Family history of urogenital cancers in patients with bladder, renal cell and prostate cancers
Article first published online: 27 AUG 2007
Copyright © 2007 Wiley-Liss, Inc.
International Journal of Cancer
Volume 121, Issue 12, pages 2748–2752, 15 December 2007
How to Cite
Randi, G., Pelucchi, C., Negri, E., Talamini, R., Galeone, C., Franceschi, S. and La Vecchia, C. (2007), Family history of urogenital cancers in patients with bladder, renal cell and prostate cancers. Int. J. Cancer, 121: 2748–2752. doi: 10.1002/ijc.23037
- Issue published online: 22 OCT 2007
- Article first published online: 27 AUG 2007
- Manuscript Accepted: 29 MAY 2007
- Manuscript Received: 30 JAN 2007
- Italian Foundation for Cancer Research
- urogenital cancer;
- family history;
- case-control study
Family history of urogenital cancers has been associated with an increased risk of cancer at the same sites. To evaluate the risks of bladder, prostate and renal cell neoplasms for family history of urogenital cancers in first-degree relatives, we analysed data from a case-control study conducted in northern Italy between 1985 and 1992 and including a total of 1,356 incident, histologically confirmed cancer cases: 727 bladder, 348 renal cell and 281 prostate cancer cases. Controls were 1,067 patients admitted to the same network of hospitals for acute, nonneoplastic conditions. Unconditional logistic regression was used to estimate odds ratios (ORs) of urogenital cancers, adjusting for age, sex, study centre, education, body mass index, smoking habit, alcohol consumption and number of brothers and, where appropriate, sisters. ORs of urogenital cancers for subjects with at least 1 first-degree relative with cancer at the same site were 6.1 [95% confidence interval (CI), 2.3–16.6] for bladder, 2.0 (95% CI: 0.6–6.2) for renal cell and 2.0 (95% CI: 0.9–4.1) for prostate cancer. © 2007 Wiley-Liss, Inc.
The risk of bladder cancer is increased by about 50–100% in first degree relatives of bladder cancer cases,4 and it appears higher for patients diagnosed before 60 years of age.5, 6 In the Swedish Family Cancer Database, a linkage study including 2,105 bladder cancer cases, the risk of bladder cancer was higher in siblings rather than in offspring of bladder cancer cases,7 but other studies did not confirm this result.1, 5
Several case-control studies found that a family history of kidney cancer was a risk factor for RCC.3, 8, 9, 10 In the nationwide Swedish Family Cancer Database, including 23,137 kidney cancer cases, the standardized incidence ratio was 1.6 for family history of the same cancer in offspring and 4.7 in siblings.11 Two other linkage studies also found increased risk for family history of kidney cancer.6, 12 Inherited genetic factors were estimated to account for 7% of bladder and 8% of kidney cancers.11
A study on 44,788 pairs of twins conducted in Scandinavian countries estimated that heritable factors are more important in prostate than in any other cancer.13 Familial history of prostate cancer has been associated with prostate cancer risk in 2 meta-analyses [odds ratios (ORs) = 2.2 and 2.5], and risk appeared to be higher if the affected relative was a brother, rather than the father.14, 15
With the aim of estimating relative and attributable risks for familial history of cancers of the urogenital tract, we analysed data from a case-control study on bladder, renal cell and prostate cancers conducted in northern Italy.
Material and methods
A total of 1,356 cases were included in the study: 727 subjects with invasive, transitional cell bladder cancer (617 males, 110 females), 348 subjects with RCC (236 males, 112 females) and 281 men with prostate cancer. All cases were aged between 25 and 79 years and had incident and histologically confirmed cancers. The median age was 63 years for bladder cancer, 60 years for RCC, and 67 years for prostate cancer. Cases were recruited in a network of general hospitals and university clinics in the areas under surveillance.
Controls were 1,067 subjects (769 males, 298 females), aged between 25 and 79 years (median age 60 years), admitted to the same network of hospitals as cases for acute, nonneoplastic, nonurological or genital tract diseases (23% were admitted for surgical conditions, 29% for fractures and other traumatic conditions, 18% for nontraumatic orthopaedic disorders, and 30% for miscellaneous other illnesses such as skin, eyes, ears, nose and throat conditions). For the analysis on prostate cancer, females were excluded from the control group. Trained interviewers identified and questioned cases and controls during their hospital stay. Overall, less than 3% of both cases and controls approached refused the interview.
The structured questionnaire included socio-demographic questions, detailed information on smoking habits (including smoking status, total duration of the habit, average number of cigarettes/day and the cigarette brand most commonly smoked), alcohol drinking (separately for wine, beer and spirits), coffee and tea consumption, frequency of use of a few selected foods, history of selected diseases, reproductive and sexual habits, history of relevant occupational exposures and modalities of diagnosis. Information on family history of urogenital cancers included parents' and siblings' histories of bladder, renal cell and prostate cancers. It was asked whether a parent had been affected or not, the total number of brothers and sisters and the number of affected siblings, if any. No verification of the cancer diagnoses in the relatives was performed.
