You are on page 1of 4

Int. J. Cancer: 66,723-726 (1996) 0 1996 Wiley-Liss, Inc.

Publlcatlon 01 the Internat,onal Un on Agamst Cancer Plrolicatlon de I Lln on Internationale Conrrr le Cancer

INTERNATIONAL RENAL-CELL-CANCER STUDY. VI. THE ROLE OF MEDICAL AND FAMILY HISTORY
Brigitte SCHLEHOFER~,~", Wolfgang POMMER', Anders MELLEMGAARD3, John H. STEWART4, Margaret MCCREDIE~, Joseph K. MCLAUGHLIN~,~~ and Jurgen WAHRENDORF' Shelley NIWA~, Per LINDBLAD', Jack S. MANDEL~, 'Division of Epidemiology, German Cancer Research Centre, Heidelbeg, Germany; 2HumboldtHospital, Berlin, Germany; 3DanishCancer Registly, Danish Cancer Society, Copenhagen, Denmark; Western Clinical School, Universityof Sydney, Sydney, Australia; SCancerEpidemiology Research Unit, NSW Cancer Council, Sydney, Australia; hWestatInc., Rockville, MD, f Cnncer Epidemiology, UniversityHospital, Uppsala, Sweden; ?!khool o f Public Health, University USA; 7Departmento o f Minnesota, Minneapolis, MN, USA; and gBiostatistics Branch, National Cancer Institute, Bethesda, MD, USA.
A number of medical conditions have been linked with renal-cell cancer, although the evidence is not consistent in every case. In a large internationalcase-control study of renalcell cancer, we examined, among other hypotheses, associations with a personal history of certain medical conditions and a family history of cancer of the kidney or thyroid. Relative risks (RR), adjusted for the effects of age, gender, body-mass index, tobacco smoking and study centre, were significantly increased by a history of kidney stones or thyroid or kidney disease. The RR were not altered by additional adjustment for hypertension, or when diagnoses were restricted to those made at least 5 or 1 0 years before 1987 (the usual "cut-off'' date). The link with kidney injury is particularly likely to be affected by recall bias. Increased RR of borderline significance were found for kidney infection (RR, 1.2) and diabetes (RR, 1.4). Having one firstdegree relative with kidney cancer was associated with a significantly increased risk of renal-cell cancer (RR, I .6;95% CI, 1. U-2.4). Seven cases reported 2 first-degree relatives with kidney cancer. No controls had first-degree relatives with kidney cancer. None of our participants reported having von Hippel-Lindau disease. The data suggests that a few conditions of the kidney are strongly associated with renal-cellcancer and that heredity plays a role in a small proportionof cases.
SUBJECTS AND METHODS

o 1996 Wiley-Liss,Inc.

A number of medical conditions have been linked in analytical epidemiological studies with renal-cell cancer (RCC), although the evidence is not consistent in every case. Statistically significant associations have been reported for kidney stones (McLaughlin et al., 1984; Maclure and Willett, 1990; Talamini et al., 1990), kidney infection (McLaughlin et al., 1984), kidney cysts (McLaughlin et al., 1984), urinary-tract infection in women (Kreiger et al., 1993), urological disease (McCredie et al., 1988), hypertension (Raynor et al., 1981; McLaughlin et al., 1984, 1992; Fraser et al., 1990; Maclure and Willett, 1990), stroke (McLaughlin et al., 1984) and myocardial infarction (Wynder et al., 1974). Nonsignificant links have been reported with diabetes (McLaughlin et al., 1984, 1992; Asal et al., 1988; Kreiger et al., 1993; Adami et al., 1993), also possible associations with thyroid disease (Rosenberg et al., 1990) and thyroid cancer (Tucker et al., 1985). Case reports have identified additional, less common conditions which appear to carry an increased risk of RCC. These include autosomal dominant inherited disease such as von Hippel-Lindau syndrome (Hudson and Wilson, 1979; Lynch and Walzak, 1980) and polycystic kidney disease (Bernstein et al., 1987), end-stage renal failure requiring dialysis (Port et al., 1989), where the risk appears to be associated with acquired cystic disease (Ishikawa et al., 1990; Matson and Cohen, 1990), and a family history of kidney cancer (Franksson et al., 1972; Cohen et al., 1979). A large international case-control study of RCC was carried out to examine, among other hypotheses, the associations with a personal history of certain medical conditions and a family history of cancer of the kidney or thyroid.

