Familial Renal Cell Carcinoma from the Swedish Family-Cancer Database

Transcription

1 EUROPEAN UROLOGY 60 (2011) available at journal homepage: Platinum Priority Kidney Cancer Editorial by Marc C. Smaldone, Veda N. Giri and Robert G. Uzzo on pp of this issue Familial Renal Cell Carcinoma from the Swedish Family-Cancer Database Hao Liu a, *, Jan Sundquist b,c, Kari Hemminki a,b a Division of Molecular Genetic Epidemiology, German Cancer Research Centre (DKFZ), Heidelberg, Germany b Centre for Primary Health Care Research, Lund University, Malmö, Sweden c Stanford Prevention Research Centre, Stanford University School of Medicine, Stanford, California, USA Article info Article history: Accepted May 15, 2011 Published online ahead of print on May 25, 2011 Keywords: Renal cell carcinoma Familial risk Heredity Abstract Background: Reliable data on familial risks are important for clinical counselling and cancer genetics. Objective: To evaluate familial risks for renal cell carcinomas (RCC) through parental and sibling probands in the largest available dataset. Design, setting, and participants: This study examined the Swedish Family-Cancer Database on 12.2 million individuals, which contains families with parents and offspring. Cancer data were retrieved from the Swedish Cancer Registry for the years , including 8513 patients with RCC. Measurements: Familial risk for offspring was defined through standardised incidence ratios (SIRs) and adjusted for many variables, including a proxy for smoking and obesity. Results and limitations: The familial risk for RCCs was 1.75 when a parent and 2.61 when a sibling was diagnosed with any kidney cancer. Also, RCCs were shown to be associated with prostate cancer (PCa) when parents or parents and siblings were diagnosed with PCa. Among siblings, the associations of RCC with melanoma, non- Hodgkin s lymphoma, and urinary bladder and papillary thyroid tumours were found. None of the results differed significantly after excluding the families with cancer pathognomonic of a von Hippel-Lindau (VHL) disease. Limitations of this study include the small number of familial cases (229 familial cases). Conclusions: The present analysis showed a high familiarity for RCC, and recessive effects may be important for familial aggregation of RCC. As a novel association, offspring RCC was in excess when parents or parents and siblings were diagnosed with PCa. There is familial clustering beyond VHL and the recent low-risk gene that probably explains a small proportion of the observed familial clustering. # 2011 European Association of Urology. Published by Elsevier B.V. All rights reserved. * Corresponding author. Division of Molecular Genetic Epidemiology, German Cancer Research Centre (DKFZ), Heidelberg, Germany. Tel ; Fax: address: h.liu@dkfz.de (H. Liu) /$ see back matter # 2011 European Association of Urology. Published by Elsevier B.V. All rights reserved. doi: /j.eururo

2 988 EUROPEAN UROLOGY 60 (2011) Introduction Renal cell carcinomas (RCC) account for >90% of adult kidney cancer and include the clear-cell, papillary, and chromophobe subtypes. Incidence rates in RCC are generally high in Europe and North America and low in Asia and South America; in addition, they vary by country. In Sweden, the incidence has remained relatively stable over the past decades [1], but in many European countries and the United States, there has been an increase of about 2% per year [2]. More than 2% of Swedish RCC patients have a parent with RCC, and the familial risk has been about 1.6 [3,4]. In addition, recessive effects (ie, high risk among siblings) might be important for familial aggregation of RCC [3,5,6]. A recent systematic meta-analysis also showed a significant positive association between a family history of kidney cancer and RCC risk [7]. However, most previous studies on family history were case-control studies, which are sometimes less valued for being retrospective. Some established risk factors for kidney cancer are cigarette smoking, excess body weight, hypertension, and familial cancer syndromes, which include von Hippel- Lindau (VHL) syndrome, tuberous sclerosis, Birt-Hogg-Dubé syndrome, and hereditary papillary RCC [8,9]. All of these syndromes are rare and probably account for a small proportion