ORs and the corresponding 95% confidence intervals (CI) of bladder, renal cell and prostate cancers, according to family history of urogenital cancers, were obtained using unconditional multiple logistic regression22 in order to account simultaneously for the effect of many potential confounders. We included in the models factors which were found to be associated with the cancers considered in this study, plus measures of family size. The regression equations included terms for age, sex, study centre, education, body mass index, smoking habit, alcohol consumption and number of brothers and, where appropriate, sisters. For bladder cancer, a further term for occupational exposure to known or suspected carcinogens (i.e., chemical industry, dyestuff, painting, pharmaceutical, coal and gas4) was included in the models. Population attributable risks were estimated according to Bruzzi and coworkers.23, 24
Table I shows the distribution of bladder, renal cell and prostate cancer cases, and of the control groups, according to sex, age and smoking habit. RCC cases and controls had similar age distributions, while bladder and prostate cancer cases were generally older than controls. Table II reports the distribution of cases and controls according to family history of urogenital cancers, with ORs and their 95% CI. A family history of cancer at the same sites was reported by 2.8% of bladder, 1.4% of renal cell and 5.4% of prostate cancer cases. The percentage of controls reporting a family history was 0.6% for bladder, 0.8% for renal cell and 2.8% for prostate (male controls only). There was a direct relation between having a family history of cancer and the risk of the same cancer, with ORs of 6.1 (95% CI: 2.3–16.6) for bladder cancer, 2.0 (95% CI: 0.6–6.2) for RCC and 2.0 (95% CI: 0.9–4.1) for prostate cancer. A family history of bladder cancer also appeared to increase the risk of prostate cancer (OR = 2.1; 95% CI: 0.5–9.5) and RCC (OR = 1.6; 95% CI: 0.4–6.9).
|No. cases (%)||No. controls (%)|
|Bladder cancer||Renal cell cancer||Prostate cancer||All||Male|
|Male||617 (84.9)||236 (67.8)||281 (100.0)||769 (72.1)||769 (100.0)|
|Female||110 (15.1)||112 (32.2)||–||298 (27.9)||–|
|<55||99 (13.6)||108 (31.0)||17 (6.0)||332 (31.2)||232 (30.2)|
|55–59||125 (17.2)||65 (18.7)||39 (13.9)||171 (16.0)||131 (17.0)|
|60–64||173 (23.8)||82 (23.6)||54 (19.2)||209 (19.6)||159 (20.7)|
|65–69||178 (24.5)||54 (15.5)||62 (22.1)||183 (17.1)||136 (17.7)|
|≥70||152 (20.9)||39 (11.2)||109 (38.8)||172 (16.1)||111 (14.4)|
|Never||123 (16.9)||138 (39.7)||69 (24.6)||394 (36.9)||169 (22.0)|
|Current/former||604 (83.1)||210 (60.3)||212 (75.4)||673 (63.1)||600 (78.0)|
|Cancer site in first degree relatives||Subjects with family history (%)||OR (95% CI)1|
|Cancer cases||Controls||Bladder||Renal cell||Prostate|
|Bladder||20 (2.8)||3 (0.9)||4 (1.4)||6 (0.6)||4 (0.5)||6.1 (2.3–16.6)2||1.6 (0.4–6.9)||2.1 (0.5–9.5)|
|Renal cell||6 (0.8)||5 (1.4)||3 (1.1)||8 (0.8)||7 (0.9)||1.4 (0.4–4.2)||2.0 (0.6–6.2)||1.1 (0.2–4.6)|
|Prostate||18 (2.5)||9 (2.6)||15 (5.4)||27 (2.5)||21 (2.8)||1.0 (0.5–1.9)||1.1 (0.5–2.3)||2.0 (0.9–4.1)|
Table III gives the distribution of subjects with a family history of cancer and ORs for bladder and prostate cancers in strata of age and smoking habit. The OR of bladder cancer was 7.6 (95% CI: 2.2–26.4) in younger subjects (<65 years) and 4.9 (95% CI: 1.0–24.7) in older subjects (≥65 years). Corresponding values for prostate cancer were 2.7 (95% CI: 0.8–8.7) and 1.5 (95% CI: 0.6–4.0). With respect to smoking status, the OR of bladder cancer was 4.8 (95% CI: 0.9–25.7) among never smokers [3 cases (2.4%) and 4 controls (1.0%)], and was 10.7 (95% CI: 2.4–48.9) among current or former smokers [17 cases (2.8%) and 2 controls (0.3%)]. Table IV shows the distribution of subjects with a family history of cancer and ORs for bladder and prostate cancers according to the type of relative affected. For bladder cancer, the OR was 6.4 (95% CI: 1.8–22.3) when a parent was affected, and 5.2 (95% CI: 1.0–25.9) when a sibling was affected. Corresponding values for prostate cancer were 1.6 (95% CI: 0.7–3.7) and 4.2 (95% CI: 0.8–22.2).