A complete description of the population-based ascertainment of cases and controls has been presented elsewhere (McLaughlin et at., 199%). Briefly, cases with a histologically verified diagnosis of RCC during the period 1989-1991 were identified through population-based registries in 4 study areas (Sydney, Denmark, Uppsala and Minnesota), while cases in Berlin and Heidelberg were identified by monitoring all hospitals and pathology departments in the study areas. Detailed information about the histological cell types of RCC was not assessed systematically in all centres. Cases were aged between 20 and 79 years at diagnosis, except in Heidelberg where the upper age limit was 75 years. Controls, frequency matched by age and gender, were selected from the same study base as the cases. The sampling frames for controls were registers covering the entire population (Denmark, Uppsala), electoral rolls (Sydney), compulsory residential lists (Berlin, Heidelberg) or Health Care Financing Administration beneficiary lists (Minnesota; controls aged 65-79 years), while random-digit dialling was used for controls aged 20 to 64 years in Minnesota. In Sydney, Denmark, Minnesota and Uppsala, permission from the attending doctor was obtained before the cases were approached; the first contact in Berlin and Heidelberg was made by hospital physicians. The overall response rate was 72.3% for cases and 74.7% for controls. Non-participation was mainly due either to death before the interview could take place (cases) or to refusal (controls). A total of 1,774 cases and 2,359 controls participated; all had face-to-face interviews in their own home, except for the German patients, who were interviewed in hospital. Following extensive planning, coordination, revision, translation into English and back-translation into the original language, the questionnaire was largely identical for all centres. Information was sought about the following items: lifetime use of tobacco, diuretics, analgesics, anti-hypertensive drugs, diet pills, hormones and alcohol; height and weight, including changes in weight; physical activity; medical and reproductive histories; family history of cancer; occupation; diet; and demographic data. Subjects with missing information on possible confounder variables, namely height and weight (used to calculate body-mass index) and use of tobacco (42 cases, 50 controls) were excluded, leaving 1732 cases (1050

'OTo whom correspondence and reprint requests should be sent, at the Division of Epidemiology, German Cancer Research Centre, Im Neuenheimer Feld 280, D-69120 Heidelberg, Germany. Fax: (49) 6221-42-2203.
"Present address: International Epidemiology Institute, Rockville, Maryland, USA. Received: October 27,1995 and in revised form February 28,1996.

724

SCHLEHOFER ETAL.

men, 682 women) and 2309 controls (1429 men, 880 women) for the pooled analysis. A series of standardized questions asked whether the subjects had ever (and at what age) been told by a doctor that they had any of the following conditions: kidney stones, kidney infection, kidney injury, bladder stones, bladder infection, hypertension, diabetes, thyroid disease (previously linked with RCC); angina, heart attack or stroke (associated with hypertension, a postulated risk factor); or analgesic nephropathy or renal papillary necrosis (linked with consumption of phenacetin-containing analgesics, a possible risk factor). In Berlin, no information was obtained about kidney injury or thyroid disease. Analgesic nephropathy, renal papillary necrosis, polycystic kidney disease or other hereditary kidney diseases were mentioned in the checklist (Sydney, Uppsala and Minnesota) or specified by the subject after a screening question on other kidney diseases (Berlin, Denmark and Heidelberg). Only diagnoses made before 1987 were included in the pooled analysis. No further analyses on specific cell types of RCC could be done, due to missing information. No validation of self-reported diseases was performed. To evaluate the role of family history, subjects in Sydney, Denmark, Uppsala and Minnesota were asked whether they had a parent or sibling with kidney or thyroid cancer, or any other cancer (which was then specified). In the 2 German centres, a general question about cancer in close relatives preceded specification of the site. The total number of siblings (with or without cancer) was obtained in all centres except Berlin. Due to missing data, the analysis of family history of kidney cancer was based on 1589 cases and 2214 controls, while that relating to family history of thyroid cancer was based on 1594 cases and 2215 controls. The relative risk (RR) of RCC was estimated by the odds ratio, which was calculated with its 95% confidence interval (CI) by logistic regression (Breslow and Day, 1980). For each of the estimations, the RR was adjusted for centre, age (tertiles of combined cases and controls), body-mass index (quartiles of controls, calculated separately for men and women) and tobacco smoking (quartiles of pack years in controls, calculated separately for men and women), since these are known risk factors for renal-cell cancer. When the regression model included men and women, adjustment for gender was also made. A x2 test of heterogeneity was performed to detect differences between centers with respect to the variables of interest.
RESULTS

