Schottenfeld, D., Severson, R. and Montie, J. E.: Risk of multiple primary cancers in prostate cancer patients in the Detroit metropolitan area: a retrospective cohort study. Prostate, 33: 75, 1997 Frodin, J. E., Ericsson, J. and Barlow, L.: Multiple primary malignant tumors in a national cancer registry—reliability of reporting. Acta Oncol, 36: 465, 1997 Gupta, R. K.: Immunoreactivity of prostate-specific antigen in male breast carcinomas: two examples of a diagnostic pitfall in discriminating a primary breast cancer from metastatic prostate carcinoma. Diagn Cytopathol, 21: 167, 1999 Schlappack, O. K., Braun, O. and Maier, U.: Report of two cases of male breast cancer after prolonged estrogen treatment for prostatic carcinoma. Cancer Detect Prev, 9: 319, 1986 Wooster, R., Mangion, J., Eeles, R., Smith, S., Dowsett, M., Averill, D. et al: A germline mutation in the androgen receptor gene in two brothers with breast cancer and Reifenstein syndrome. Nat Genet, 2: 132, 1992 Ravandi-Kashani, F. and Haynes, T. G.: Male breast cancer: a review of the literature. Eur J Cancer, 34: 1341, 1998 Valeri, A., Fournier, G., Morin, V., Morin, J. F., Drelon, E., Mangin, P. et al: Early onset and familial predisposition to prostate cancer significantly enhance the probability for breast cancer in first degree relatives. Int J Cancer, 86: 883, 2000 Olsson, H., Andersson, H., Johansson, O., Moller, T. R., Kristoffersson, U. and Wenngren, E.: Population-based cohort investigations of the risk for malignant tumors in first-degree relatives and wives of men with breast cancer. Cancer, 71: 1273, 1993 Haraldsson, K., Loman, N., Zhang, Q. X., Johannsson, O., Olsson, H. and Borg, A.: BRCA2 germ-line mutations are frequent in male breast cancer patients without a family history of the disease. Cancer Res, 58: 1367, 1998 Sigurdsson, S., Thorlacius, S., Tomasson, J., Tryggvadottir, L., Benediktsdottir, K., Eyfjord, J. E. et al: BRCA2 mutation in Icelandic prostate cancer patients. J Mol Med, 75: 758, 1997 Cancer risks in BRCA2 mutation carriers. The Breast Cancer Linkage Consortium. J Natl Cancer Inst, 91: 1310, 1999 Brenner, D. J., Curtis, R. E., Hall, E. J. and Ron, E.: Second malignancies in prostate carcinoma patients after radiotherapy compared with surgery. Cancer, 88: 398, 2000 Levi, F., La Vecchia, C., Randimbison, L., Te, V. C. and Erler, G.: Incidence of invasive cancers following cutaneous malignant melanoma. Int J Cancer, 72: 776, 1997 Hoar, S. K., Wilson, J., Blot, W. J., McLaughlin, J. K., Winn, D. M. and Kantor, A. F.: Second cancer following cancer of the digestive system in Connecticut, 1935– 82. Natl Cancer Inst Monogr, 68: 49, 1985 Stattin, P., Bylund, A., Rinaldi, S., Biessy, C., Dechaud, H., Stenman, U. H. et al: Plasma insulin-like growth factor-I, insulin-like growth factor-binding proteins, and prostate cancer risk: a prospective study. J Natl Cancer Inst, 92: 1910, 2000
Vol. 169, 1345–1348, April 2003 Printed in U.S.A.