of familial RCC, and other susceptibility genes are to be expected [9,10]. Several types of associated tumours have been reported in families with hereditary RCCs, but population-based studies on RCCs and associated tumours in families have been limited [3,6,8,11]. In the present study, we use the 2010 update of the Swedish Family-Cancer Database to address questions about familial risks for RCCs and their association with other familial cancers at a national level. Compared to the previous study from the database, an extended population and hence some 100% more familial RCC cases are available [3] (229 familial RCC cases now vs 103 familial RCC cases earlier). Also, we have the possibility of considering both parent offspring and sibling effects in the offspring population up to 76 yr of age. This report, the largest cohort study of familial RCC, offers new insight into the aetiology of the disease. 2. Patients and methods Statistics Sweden maintains a Multigeneration Register in which individuals (called offspring in this study) born in Sweden in 1932 and later are registered with their parents (those pleading parenthood at birth) and are organised according to families. Practically, only the two oldest generations, parents and offspring, contributed cases to family studies [12]. The data on families and cancers have complete coverage, barring some groups of deceased offspring, which affects those born in the 1930 s and those who died before This small group of offspring with missing links to parents has a negligible effect on the estimates of familial risk [13]. The Multigeneration Register was linked by the individually unique national registration number to the Cancer Registry from the years Cancer cases were retrieved from the Swedish Cancer Registry, which relies on separate compulsory notifications from clinicians who diagnosed the neoplasms as well as pathologists and cytologists. Cancer registration is considered to be close to 100% currently [14]. Because there is incomplete information in the database about death among cancer cases between 1959 and 1960, we used the follow-up period between 1 January 1961 and 31 December The site of cancer is registered based on a four-digit diagnostic code according to the International Classification of Diseases, 7th revision (ICD-7): Code was used for renal parenchyma tumour; hemangioblastoma (including hemangioma) was defined as benign tumour of the central nervous system (ICD-7 code 193), with histology codes 501 and 511, and adrenal tumours (ICD-7 code 195.0) included pheochromocytomas (histology code 446). From 1993 onward, International Classification of Disease for Oncology, 2nd edition (ICD-O-2)/ICD with histopathologic data was used. According to this classification, it was possible to distinguish papillary thyroid cancer. Standardised incidence ratios (SIRs) were used to measure the cancer risks for offspring when their parents, siblings, or both were diagnosed with specific cancers (ie, using parents and siblings as probands). The reference rate was calculated for offspring whose parents or siblings had no specified cancer as the number of cases divided by person-years at risk. SIR was the ratio of the observed (O) to expected number of cases. We used chronic obstructive pulmonary disease (COPD) as a proxy for smoking. Adjustment for obesity was based on prior hospitalisation for obesity. Data on COPD and obesity were obtained from the Swedish Hospital Discharge Register, which records complete data on all discharges, with dates of hospitalisation and diagnoses in some regions since 1964 and nationwide since The expected numbers were calculated from 5-yr age, sex, period, residential area, socioeconomic status, obesity, and smokingspecific standardisedratesfor offspring whoseparents were not diagnosed with kidney cancer. This is an indirect standardisation for the listed possible intervening variables. Confidence intervals (CI) of 95% were calculated assuming a Poisson distribution of subjects [15]. Follow-up was started for each offspring at birth, immigration, or 1 January 1961, whichever came latest. When the ICD-O-2/ICD histology was used, followup was started on 1 January Follow-up was terminated on diagnosis of first cancer, death, emigration, or the closing date of the study, 31 December Risks for siblings were calculated using the cohort method, considering dependence between the pairs in the assessment of 95% CIs, as described elsewhere [16]. Some analyses were carried out in reverse order, using family members with any kidney cancer as probands and calculating SIRs for any cancer among offspring. 