TABLE I - PERCENTAGE DISTRIBUTION OF CASES WITH RENAL-CELL


CANCER AND POPULATION-BASED CONTROLS BY DISEASE STATUS2 Men
Cases Women

Controls

Cases

Controls

Kidney stones infection iniurv3 BladdG stones infection Hypertension Diabetes Thyroid disease3

15.2 5.4 4.4 1.7 9.5 29.9 7.1 3.5

11.2 3.8 1.2 1.3 10.3 22.3 5.5 2.1

7.1 13.0 3.5 1.3 37.4 40.5 6.8 16.9

5.5 11.5 0.9 0.7 35.8 28.0 5.0 12.4

Data pooled from International Case-control Study in Australia (Sydney), Denmark, Germany (Berlin, Heidelberg), Sweden (Uppsala) and USA (Minnesota).-2Diagnosed by a doctor before 1987 according to self-report of subject; multiple assignmentswere p~ssible.-~Excludes Berlin. borderline significance were found for kidney infection (1.2) and diabetes (1.4). In the case of diabetes and thyroid disease, duration of disease did not alter the risk. Angina (RR, 1.2), heart attack (RR, 1.1) and stroke (RR, 0.7) were not associated with increased risk of RCC. Kidney injury appeared to be linked with RCC (RR, 3.2; 95% CI 1.9-5.5). An injury to the kidney was reported by 17% of cases and 4% of controls in Sydney, and approximately 2% of cases and less than 1%of controls in the other centres (Denmark, Heidelberg, Minnesota, Uppsala). Data on the following conditions were not obtained from all centres; for these, proportions are based only on subjects from the centres seeking the information. Polycystic disease was reported by 8 cases (0.7%) and 8 controls (0.5%), analgesic nephropathy by 3 cases only (0.3%), papillary necrosis by 1 case only (0.1%), other inherited kidney disease by 16 cases (1.4%) and 8 controls (0.5%) and other kidney disease by 52 cases (3.6%) and 36 controls (2.2%). No further specification of disease was possible except in the German centres, where 5 cases (1.1%)and 4 controls (0.6%) had acquired cystic disease (multiple or solitary).
Family history The majority of subjects reported no parents or siblings with a diagnosis of cancer of the kidney (type unspecified) or of the thyroid (Table 111). Having one first-degree relative with kidney cancer was associated with a significantly increased risk of RCC (RR 1.6; 95% CI 1.1-2.4). No controls reported having any first-degree relative with kidney cancer; but 7 cases reported 2 first-degree relatives with kidney cancer. Out of these, in 3 cases the mother and a sibling (in 2 cases a brother; the third was not specified) were affected with RCC, in one case the father and a sibling (not specified), in 2 further cases brother and sister, and in one case 2 sisters. The ages at diagnosis of these 7 cases ranged from 50 to 74 years (median 67 years). The adjusted risk of renal-cell cancer conferred by having any first-degree relative with kidney cancer did not differ significantly when subjects were subdivided into those less than 60 years of age (RR, 2.4; 95% CI 1.1-5.2) and those 60 years or more (RR, 1.7; 95% CI 1.1-2.6). The likelihood of RCC was not increased significantly by having a parent or sibling with thyroid cancer (Table 111). A variety of other cancers was reported among close relatives of cases and controls, but analysis was precluded by frequent lack of specificity in identification of the site of cancer.