DOI: 10.1097/01.ju.0000056706.88960.7c
SECOND PRIMARY CANCERS IN MEN WITH PROSTATE CANCER: AN INCREASED RISK OF MALE BREAST CANCER ¨ RN TAVELIN CAMILLA THELLENBERG, BEATRICE MALMER, BJO
AND
¨ NBERG HENRIK GRO
From the Department of Radiation Sciences, Oncology, Umeå University, Umeå, Sweden
ABSTRACT
Purpose: Evaluation of second primary cancers provides valuable insight about etiology and shared risk factors. Studies of second primary cancers following prostate cancer conclude that overall risk of second primary cancers decreases. However, risk of bladder cancer and kidney cancer increases. We examine the risk of common and rare second primary cancers following prostate cancer in a large population based cohort to identify possible common etiological factors. Materials and Methods: All prostate cancer cases in the Swedish Cancer Registry (135, 713) from 1958 to the end of 1996 constituted the study base. Risk (standardized incidence ratio) of second primary cancers was calculated as the ratio between observed and expected number of cancers. We used 2-tailed 95% confidence intervals (CI) to test significance. Results: An overall increased risk (standardized incidence ratio 1.17, 95% CI 1.15–1.19) of second primary cancers was found but was only seen in the first 6 months of followup (ratio 3.45, 3.32–3.57). The most interesting finding was an increased risk (ratio 2.01, 95% CI 1.44 –2.74) of male breast cancer. Other tumor sites with increased risk were the small intestine (standardized incidence ratio 1.39, 95% CI 1.09 –1.51), skin melanoma (ratio 1.33, 95% CI 1.16 –1.51) and endocrine tumors (ratio 1.41, 95% CI 1.13–1.74). Conclusions: A small but increased risk of second primary cancers following prostate cancer was found, most likely due to increased surveillance during the first 6 months after diagnosis. However, following prostate cancer there is an increased risk of endocrine related second primary cancers such as male breast cancer and carcinoids in the small intestine. To our knowledge these associations have not been reported previously, and they warrant more study. KEY WORDS: prostatic neoplasms; neoplasms; second primary; breast neoplasms; male; registries
Prostate cancer is the most common cancer in Sweden with approximately 7,400 new cases every year. It is mostly diagnosed in elderly men but the impact of prostate cancer mortality is substantial as about 2,500 men die of this disease each year. Despite the high incidence in the population, few etiological factors have been identified. To date the only well-known and widely accepted risk factors are age, ethnicity and positive family history.1 Studying the incidence of second primary cancers provides insight into shared etiological factors of the initial and subsequent cancer. Several different explanations for the occurrence of second primary cancer have been identified, including common genetic predisposition (for example the LiFraumeni syndrome and hereditary nonpolyposis colon cancer), shared environmental factors such as tobacco use, or treatment effects of chemotherapy, radiotherapy or hormonal therapy used to treat the initial cancer. Previous studies conclude that generally there is a decreased overall risk of second primary cancer after prostate cancer.2– 4 However, some studies have shown an increased risk of bladder and kidney cancer but they involved a small number of patients.4 – 6 We attempted to estimate accurately the risk of subsequent cancers after prostate cancer in a large population based cohort. By performing a larger study it is possible to study unusual tumors that are difficult to study in small cohorts. High quality data from the Swedish Cancer Registry from 1958 to 1996 were used, which allowed for nearly complete followup of incidence data from the same well-defined population as the study cohort.