3. Results There were 9763 cases of kidney cancer in offspring from between 1961 and 2008 in the database. Among 9763 kidney cancer cases, 8513 cases were RCC cases (5257 in men and 3256 in women). Familial risks for RCCs in offspring are shown in Table 1 when parents (without affected siblings) and siblings (without affected parents) were used as probands. The SIR for RCCs was 1.75 when a parent and 2.61 when a sibling was diagnosed with any kidney cancer. At discordant cancer sites, parental prostate and adrenal gland cancers (including pheochromocytoma) were significantly associated with an elevated risk of RCC; the associations of RCCs with urinary bladder tumour (1.51) and melanoma (1.51) as well as non-hodgkin s lymphoma (NHL; 1.37) were also identified among siblings. An association with papillary thyroid cancer among siblings was significant (3.43). However, no significant association of papillary RCC with papillary thyroid cancer was observed in the present study (data not shown). We carried out analyses of RCCs in reverse order by calculating SIRs for any cancer in offspring using parents

3 EUROPEAN UROLOGY 60 (2011) Table 1 Standardised incidence ratios for renal cell carcinomas in offspring whose parents or siblings have been diagnosed with any cancer ( ) Cancer site of probands Parents only Siblings only Parents and siblings O SIR 95% CI * O SIR 95% CI * O SIR 95% CI * Upper aerodigestive tract Salivary gland Oesophagus Stomach Small intestine Colorectum Liver Pancreas Lung Breast Cervix Endometrium Ovary Other female genital Prostate Testis Kidney Urinary bladder Melanoma Skin Eye Nervous system Hemangioblastoma Thyroid gland Papillary ** Endocrine glands Adrenal glands Connective tissue NHL Hodgkin s disease Myeloma Leukaemia O = observed; SIR = standardised incidence ratio; CI = confidence interval; NHL = non-hodgkin s lymphoma. * Follow-up from 1993 onwards. ** Bold type: 95% CI does not include and siblings with any kidney cancer as probands; for any discordant site, this approach is a completely independent analysis for the parent-to-offspring comparison. In Table 2, with the exception of kidney cancer, some sites of the significant increases in Table 1 were also found in the reverse analyses (eg, urinary bladder cancer [1.59], melanoma [1.31], NHL [1.34], and papillary thyroid cancer [2.96]). Also, the SIRs for pancreatic endocrine tumours (5.77) and hemangioblastoma (3.20) were significantly increased from parental probands in the reverse analysis. Table 3 shows the age-specific familial risks for RCC. In addition to RCC, only sibling thyroid cancer was significantly associated with RCC diagnosed before 50 yr of age. The significant associations with melanoma, urinary bladder, and papillary thyroid cancers among siblings were limited to RCC diagnosed after 50 yr of age. Familial risks were also analysed by histology; data for clear-cell RCC (ccrcc) are not shown, because the results did not essentially differ from those for all RCC. For ccrcc, the SIR was ( ) or 1.92 ( ) when parents and siblings or siblings, respectively, were diagnosed with any kidney cancer. At discordant cancer sites, parental nervous system tumour was associated with an increased risk of ccrcc (SIR: 1.62; n = 35; 95% CI, ). Considering the influences of VHL, smoking, and obesity, we repeated all the above calculations after successively excluding 136 putative VHL families (defined as any family with two cases of hemangioblastoma or one case of hemangioblastoma and a second case of RCC or pheochromocytoma [17,18]), individuals with obesity, or individuals with COPD; none of the results differed significantly. In addition, removal of COPD and obesity two risk factors for kidney cancer did not change the SIRs for familial kidney cancer (data not shown). 4. Discussion All kidney cancer diagnoses reported to the Swedish Cancer Registry are histologically verified, and we can assume that the diagnostic accuracy is as good as it can be at a national level [14]. Although the present study was nationwide, the number of cases was small for many familial associations. Familial clustering of cancer may be the result of environmental factors shared by family members or the result of shared genes [5,19,20]. Our previous studies have examined the contribution of environmental sharing to the