Medical history The distribution of subjects reporting a diagnosis of each disease previously linked to RCC is given in Table I. No subject reported having von Hippel-Lindau disease. A higher proportion of cases than controls reported each of the kidney conditions, diabetes, or thyroid disease. Of the 3 specified conditions associated with hypertension, angina and heart attack were slightly more common among cases than controls (7.8% vs. 6.3% and 6.4% vs. 5.7% respectively) but the reverse was true for stroke (cases, 1.7% vs. controls, 2.4%). These apparent associations were investigated further by multivariate analysis with adjustment for centre, age, gender, body mass index and tobacco smoking (Table 11). The risk of RCC was increased significantly by a history of hypertension (RR 1.7; CI 1.4-1.9), an association which is explored in greater detail elsewhere (McLaughlin et a/., 1 9 9 5 ~ )Signifi. cantly raised RR were also found for kidney stones (1.3; CI 1.1-1.7) and thyroid disease (1.6; CI 1.3-2.2), estimates which were not altered by additional adjustment for hypertension or when diagnoses were restricted to those made at least 5 or 10 years before 1987 (the usual cut-off date). Increased RR of

RENAL-CELL CANCER, MEDICAL AND FAMILY HISTORY

725

TABLE 11 - RELATIVE RISK (RR) AND 95% CONFIDENCE INTERVALS (CI) OF RENAL-CELL CANCER ASSOCIATED WITH SELF-REPORTED DISEASE

Disease2

Cases

Controls

Men RR (95% CI)

x2 het3

Cases

Controls

Women RR (95% CI)

x2 het3

Both sexes RR (95% C1)

Kidney stones Kidney infection Kidney injury4 Bladder stones Bladder infection Hypertension Diibetes Thyroid disease4

160 57 40 18 99 313 75 32

48 88 21 9 78 25 1.4 1.0-1.9 1.7 11.0-3.0j 0.77 0.96 46 102 44 95

: 36 [ : : ? I ; : : {

0.24 0.47

1.4 (1.0-1.8) 1.6 (1.3-2.2

Data ooled from International Case-control Study in Australia (Sydney), Denmark, Germany (Berlin, Heidelberg), Sweden (UPPsala and USA (Minnesota).-2Adjusted for centre, age, gender (where appropriate), body-mass index, pack-years of tobacco; reference group includes subjects not affected with specific disea~e.-~x* test for heter~geneity-~Excludes Berlin.
TABLE 111 RELATIVE RISK (RR) AND 95% CONFIDENCE INTERVALS (CI) FOR RENAL-CELLCANCER ASSOCIATED WITH FAMILY HISTORY OF KIDNEY OR THYROID CANCER

Number of first-degree relatives with:

Men and women Cases Controls RR2 (95% C1)

Kidney cancer3 one two4 Thyroid cancer3 one two

56 7 22
0

50 0 17

1.6 5 -

(1.1-2.4)
6

1.6

(0.8-3.1)

Data pooled from International Case-control Study in Australia (Sydney), Denmark, Germany (Berlin, Heidelberg), Sweden (Uppsala) and USA (Minnesota).-2Adjusted for centre, age, gender, body-mass index, pack-years of tobacco; it was not possible to adjust for number of siblings, since not all centres asked this question; reference group includes subjects reporting no relative with specific cancer; information on kidney cancer was missing for 143 cases and 9.5 control^.-^^^ test for heterogeneity between centres: any first-degree relative with kidney cancer, men: p = 0.38; women:p = 0.15; any first-degreerelative with thyroid cancer, men: p = 0.60; women: p = 0.55.-4No subject reported more than 2 first-degree relatives with kidney ~ancer.-~Cannot be calculated.6Lr)werconfidence interval cannot be calculated in the full model.
DISCUSSION