MATERIALS AND METHODS
The study cohort was defined as all men diagnosed with malignant adenocarcinoma of the prostate gland (ICD7 177) who were reported to the Swedish Cancer Registry from 1958 to 1996. Initially all cases were included in the study (including cancers discovered during autopsy) for a total of 135,713. Followup is almost complete because the official records of population data in Sweden are accurate. A second primary cancer was defined as a malignant tumor diagnosed on the same date or after prostate cancer was diagnosed. Accuracy in reporting second primary cancers in the Swedish Cancer Registry has been reviewed extensively and is good, with 94% concordance between data in the registry, and medical and pathological records.7 Person-years were calculated from the date of prostate cancer diagnosis to the date of the second primary cancer, date of death, loss to followup (that is, migration out of Sweden) or December 31, 1996 (whichever came first). The program PYRS (IARC, Lyon, France) was used. The expected number of cancer cases was calculated by multiplying the calendar interval and age specific cancer incidence rate for Sweden by person-years in 5-year intervals. Cancer incidence rates for Sweden for 1958 to 1996 were obtained from the Swedish Cancer Registry and used in all calculations. The standardized incidence ratio was defined as the ratio between the observed and expected number of cases. A 2-tailed 95% confidence interval (CI) of the standardized incidence ratio was calculated according to Byar’s formula. To evaluate if the observed number of cases differed significantly from the expected number, the aforementioned CI was used. A difference was considered significant if the CI did not overlap unity. We used a time
Accepted for publication November 27, 2002. Supported by grants from the Lion’s Cancer Research Foundation, Umeå University. 1345
1346
MALE BREAST CANCER RISK IN PATIENTS WITH PROSTATE CANCER
cutoff of 6 months because we did not primarily sort out autopsy cases, which would exclude second primary cancers diagnosed at autopsy with the prostate cancer and in the diagnostic evaluation at the first cancer diagnosis. A cutoff of 6 months also would possibly identify clinically significant second primary cancers.
tinal tract cancers (except small intestine) and smoking related cancers displayed decreased or no difference in risk. Brain tumors expressed no increased risk after the first 6 months of followup but when analyzing brain tumors (gliomas and meningiomas) separately a small increase in risk of meningiomas was noted.
RESULTS
DISCUSSION
There were 10,526 second primary cancers observed among 135,713 patients with prostate cancer, of which 8,984 were expected, for a modest overall increased risk (standardized incidence ratio 1.17 95% CI 1.15–1.19) of second primary cancers. However, when looking at second primary cancers diagnosed after the initial 6 months following prostate cancer diagnosis, there was a decreased risk (standardized incidence ratio 0.93 95% CI 0.91– 0.94), which was more pronounced at older ages (table 1). For the 2 youngest cohorts (age less than 60 and 60 to 64 years) risk was not significantly decreased. A significantly increased risk (standardized incidence ratio 2.01 95% CI 1.44 –2.74) of male breast cancer after prostate cancer was observed in 40 cases (table 2). The increased risk was higher for the first 6 months after prostate cancer diagnosis (nonsignificant) but persisted later when most of the tumors were diagnosed. A younger age at prostate cancer diagnosis is (less than 60 years) increased the risk (standardized incidence ratio 5.29, 95% CI 1.06 –15.47) of subsequent breast cancer, and the risk continued to be increased significantly