4 990 EUROPEAN UROLOGY 60 (2011) Table 2 Standardised incidence ratios for cancers in offspring whose parents or siblings have been diagnosed with kidney cancer ( ) Cancer site in offspring Parents or siblings with kidney cancer Parents only Siblings only Parents and siblings O SIR 95% CI * O SIR 95% CI * O SIR 95% CI * Upper aerodigestive tract Salivary gland Oesophagus Stomach Small intestine Colorectum Liver Pancreas Pancreatic endocrine tumour Lung Breast Cervix Endometrium Ovary Other female genital Prostate Testis RCC Urinary bladder Melanoma Skin Eye Nervous system Hemangioblastoma Thyroid gland Papillary ** Endocrine glands Adrenal glands Connective tissue NHL Hodgkin s disease Myeloma Leukaemia O = observed; SIR = standardised incidence ratio; CI = confidence interval; RCC = renal cell carcinoma; NHL = non-hodgkin s lymphoma. * Bold type: 95% CI does not include ** Follow-up from 1993 onwards. familial risk by assessing spouse correlations for cancer, including RCC; no significant result was found for RCC [20,21]. The caveat is that childhood shared effects would be not detected by comparing risk among spouses. However, in analysis of all types of family relationships, a low degree of environmental sharing was noted for RCC. Thus, we believe that the present results address mainly heritable effects. Our analysis revealed a high familial risk for RCCs when parents or siblings were diagnosed with kidney cancer SIRs being 1.75 and 2.61, respectively and the excess risk in siblings could be a recessive effect [6]. The risk was higher for early-onset RCC patients (<50 yr of age). Our previous study attempted to solve this problem by comparing the familial risks between siblings according to their age difference; the findings offered evidence on recessive effects in early-onset RCC [22]. A dominant predisposition to RCC in association with VHL disease is well documented [23]. In the present analysis, VHL-related tumours associated with an increased RCC risk between two generations were hemangioblastoma and pancreatic endocrine tumour (reverse analysis) and adrenal gland tumour (pheochromocytoma). However, we found only two RCC cases among siblings with features of a VHL disease, which suggested that high familial clustering of RCC among siblings cannot be explained by the occurrence of the VHL disease. Moreover, because even the other known familial cluster of RCC, including chromosome 3 translocations, tuberous sclerosis, and Birt-Hogg-Dubé syndrome, show dominant inheritance, we suggest that an unknown recessive condition might have an important role for familial RCC [8,24]. The present study also suggested that there is familial clustering beyond VHL syndrome, because familial risk did not differ significantly after excluding the putative VHL families. Despite recent successes in kidney cancer genomewide association study [25], these newly identified low-penetrance genes probably explain a small proportion of the observed familial clustering. Apart from VHL-related tumours, significantly increased risks were noted for several other discordant sites. The analyses covered four independent comparisons through paternal, sibling, parental and sibling, and the reverse parental probands; the reverse analysis between siblings provided no independent evidence. Finding at least two

5 Table 3 Standardised incidence ratios for renal cell carcinomas by age of onset in offspring whose parents or siblings have been diagnosed with any cancer ( ) Cancer site of probands Parents only Siblings only Parents and siblings <50 yr of age 50 yr of age <50 yr of age 50 yr of age <50 yr of age 50 yr of age O SIR 95% CI * O SIR 95% CI * O SIR 95% CI * O SIR 95% CI * O SIR 95% CI * O SIR 95% CI * Upper aerodigestive tract Salivary gland Oesophagus Stomach Small intestine Colorectum Anus Liver Pancreas Nose Lung Breast Cervix Endometrium Ovary Other female genital Prostate Testis Kidney Urinary bladder Melanoma Skin Eye Nervous system Hemangioblastoma Thyroid gland Papillary ** Endocrine glands Adrenal glands Bone Connective tissue NHL Hodgkin s disease Myeloma Leukaemia O = observed; SIR = standardised incidence ratio; CI = confidence interval; NHL = non-hodgkin s lymphoma. * Bold type: 95% CI does not include ** Follow-up from 1993 onwards. EUROPEAN UROLOGY 60 (2011)