This pooled analysis of data from parallel case-control studies in 6 centres indicates that there is likely to be some inherited predisposition to RCC in addition to those cases which are associated with certain rare hereditary conditions such as von Hippel-Lindau disease, which we could not identify in our participants. However, when the possibility of recall bias is taken into consideration, the RR of 1.6 conferred by having a first-degree relative with (unverified) kidney cancer suggests only a modest effect, which is probably less strong than for cancers of the breast or colon (Easton, 1994). Nevertheless, investigation of the genotype in informative families may shed further light on the still relatively obscure aetiology of RCC (Maher and Yates, 1991). In agreement with large clinical series (Zeier et al., 1988; Gabow and Bennett, 1991), this study does not support a link with autosomal dominant polycystic kidney disease, the most common inherited renal condition. However, a recent clinical study by Zbar et al. (1995) also supports the hypothesis that the predisposition to develop specific types of renal-cell carcinoma, namely papillary RCC, may also be inherited. Furthermore, in a detailed histopathological investigation, Bernstein et al. (1987) demonstrated that epithelial hyperplasia and micropolyp formation were characteristic of a number of inherited and acquired cystic diseases of the kidney, including von Hippel-Lindau and autosomal dominant polycystic disease, tuberous sclerosis, localized and acquired cystic disease, and that each of these conditions carried an increased risk of

neoplasia in the cysts. Thus, except perhaps in the case of von Hippel-Lindau disease, where the risk of carcinoma, often multiple, is very high, the evidence favours a proclivity to cancer in renal cystic disease irrespective of aetiology. Several more common kidney diseases have been associated with RCC. In this study a weak link was found with stones and possibly infection (when in the kidney but not when in the bladder) and a stronger link with hypertension. No plausible explanation readily suggests itself for the association with these conditions, which rarely cause pathology in the vicinity of the proximal renal tubules, the origin of the majority of renal adenocarcinomas. In essential hypertension and idiopathic hypercalciuria (the most common cause of urinary calculi in Western populations), subtle abnormalities of tubular ionic transport are present. In the case of calculi a shared association with high dietary animal protein (Maclure and Willett, 1990; Iguchi et aL, 1990; Trinchieri et al., 1991; Curhan et al., 1993) might provide the link with cancer. In this study, however, we found no unequivocal association between protein intake and risk of renal-cell cancer (Wolk et d.,1996). We are inclined to discount the other positive association found, that with trauma, as it may have been unduly influenced by recall bias. Our evidence supports previous findings of a likely association with two metabolic conditions, diabetes and thyroid disease. In respect of the latter, the possibility of cases having been exposed to radioactive iodine (Hoffman et al., 1982) has not been ruled out as the explanation for the increased risk. The association with diabetes remains after adjustment for 2 confounding factors, body-mass index and hypertension; nevertheless, it is still possible that all of these 3 conditions have a similar aetiological relationship to RCC. Recall bias is likely to have been greatest in relation to kidney injury and a family history of cancer, but would have operated to some extent in respect of the other kidney conditions. However, neither recall bias nor misclassification bias is likely to have been a factor in the positive findings seen with hypertension, diabetes and thyroid disease, which would not be associated in the lay mind with the kidney, and in which the diagnosis usually is straightforward and accompanied by treatment which will have aided recall. That hypertension, but not other common cardiovascular conditions associated with high blood pressure, was linked to RCC suggests that other factors, such as hypercholesterolaemia (Wynder et al., 1974), predisposing to arterial disease confer no risk for kidney cancer. In summary, there are few conditions of the kidney, other than cystic disease, that are strongly associated with renal adenocarcinoma. Heredity plays a role in a small proportion of cases, and metabolic factors that cannot be specified more precisely may explain the link between a number of systemic conditions and increased susceptibility to cancer of the kidney.

726

SCHLEHOFER ETAL.