for patients older than 70 (1.85, 1.18 – 2.75) (table 3). From 1958 to 1980 the risk of male breast cancer was somewhat higher than that in the following period studied (standardized incidence ratio 2.12, 95% CI 1.35–3.19 versus 1.48, 0.77–2.59). An increased risk of tumors developing in the small intestine (standardized incidence ratio 2.20, 95% CI 1.83–2.61), mostly carcinoids, was observed. Carcinoids make up about 25% to 50% of all tumors in the small bowel. In this study the second primary cancer in the small bowel consisted of 70.5% carcinoids, 18.6% adenocarcinomas and 10.9% various histopathologies. The risk was higher the earlier in life the patient was diagnosed with prostate cancer but persisted across the age cohorts (table 3). After 6 months of followup there was an increased risk of malignant melanomas and endocrine tumors (parathyroid, thymus, pituitary, adrenal gland and endocrine pancreas). Among endocrine tumors parathyroid tumors were the most common (46.8% of all observed cases). Melanoma showed only a slight change in risk with respect to early or late diagnosis after prostate cancer, and its risk decreased 10 years after prostate cancer diagnosis regardless of age at diagnosis (table 4). A high increased risk was present for kidney (standardized incidence ratio 7.06, 95% CI 6.71–7.90) and bladder cancer (7.29, 6.71–7.90) for the first 6 months of followup and then disappeared in the longer followup. The exception to this pattern was in young men (age less than 60 at prostate cancer diagnosis) having an increased risk of bladder cancer even after the first 6 months. Gastrointes-
Our study revealed a new and interesting finding of an increased risk for male breast cancer. A possible explanation might be that these tumors metastasize from prostate cancer and are mistakenly diagnosed as mammary gland tumors. Recently it has been shown that prostate specific antigen is present in breast tissue and, if this is not recognized earlier, it can lead to misclassifications and fewer diagnoses of breast cancer.8 A more plausible explanation may be that hormonal treatment of the prostate cancer with estrogens induced the breast cancer.9 This explanation might be supported by the fact that the risk of breast cancer was higher in the early period of our study from 1958 to 1980 when estrogens were used more often in the treatment of prostate cancer. Similarly case reports of breast cancer and the Reifenstein syndrome have been reported.10 The Reifenstein syndrome is characterized by infertility, testicular feminization, frequent hypospadias and other defects. These men have germline mutations in the androgen receptor that cause androgen insufficiency and, therefore, estrogen excess. Another genetic condition is Klinefelter’s syndrome, associated with an increased risk of male breast cancer but not with prostate cancer.11 Genetic factors may explain the association of breast and prostate cancer because several studies have reported an increased risk of male and female breast cancer in families with multiple cases of prostate cancer.12 However, Olsson et al reported a decreased risk of prostate cancer in relatives of men with breast cancer.13 Mutations in the BRCA2 gene have been associated with prostate cancer and male breast cancer. In a population based study of 34 male breast cancer cases from southern Sweden mutations in the BRCA2 gene were found in approximately 20%.14 In families with breast cancer from Iceland with the BRCA2 founder mutation 995 del5 a clustering of prostate cancer and male breast cancer is present. Men carrying the mutation seem to have aggressive disease.15 The Breast Cancer Linkage Consortium found an RR of 4.65 (95% CI 3.48 – 6.22) for prostate cancer in BRCA2 mutation