6 992 EUROPEAN UROLOGY 60 (2011) significant associations provides strong evidence for a true familial association. Using these criteria, the associations of RCC with prostate and lung tumours were confirmed. Because the association between RCC and urinary bladder tumour was significant among siblings and borderline significant through parental proband in the reverse analysis, a true association between the two tumours was also confirmed. The increased risk for prostate cancer (PCa) after kidney cancer was identified in previous population-based studies [3,26], but the association between these two tumours in families has been rarely reported. The associations that remained unconfirmed were between RCC and melanoma, NHL, and papillary thyroid tumour. However, although no clear common factors can fully explain association between RCC and melanoma in the present study, it is intriguing that both tumours are dependent on the same cytokines, such as interleukin-2 [27]. The suggestion of an association between RCC and NHL is consistent with other studies [28,29], but the reasons for it remain largely unexplained. Also, the association of RCC with papillary thyroid cancer was observed, probably the result of a heritable syndrome of familial papillary thyroid cancer and papillary renal neoplasia [30]. However, this association was not observed between papillary RCC and papillary thyroid cancer; because of rare familial cases, this syndrome appears less likely. A limitation of this study is that no biologic samples were available from patients, thus excluding possibilities of molecular search for the suggested recessive inheritance. As consanguinity is rare in Sweden, it is not likely that two identical alleles would underlie the susceptibility; homozygosity mapping of alleles in affected individuals would thus not be the genetic method of choice. A more likely scenario is that two heterogeneous recessive alleles at the susceptibility locus could cause the disease in a heterozygous state. 5. Conclusions The present analysis from a national population-based database showed a high familiarity for RCC, and recessive susceptibility genes might play a role in familial aggregation of RCC. As a novel association, offspring RCC was in excess when parents or parents and siblings were diagnosed with PCa. Moreover, there is familial clustering beyond VHL and the recent low-risk gene, and this probably explains a small proportion of the observed familial clustering. Author contributions: Hao Liu had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Hemminki, Liu. Acquisition of data: Hemminki, Sundquist. Analysis and interpretation of data: Liu. Drafting of the manuscript: Liu. Critical revision of the manuscript for important intellectual content: Hemminki, Sundquist. Statistical analysis: Liu. Obtaining funding: Hemminki. Administrative, technical, or material support: Hemminki. Supervision: Hemminki. Other (specify): None. Financial disclosures: I certify that all conflicts of interest, including specific financial interests and relationships and affiliations relevant to the subject matter or materials discussed in the manuscript (eg, employment/affiliation, grants or funding, consultancies, honoraria, stock ownership or options, expert testimony, royalties, or patents filed, received, or pending), are the following: None. Funding/Support and role of the sponsor: Deutsche Krebshilfe, the Swedish Cancer Society, and the Swedish Council for Working Life and Social Research provided support for the collection, management, and interpretation of the data in this study as well as approval of the manuscript. The Family-Cancer Database was created by linking registers maintained at Statistics Sweden and the Swedish Cancer Registry. References [1] Lindblad P. Epidemiology of renal cell carcinoma. Scand J Surg 2004;93: [2] Eurocim