REFERENCES

ADAMI, H.-O., MCLAUGHLIN, J., EKBOHM, A,, BERNE, C., SILVERMAN, occurrence, prevalence and renal cancers. Medicine, 69, 217-226 (1990). D., HACKER, D. and PERSSON, I., Cancer risk in patients with diabetes mellitus. Cancer Causes Control, 2,307-314 (1993). MCCREDIE, M., FORD, J.M. and STEWART, J.H., Risk factors for cancer ASAL, N.R., GEYER, J.R., RISSER, D.R., LEE, E.T., KADAMANI, S. and of the renal parenchyma. Int. J. Cancer, 42,13-16 (1988). CHERNG, N., Risk factors in renal-cell carcinoma. 11. Medical history, MCLAUGHLIN, J.K., CHOW, W-H., MANDEL, J.S., MELLEMGAARD, A., occupation, multivariate analysis and conclusions. Cancer Detect. MCCREDIE, M., LINDBLAD, P., SCHLEHOFER, B., POMMER, W., NIWA, S. Prevent., 13,263-279 (1988). and ADAMI, H.-O., International Renal-Cell-Cancer Study. VIII. Role BERNSTEIN, J., EVAN, A.P. and GARDNER, K.D., Epithelial hyperplasia of diuretics, other anti-hy ertensive medications and hypertension. in human polycystic kidney diseases. Its role in pathogenesis and risk Int. J. Cancer, 63,216-221 8995,). of neoplasia. Amer. J. PathoL, 129,92-101 (1987). MCLAUGHLIN, J.K., GAO, Y.-T., GAO, R.-N., ZHENG, W., JI, B.-T., BRESLOW, N.E. and DAY,N.E., Statistical methods in cancer research: BLOT,W.J. and FRAUMENI, J.F. JR., Risk factors for renal-cell cancer analysis of case-control studies. IARC Scientific Publication 32, IARC, in Shanghai, China. Int. J. Cancer, 52,562-565 (1992). Lyon (1980). MCLAUGHLIN, J.K., LINDBLAD, P., MELLEMGAARD, A., MCCREDIE, M., COHEN, A.J., LI, F.P., BERG,S., MARCHETTO, D.J., TSAI,S., JACOBS, MANDEL, J.S., SCHLEHOFER, B., POMMER, W. and ADAMI, H.-O., R.S., Hereditary renal-cell carcinoma associated S.C. and BROWN, International Renal-Cell-Cancer Study. I. Tobacco use. Int. J. Cancer, with a chromosomal translocation. New Engl. J. Med., 301, 592-595 60,194-198 (199%). (1979). MCLAUGHLIN, J.K., MANDEL, J.S., BLOT, W.J., SCHUMAN, L.M., MEHL, CURHAN, G.C., WILLETT, W.C., RIMM, E.B. and STAMPFER, M.J., A E.S. and FRAUMENI, J.F. JR., A population-based case-control study of prospective study of dietary calcium and other nutrients and the risk of renal-cell carcinoma. J. nat. Cancer Inst., 72,275-284 (1984). symptomatic kidney stones. New Engl. J. Med., 328,833-838 (1993). PORT,F.K., RAGHEB, N.E., SCHWARTZ, A.G. and HAWTHORNE, V.M., EASTON, D.F., The inherited component of cancer. Brit. rned. Bull., 50, Neoplasms in dialysis patients: a population-based study. Amer. J. 527-535 (1994). Kidney Dis., 14,119-123 (1989). FRANKSSON, C., BERGSTRAND, A., LJUNGDAHL, I., MAGNUSSON, G. and RAYNOR, W.J., SHEKELLE, R.B., ROSSOF, A.H., MALIZA, C. and PAUL, NORDENSTAM, H., Renal carcinoma (hypernephroma) occurring in 5 O., High blood pressure and 17-year cancer mortality in the Western sib1ings.J. Urol., 108,58-61 (1972). Electric Health Study. Amer. J. Epidemiol., 113,371-377 (1981). FRASER, G.E., PHILLIPS, R.L. and BEESON,W.L., Hypertension, A.G., DEXEUS, F., SWANSON, D.A. and VON ESCHENBACH, anti-hypertensive medication and risk of renal carcinoma in California ROSENBERG, A.C., Relationship of thyroid disease