carriers, with lifetime risk estimations of 19.8% for prostate cancer and 6% for male breast cancer by carriers age 70 years.16 The incidence of male breast cancer in Sweden is similar to that of the rest of Europe but lower than that of black and Jewish populations. The reported high risk of kidney and bladder cancer shown in some studies4, 5 could not be verified in our study except in the first 6 months. Most bladder cancers were diagnosed on the same day or the day after prostate cancer diagnosis (data not shown). The diagnosis of prostate and bladder cancer is likely due to autopsy cases or cases diagnosed in the routine diagnostic evaluation, cystoscopy or urography. A few studies
TABLE 1. Differences in risk of second primary cancer depending on age at prostate cancer diagnosis 0–5 Mos. After Diagnosis Age (yrs.)
No. Cases
Standardized Incidence Ratio (95% CI)
Less than 60 60–64 65–69 70–79 Greater than 80
67 146 354 1,335 1,104
All ages
3,006
6 or More Mos. After Diagnosis
All Cases
No. Cases
Standardized Incidence Ratio (95% CI)
No. Cases
Standardized Incidence Ratio (95% CI)
6.12 (4.74–7.78) 4.12 (3.48–4.85) 3.77 (3.39–4.18) 3.05 (2.89–3.22) 3.75 (3.53–3.98)
295 691 1,328 3,810 1,396
1.07 (0.94–1.20) 1.01 (0.94–1.09) 0.95 (0.91–1.01) 0.92 (0.89–0.95) 0.85 (0.81–0.90)
362 837 1,682 5,145 2,500
1.26 (1.13–1.40) 1.21 (1.13–1.29) 1.14 (1.08–1.19) 1.12 (1.10–1.16) 1.30 (1.25–1.35)
3.45 (3.32–3.57)
7,520
0.93 (0.91–0.95)
10,526
1.17 (1.15–1.19)
1347
MALE BREAST CANCER RISK IN PATIENTS WITH PROSTATE CANCER TABLE 2. Sites of second primary cancers diagnosed after prostate cancer diagnosis 0–5 Mos. After Diagnosis ICD7 Ca Site (No. cases)
6 or More Mos. After Diagnosis
All Cases
Standardized Incidence Ratio Standardized Incidence Ratio Standardized Incidence Ratio No. Cases No. Cases No. Cases (95% CI) (95% CI) (95% CI)
Head/neck (140/148) Esophagus (150) Stomach (151) Small intestine (152) Colon (153) Rectum (154) Liver (155) Pancreas (157) Lung (162) Male breast (170) Kidney (180) Bladder (181) Melanoma (190) Brain (193) Endocrine (195) Hematological mal. (200–5)
34 29 223 55 308 133 149 152 371 5 279 592 24 114 38 250
1.26 (0.87–1.76) 2.01 (1.35–2.89) 2.54 (2.22–2.89) 9.76 (7.35–12.71) 3.22 (2.87–3.60) 2.18 (1.83–2.59) 4.42 (3.74–5.19) 3.72 (3.16–4.37) 3.49 (3.14–3.86) 2.63 (0.85–6.14) 7.06 (6.71–7.90) 7.29 (6.71–7.90) 1.32 (0.85–1.97) 6.20 (5.11–7.44) 5.37 (3.80–7.37) 3.15 (2.77–3.57)
207 98 667 74 945 517 279 321 730 35 316 768 225 175 88 671
0.86 (0.74–0.98) 0.76 (0.62–0.93) 0.87 (0.80–0.94) 1.39 (1.09–1.75) 1.03 (0.97–1.10) 0.92 (0.84–1.00) 0.86 (0.77–0.97) 0.87 (0.77–0.97) 0.78 (0.72–0.84) 1.95 (1.36–2.71) 0.90 (0.80–1.00) 1.00 (0.93–1.07) 1.33 (1.16–1.51) 1.09 (0.94–1.26) 1.41 (1.13–1.74) 0.91 (0.84–0.98)
241 127 890 129 1,253 650 428 473 1,101 40 595 1,360 249 289 126 921
0.90 (0.79–1.02) 0.89 (0.74–1.06) 1.04 (0.97–1.11) 2.20 (1.83–2.61) 1.24 (1.17–1.31) 1.04 (0.96–1.12) 1.20 (1.09–1.32) 1.15 (1.05–1.26) 1.05 (0.99–1.12) 2.01 (1.44–2.74) 1.52 (1.40–1.65) 1.59 (1.51–1.68) 1.33 (1.17–1.50) 1.62 (1.44–1.81) 1.81 (1.51–2.16) 1.13 (1.05–1.20)
TABLE 3. Prostate cancer diagnoses at different patient ages and influence on risk of second primary cancers Age Less Than 60 ICD7 Ca Site (No. cases)
Age 60–69
Age Greater Than 70
Standardized Incidence Ratio Standardized Incidence Ratio Standardized Incidence Ratio No. Cases No. Cases No. Cases (95% CI) (95% CI) (95% CI)
Small intestine (152) 5 2.62 (0.84–6.11) Colon (153) 18 0.72 (0.43–1.14) Rectum (154) 14 0.77 (0.42–1.29) Male breast (170) 3 5.29 (1.06–15.47) Kidney (180) 14 0.94 (0.51–1.58) Bladder (181) 47 1.77 (1.30–2.36) Melanoma (190) 12 1.26 (0.65–2.20) Endocrine (195) 3 0.80 (0.16–2.33) Second primary cancer diagnosed after the first 6 months only.