|Strata||Bladder cancer||Prostate cancer|
|No. cases:controls||OR (95% CI)1||No. cases:controls||OR (95% CI)1|
|<65||12:4||7.6 (2.2–26.4)2||5:12||2.7 (0.8–8.7)|
|≥65||8:2||4.9 (1.0–24.7)3||10:9||1.5 (0.6–4.0)|
|Never||3:4||4.8 (0.9–25.7)||3:4||1.9 (0.4–10.0)|
|Current/former||17:2||10.7 (2.4–48.9)2||12:17||1.8 (0.8–4.3)|
|Bladder cancer||Prostate cancer|
|No. cases: controls||OR (95% CI)1||No. cases: controls||OR (95% CI)1|
|Parent||12:4||6.4 (1.8–22.3)2||9:19||1.6 (0.7–3.7)|
|Sibling||8:2||5.2 (1.0–25.9)3||6:2||4.2 (0.8–22.2)|
Our study contributes to further quantify the direct association between family history of bladder, renal cell, and prostate cancers in first degree relatives and risk of cancer at the same sites. We found a strong association for bladder cancer, with a 6-fold increase in risk, while for RCC and prostate cancer, there was suggestive evidence that a family history of cancer is associated with cancer risk at these sites (doubled risk). Although not significant, we also found an increased risk of prostatic cancer in subjects who had relatives with bladder cancer, confirming the results of a larger Italian study which showed a 3.5 (95% CI: 1.6–7.4) elevated risk.2 However, none of the other associations between one cancer site and a different one was appreciably above unity. Although the number of subjects with family history was limited, these results are consistent with most studies on this issue.7, 11, 14, 15
Despite the large size of the study, the number of cases for each cancer site and more importantly the number of cases and controls with family history for selected cancers was limited. Consequently, the low statistical power for several associations is a major limitation of this study. The use of hospital controls has also been debated.22 However, we excluded from the control group subjects admitted for neoplastic conditions and for all conditions related to the known risk factors for prostate, renal cell and bladder cancers, including tobacco-related ones, and metabolic diseases related to overweight and obesity. In any case, hospital admission for controls is unlikely to be related to the same genetic aspects as familial prostate, renal cell or bladder cancers. Among the strengths of our study are the comparable catchment areas, and the almost complete participation of cases and controls. Moreover, the present study was able to disclose several of the recognized associations for the diseases considered. Thus, bladder cancer was related to tobacco25 and occupational factors26; RCC was related to tobacco27 and urogenital diseases20; and prostate cancer was related to socioeconomic status and sexual activity.28
Since family history was self-reported, it is possible that cases and controls gave inaccurate information, and that cancer cases tended to recall cancers in the family better than controls. A review of studies evaluating the accuracy and completeness of reporting of family history in first degree relatives of cancer patients and controls found satisfactory results for family history of breast, colon and prostate cancers, and less so for endometrial and ovarian cancers.29 The positive predictive value of self-reported family history (i.e. the proportion of subjects who actually had a family history among those who report it) was generally higher for cancer cases than controls across various cancer sites considered. In a study from Sweden30 analyzing the reliability of self-reported family history of cancer in a case-control study that was based on 1,508 cases of lymphoma and 1,229 control subjects, cases reported a family history of any cancer with similar sensitivity (0.85 vs. 0.8) and specificity (0.89 vs. 0.92) than control subjects. Younger men, better educated or first-born showed higher specificity in reporting family history, while women and subjects with a relative diagnosed 20 or more years before showed higher sensitivity, but the differences were marginal. An Italian study, based on the same population and study design of the present work, investigated the reproducibility of self-reported family history of digestive and respiratory tract cancers in first degree relatives, interviewing 294 controls in the hospital setting, and subsequently at home.31 The reproducibility of data on family history was satisfactory, although controls tended to report family history of cancer more frequently in hospital than at home. Thus, the common hospital setting for cases and controls in the present study, and hence the similar attention to medical history, may have improved the comparability of information.
Our findings of an elevated risk of bladder cancer and an increased risk for RCC and prostate cancer in subjects with a family history of cancer at the same site are in broad agreement with other reports, although our point estimate for bladder cancer is somewhat higher than in other studies.4 Because of the small number of subjects reporting a family history of bladder cancer (20 bladder cancer cases and 6 controls), risk estimates are subject to considerable random variation. Thus, the whole range of the CI should be considered, rather than the point estimate only. Possible explanations for the higher OR include, besides chance, a difference in the population studied or a more marked differential recall in this study.