to renal-cell carcinoma. An Seventh-Day Adventists. Int. J. Epidemiol., 19,832-838 (1990). epidemiologic study. Urdogy, 35,492-498 (1990). GABOW, P.A. and BENNETT, W.M., Renal manifestations: complicaTALAMINI, R., BARON, A.E., BARRA, S., BIDOLI,E., LA VECCHIA, C., tion management and long-term outcome of autosomal dominant NEGRI, E., SERRAINO, D. and FRANCESCHI, S., A case-control study of polycystic kidney disease. Semin. Nephrol., 11,643-652 (1991). risk factor for renal-cell cancer in northern Italy. Cancer Causes HOFFMAN, D.A., MCCONAHEY, W.M., FRAUMENI, J.F. and KURLAND, Control, 1,125-131 (1990). L.T., Cancer incidence following treatment of hyperthyroidism. Int. J. TRINCHIERI, A., MANDRESSI, A., LUONGO, P., LONGO, G. and PISANI, Epidemiol., 11,218-224 (1982). P., The influence of diet on urinary risk factors for stones in healthy HUDSON, H.C. and WILSON, C.E., Bilateral renal-cell carcinoma in von subjects and idiopathic renal calcium stone formers. Brit. J. Urol., 67, Hippel-Lindau disease: treated with staged bilateral nephrectomy and 230-236 (1991). dialysis. Sth. rned. J. (Bgham, AL), 72,363-365 (1979). TUCKER,M.A., BOICE, J.D. and HOFFMAN, D.A., Second cancer IGUCHI, M., UMEKAWA, T. and ISHIKAWA, Y., Dietary intake and habits following cutaneous melanoma and cancers of the brain, thyroid, of Japanese renal-stone patients. J. Urol., 143,1093-1095 (1990). connective tissue, bone and eye in Connecticut, 1935-82. Nut. Cancer ISHIKAWA, I., SAITO, Y., SHIKURA, N., KITADA, H., SHINODA, A. and Inst. Monogr., 68,161-189 (1985). SUZUKI, S., Ten-year prospective study on the development of renal- WOLK,A,, GRIDLEY, G., NIWA,S., LINDBLAD, P., MCCREDIE, M., cell carcinoma in dialysis patients. Amer. J. Kidney Dis., 16, 452-458 MELLEMGAARD, A,, MANDEL, J., WAHRENDORF, J., MCLAUGHLIN, J. (1990). and ADAMI, H.O., International Renal-Cell-Cancer study. VII. Role of KREIGER, N., MARRETT, L.D., DODOS, L., HILDITCH, S. and DARLING- Diet. Int. J. Cancer, 65,67-73 (1996). TON, G.A., Risk factors for renal-cell carcinoma: results of a populationWYNDER, E.L., MABUCHI, K. and WHITMORE, W.F., Epidemiology of based case-control study. Cancer Causes Conirol, 4,101-110 (1993). adenocarcinoma of the kidney. J. nut. Cancer Inst., 53, 1619-1634 LYNCH, H.T. and WALZAK, M.P., Genetics in urologic cancer. Urol. (1974). Clin. North Amer., 7,815-829 (1980). ZBAR,B., GLENN, G., LUBENSKY, I., CHOYKE, P., WALTHER, M.M., MACLURE, K.M. and WILLETT, W., A case-control study of diet and MAGNUSSON, G., BERGERHEIM, U.S.R., PETTERSSON, S., AMIN,M., risk of renal adenocarcinoma. Epidemiology, 1,430-440 (1990). HURLEY, K. and LINEHAN, W.M., Hereditary papillary renal-cell MAHER,E.R. and YATES,J.R.W., Familial renal-cell carcinoma: carcinoma: clinical studies in 10 families. J. UroL, 153,907-912 (1995). clinical and molecular genetic aspects. Brit. J. Cancer, 63, 176-179 ZEIER,M., GEBERTH, S., RITZ, E., JAEGER, T. and WALDHERR, R., (1991). Adult dominant polycystic kidney disease-clinical problems. NephMATSON, M.A. and COHEN, E.P., Acquired cystic kidney disease: ron., 49,177-183 (1988).