24 203 126 8 90 233 71 28
1.74 (1.11–2.58) 0.97 (0.84–1.11) 0.88 (0.74–1.05) 1.81 (0.78–3.57) 0.91 (0.75–1.11) 1.15 (1.00–1.30) 1.38 (1.08–1.74) 1.39 (0.92–2.01)
45 724 377 24 121 488 142 57
1.21 (0.88–1.61) 1.06 (0.99–1.14) 0.94 (0.84–1.04) 1.85 (1.18–2.75) 0.89 (0.78–1.12) 0.90 (0.82–0.98) 1.31 (1.10–1.54) 1.48 (1.12–1.92)
TABLE 4. Risk of second primary cancer after prostate cancer according to time after prostate cancer diagnosis 0.5–5 Yrs. Ca. Site
Small intestine Colon Rectum Male breast Kidney Bladder Melanoma Brain Endocrine
No. Cases
Standardized Incidence Ratio (95% CI)
48 579 307 25 198 498 142 103 64
1.51 (1.11–2.00) 1.07 (0.98–1.16) 0.91 (0.81–1.01) 2.18 (1.41–3.22) 0.91 (0.79–1.05) 1.08 (0.99–1.18) 1.39 (1.17–1.64) 1.03 (0.84–1.25) 1.64 (1.26–2.09)
5–10 Yrs.
10 or More Yrs.
No. Cases
Standardized Incidence Ratio (95% CI)
No. Cases
Standardized Incidence Ratio (95% CI)
19 248 155 7 85 178 63 44 13
1.39 (0.84–2.17) 1.05 (0.93–1.19) 1.08 (0.92–1.27) 1.47 (0.59–3.03) 0.96 (0.77–1.19) 0.90 (0.77–1.04) 1.47 (1.13–1.88) 1.10 (0.80–1.48) 0.82 (0.44–1.41)
7 118 55 3 33 92 20 28 11
0.90 (0.36–1.85) 0.84 (0.70–1.01) 0.67 (0.51–0.88) 1.11 (0.22–3.25) 0.70 (0.48–0.98) 0.81 (0.65–0.99) 0.83 (0.51–1.28) 1.34 (0.89–1.94) 1.38 (0.69–2.46)
address the question of whether full dose radiotherapy of the prostate gland and pelvic lymphatic nodules for prostate cancer induces second primary cancer in adjacent tissues. Brenner et al showed that in a large number of patients with prostate cancer treated with surgery or radiotherapy there was a significant increase in risk of solid tumors and infield sarcomas in the radiotherapy group.17 In addition, there was an increased risk of bladder cancer among men treated with radiation. In our study the increased risk of bladder cancer occurred in the early period after prostate cancer diagnosis and, thus, could not be due to radiation treatment. No increased risk was present for sarcomas. However, no information on primary treatment is available in the Swedish Cancer Registry. Malignant skin melanomas show an increased risk and persist for years after prostate cancer diagnosis. Greenberg et al also reported an increased risk of melanoma following prostate cancer.5 This finding may be due to hormonal influence as suggested by Levi et al in their study of second primary cancers following cutaneous malignant melanoma.18 They found a standardized incidence ratio of 2.1 (95% CI
1.2–3.4) for prostate cancer as second primary but their study was based on few cases. The finding of an increased risk of small bowel cancer after prostate cancer is intriguing but has been reported previously. A survey of multiple primaries following cancer of the digestive system in Connecticut revealed an increase in risk of 3.1 for prostate cancer after small intestine tumor diagnosis.19 However, no histology was reported in this study. An increased risk of prostate cancer with high levels of insulinlike growth factor-1 has been reported,20 and the strong influence of insulin-like growth factor-1 on the growth of carcinoids could explain this association. A third of the entire small bowel tumors in our study were diagnosed at autopsy, and more than half of the cases had carcinoid histology. These numbers are reflected by the high standardized incidence ratio of 9.76 (95% CI 7.35–12.71) in the first 6 months of followup. According to the Data Swedish Cancer Registry approximately 25% of cancers of the small intestine were generally diagnosed at an autopsy. As in other studies2– 4 there was a decreased risk of subsequent malignancies that were more pronounced in older pa-
1348
MALE BREAST CANCER RISK IN PATIENTS WITH PROSTATE CANCER TABLE 5. Cohort studies of second primary cancers following prostate cancer References/Study Base
No. Pts.
Standardized Incidence Ratio (95% CI)
Kleinerman et al/Connecticut Tumor Registry2 Osterlind et al/Danish Tumour Registry3 Greenberg et al/Georgia Center for Cancer Statistics5
18,135
0.85 (0.80–0.90)
Lung Ca (0.7), digestive system (0.8)
19,886
0.51 (0.47–0.55)
McCredie et al/New South Wales Central Cancer Registry, Australia4 Pawlish et al/Metropolitan Detroit Cancer Surveillance System6 Present study/Swedish Cancer Registry
23,067
0.79 (0.75–0.84)
Respiratory Ca (0.5, 0.4–0.6), digestive system (0.5, 0.5–0.6), skin melanoma (1.8, nonsignificant) Bladder (2.2, p ⬍0.01) risk confined to first 3 yrs., kidney (3.3, p ⬍0.01) for year 1 only, skin melanoma (2.5, p ⬍0.05) Kidney (1.3, 1.0–1.7), lung (0.6, 0.5–0.7), gastric (0.7, 0.6– 0.9), pancreas (0.6, 0.4–0.9)
9,794
1.01 (0.95–1.07)
135,713
1.16 (1.12–1.19) 0.92 (0.90–0.94)
3,675
1.2 (1.1–1.4)
tients. This decrease suggests a tendency to overlook symptoms that might be due to a new malignancy in the elderly, and others have suggested such surveillance bias as an explanation. Most of the previous studies, including ours, have revealed decreased risk of smoking related cancers such as lung, and head and neck tumors.2– 4 In our study several uncommon cancers (male breast cancer, small bowel tumors and endocrine tumors) were found at a higher frequency than expected. A reason that these uncommon cancers have not been reported may relate to the small populations of the previous studies. At least 6 times as many patients as the largest of the previous studies were included in our study (table 5). Despite that fact, however, the number of male breast cancer cases was small. Using the Swedish Cancer Registry, which consists of a well-defined population, made it possible to produce high quality data on second primary cancers. In addition, almost complete followup should provide a clear indication of the risks, and there is a low percentage of misclassifications in the registry as demonstrated by Frodin et al.7 A problem associated with our study is the fact that no data regarding primary treatment or disease stage are reported to the registry. In addition, performing a study this large may generate random associations by multiple testing. Coincidental discovery of a second primary cancer at autopsy might not have clinical importance but could be noteworthy when explaining common etiologies. Different time cutoffs have been used in other studies when calculating risks, and there are no data that indicate 1 or 6 months as superior. However, we believe that 6 months is sufficient to complete diagnostic evaluation of prostate cancer.
Bladder (1.57, 1.34–1.83) significant for first 2 months only, lung (0.83, 0.74–0.93), male breast (1.66, 0.54–3.88) All second primary Ca together Followup from 6 months after diagnosis of prostate Ca male breast cancer (1.95, 1.36–2.71), small intestine (1.39, 1.09– 1.75), endocrine tumors (1.41, 1.13–1.74)
5. 6.
7. 8.
9. 10.
11. 12.
13.
14.
CONCLUSIONS
We report significantly increased risk of male breast cancer following prostate cancer and a decreased overall risk for second primary cancer. Several uncommon endocrine cancers and skin melanomas also showed significantly increased risk of occurrence after prostate cancer. These findings warrant further detailed investigations.
15.
REFERENCES
18.
1. Steinberg, G. D., Carter, B. S., Beaty, T. H., Childs, B. and Walsh, P. C.: Family history and the risk of prostate cancer. Prostate, 17: 337, 1990 2. Kleinerman, R. A., Liebermann, J. V. and Li, F. P.: Second cancer following cancer of the male genital system in Connecticut, 1935– 82. Natl Cancer Inst Monogr, 68: 139, 1985 3. Osterlind, A., Rorth, M. and Prener, A.: Second cancer following cancer of the male genital system in Denmark, 1943– 80. Natl Cancer Inst Monogr, 68: 341, 1985 4. McCredie, M., Macfarlane, G. J., Stewart, J. and Coates, M.: Second primary cancers following cancers of the kidney and
Comments (Standardized Incidence Ratio, 95% CI)
16. 17.
19.
20.
prostate in New South Wales (Australia), 1972–91. Cancer Causes Control, 7: 337, 1996 Greenberg, R. S., Rustin, E. D. and Clark, W. S.: Risk of genitourinary malignancies after cancer of the prostate. Cancer, 61: 396, 1988 Pawlish, K. S., Schottenfeld, D., Severson, R. and Montie, J. E.: Risk of multiple primary cancers in prostate cancer patients in the Detroit metropolitan area: a retrospective cohort study. Prostate, 33: 75, 1997 Frodin, J. E., Ericsson, J. and Barlow, L.: Multiple primary malignant tumors in a national cancer registry—reliability of reporting. Acta Oncol, 36: 465, 1997 Gupta, R. K.: Immunoreactivity of prostate-specific antigen in male breast carcinomas: two examples of a diagnostic pitfall in discriminating a primary breast cancer from metastatic prostate carcinoma. Diagn Cytopathol, 21: 167, 1999 Schlappack, O. K., Braun, O. and Maier, U.: Report of two cases of male breast cancer after prolonged estrogen treatment for prostatic carcinoma. Cancer Detect Prev, 9: 319, 1986 Wooster, R., Mangion, J., Eeles, R., Smith, S., Dowsett, M., Averill, D. et al: A germline mutation in the androgen receptor gene in two brothers with breast cancer and Reifenstein syndrome. Nat Genet, 2: 132, 1992 Ravandi-Kashani, F. and Haynes, T. G.: Male breast cancer: a review of the literature. Eur J Cancer, 34: 1341, 1998 Valeri, A., Fournier, G., Morin, V., Morin, J. F., Drelon, E., Mangin, P. et al: Early onset and familial predisposition to prostate cancer significantly enhance the probability for breast cancer in first degree relatives. Int J Cancer, 86: 883, 2000 Olsson, H., Andersson, H., Johansson, O., Moller, T. R., Kristoffersson, U. and Wenngren, E.: Population-based cohort investigations of the risk for malignant tumors in first-degree relatives and wives of men with breast cancer. Cancer, 71: 1273, 1993 Haraldsson, K., Loman, N., Zhang, Q. X., Johannsson, O., Olsson, H. and Borg, A.: BRCA2 germ-line mutations are frequent in male breast cancer patients without a family history of the disease. Cancer Res, 58: 1367, 1998 Sigurdsson, S., Thorlacius, S., Tomasson, J., Tryggvadottir, L., Benediktsdottir, K., Eyfjord, J. E. et al: BRCA2 mutation in Icelandic prostate cancer patients. J Mol Med, 75: 758, 1997 Cancer risks in BRCA2 mutation carriers. The Breast Cancer Linkage Consortium. J Natl Cancer Inst, 91: 1310, 1999 Brenner, D. J., Curtis, R. E., Hall, E. J. and Ron, E.: Second malignancies in prostate carcinoma patients after radiotherapy compared with surgery. Cancer, 88: 398, 2000 Levi, F., La Vecchia, C., Randimbison, L., Te, V. C. and Erler, G.: Incidence of invasive cancers following cutaneous malignant melanoma. Int J Cancer, 72: 776, 1997 Hoar, S. K., Wilson, J., Blot, W. J., McLaughlin, J. K., Winn, D. M. and Kantor, A. F.: Second cancer following cancer of the digestive system in Connecticut, 1935– 82. Natl Cancer Inst Monogr, 68: 49, 1985 Stattin, P., Bylund, A., Rinaldi, S., Biessy, C., Dechaud, H., Stenman, U. H. et al: Plasma insulin-like growth factor-I, insulin-like growth factor-binding proteins, and prostate cancer risk: a prospective study. J Natl Cancer Inst, 92: 1910, 2000