In agreement with a case-control study from Texas,1 though not with all studies,5, 32 the association of bladder cancer risk with family history of bladder cancer was weaker in never smokers than in current or former smokers, supporting the hypothesis that environmental factors may modulate familial risk.1
The risk of bladder cancer was similar for having a parent or a sibling affected. Previous results were contrasting.1, 5, 7 For prostate cancer, the risk was appreciably higher when the affected relative was a brother rather than the father. Our results were in agreement with 2 meta-analyses, where ORs for history of prostate cancer in a brother were 2.9 and 3.4, and in the father were 2.1 and 2.5, respectively.14, 15 In another Italian case-control study,2 the OR was 6.6 (95% CI: 2.6–17) when the affected relative was a brother, and 3.1 (95% CI: 1.8–5.3) when the affected relative was the father. The higher risk of prostate cancer for men with an affected brother compared to those with an affected father in this and other studies suggests that a recessive or X-linked model of inheritance plays a role in familial prostate cancer.33, 34, 35, 36 However, this may also reflect an elder age at diagnosis for fathers or that the bias due to increased PSA testing may be more relevant in brothers of affected individuals.37 Male carriers of germ-line mutations of the genes BRCA1 and BRCA2, 2 genes predisposing to breast and ovarian cancers, have been found to be at high risk of prostate cancer.38, 39 Furthermore, another study suggested that the association of prostate cancer with BRCA2 may depend on the type of BRCA2 mutation and thus vary in different populations.40
For RCC, several autosomal dominant inherited syndromes have been described,41, 42 the most common being the von Hippel-Lindau syndrome, characterized by excesses of RCC and other neoplasms and caused by germ-line mutations in the von Hippel-Lindau tumor suppressor gene which is also involved in sporadic RCC.41, 42 These syndromes are rare and probably, as for many other cancers, most of the familial risk in older patients is not due to these highly penetrant genes.43 Other susceptibility genes may exist, with lower penetrance but much higher frequency in the population, which might account for more cases of RCC. Identification of these genes is extremely difficult because their low penetrance does not cause striking familial aggregations.43 The observation that recessive effects may be important in familial RCC, in contrast with von Hippel-Lindau and other identified dominant familial RCC syndromes, supports the existence of such lower penetrance susceptibility genes.44
Activation and deactivation of aromatic amines in humans are regulated by N-acetyltransferase (NAT) 1 and NAT2 genes. Slow acetylator NAT2 genotype has been associated with a modest increase of bladder cancer risk.4 Potential interaction between NAT2 gene and Glutathione S-transferase Mu1 (GSTM1) gene, which is responsible of an enzyme involved in the detoxification of a number of carcinogens, should be addressed in future investigations, also in relation to environmental factors influencing bladder cancer risk.4
As concerns multiple cases of urogenital cancers in the family, almost all subjects with family history of cancer reported at most one affected relative. Only 6 subjects reported multiple cancers in the family (less than 5% of subjects with familial history of cancer). Among cases, 1 subject with bladder cancer reported 2 affected siblings, 2 subjects with renal cell and bladder cancers reported 3 affected siblings and 1 subject with prostate cancer reported 4 affected siblings. Among controls, 2 subjects reported 1 sibling and 1 parent affected.
The estimates for population attributable risks for site-specific family history of cancer were 2% (95% CI: 1–4%) for bladder cancer, 1% (95% CI: 0–2%) for RCC and 3% (95% CI: 0–6%) for prostate cancer. Therefore, our estimates are lower than those from the Swedish Family-Cancer Database that were 4.6, 3.5 and 18.6%, respectively, especially for prostate cancer.45 Apart from chance, this difference between estimates may be attributed to differences between this Italian and the Swedish population, including the age distribution of cases and their relatives, family size, background risk of bladder, renal cell and prostate cancers and the frequency of susceptibility genes in the population.2
In any case, the population-attributable risk for family history was considerably lower than those for environmental factors in the same dataset. Thus, about 50% of bladder cancers were due to tobacco smoking, 4% to occupation, 12% to urinary tract infections and 16% to dietary indicators.25 For RCC, attributable risks were 25% for tobacco, 7% for urinary tract diseases and 17% for dietary indicators.46 Given the lack of strong associations, however, we were unable to provide estimates of population attributable risk for prostate cancer.
Italian Association for Cancer Research and the Italian League against Cancer. G. Randi and C. Galeone were supported by fellowships from the Italian Foundation for Cancer Research. The authors thank Ms. I. Garimoldi for editorial assistance.