version 4.0. European incidence database V23, 730 entity dictionary. Lyon, France; European Network of Cancer Registries; [3] Czene K, Hemminki K. Kidney cancer in the Swedish Family Cancer Database: familial risks and second primary malignancies. Kidney Int 2002;61: [4] Hemminki K, Czene K. Attributable risks of familial cancer from the Family-Cancer Database. Cancer Epidemiol Biomarkers Prev 2002;11: [5] Gudbjartsson T, Jonasdottir TJ, Thoroddsen A, et al. A populationbased familial aggregation analysis indicates genetic contribution in a majority of renal cell carcinomas. Int J Cancer 2002;100: [6] Hemminki K, Li X. Familial renal cell cancer appears to have a recessive component. J Med Genet 2004;41:e58. [7] Clague J, Lin J, Cassidy A, et al. Family history and risk of renal cell carcinoma: results from a case-control study and systematic metaanalysis. Cancer Epidemiol Biomarkers Prev 2009;18: [8] Takahashi M, Kahnoski R, Gross D, Nicol D, Teh BT. Familial adult renal neoplasia. J Med Genet 2002;39:1 5. [9] Linehan WM, Srinivasan R, Schmidt LS. The genetic basis of kidney cancer: a metabolic disease. Nat Rev Urol 2010;7: [10] Dalgliesh GL, Furge K, Greenman C, et al. Systematic sequencing of renal carcinoma reveals inactivation of histone modifying genes. Nature 2010;463: [11] Hemminki K, Li X, Czene K. Familial risk of urological cancers: data for clinical counseling. World J Urol 2004;21: [12] Hemminki K, Li X, Plna K, Granstrom C, Vaittinen P. The nation-wide Swedish Family-Cancer Database updated structure and familial rates. Acta Oncol 2001;40: [13] Hemminki K, Li X. Familial risk of cancer by site and histopathology. Int J Cancer 2003;103: [14] Centre for Epidemiology. Cancer incidence in Sweden. Stockholm, Sweden: National Board of Health and Welfare; [15] Breslow NE, Day NE. Statistical methods in cancer research. Volume II the design and analysis of cohort studies. IARC Scientific Publications No. 82. Lyon, France: International Association for Research on Cancer; [16] Hemminki K, Vaittinen P, Dong C, Easton D. Sibling risks in cancer: clues to recessive or X-linked genes? Br J Cancer 2001;84: [17] Lonser RR, Glenn GM, Walther M, et al. von Hippel-Lindau disease. Lancet 2003;361:

7 EUROPEAN UROLOGY 60 (2011) [18] Hemminki K, Li X, Collins VP. A population-based study of familial central nervous system hemangioblastomas. Neuroepidemiology 2001;20: [19] Lichtenstein P, Holm NV, Verkasalo PK, et al. Environmental and heritable factors in the causation of cancer analyses of cohorts of twins from Sweden, Denmark, and Finland. N Engl J Med 2000;343: [20] HemminkiK,DongC,VaittinenP.Cancer riskstospousesand offspring in the Family-Cancer Database. Genet Epidemiol 2001;20: [21] Hemminki K, Jiang Y. Cancer risks among long-standing spouses. Br J Cancer 2002;86: [22] Hemminki K, Li X. Age-specific familial risks for renal cell carcinoma with evidence on recessive heritable effects. Kidney Int 2004; 65: [23] Latif F, Tory K, Gnarra J, et al. Identification of the von Hippel- Lindau disease tumor suppressor gene. Science 1993;260: [24] Woodward ER, Clifford SC, Astuti D, Affara NA, Maher ER. Familial clear cell renal cell carcinoma (FCRC): clinical features and mutation analysis of the VHL, MET, and CUL2 candidate genes. J Med Genet 2000;37: [25] Purdue MP, Johansson M, Zelenika D, et al. Genome-wide association study of renal cell carcinoma identifies two susceptibility loci on 2p21 and 11q13.3. Nat Genet 2011;43:60 5. [26] Barocas DA, Rabbani F, Scherr DS, Vaughan Jr ED. A populationbased study of renal cell carcinoma and prostate cancer in the same patients. BJU Int 2006;97:33 6. [27] Nathan PD, Eisen TG. The biological treatment of renal-cell carcinoma and melanoma. Lancet Oncol 2002;3: [28] Negri E, Talamini R, Montella M, et al. Family history of hemolymphopoietic and other cancers and risk of non-hodgkin s lymphoma. Cancer Epidemiol Biomarkers Prev 2006;15: [29] Pottern LM, Linet M, Blair A, et al. Familial cancers associated with subtypes of leukemia and non-hodgkin s lymphoma. Leuk Res 1991;15: [30] Malchoff CD, Sarfarazi M, Tendler B, et al. Papillary thyroid carcinoma associated with papillary renal neoplasia: genetic linkage analysis of a distinct heritable tumor syndrome. J Clin Endocrinol Metab 2000;85: