Vitamin D & Exposure in Relation to Kidney Cancer

by Sara Karami

B.S., Biology, James Madison University, 2002 M.P.H, Epidemiology, The George Washington University, 2004

A Dissertation submitted to

The Faculty of The Columbian College of Arts and Science of The George Washington University in partial fulfillment of the requirements for the degree of Doctor of Philosophy

August 31, 2009

Dissertation directed by

Katherine L. Hunting Professor of Environmental and Occupational Health and of Epidemiology and Biostatistics

and

Lee E. Moore Epidemiological Investigator, NIH, NCI

The Columbian College of Arts and Science of The George Washington University certifies that Sara Karami has passed the Final Examination for the degree of Doctor of

Philosophy as of August 12, 2009. This is the final and approved form of the dissertation.

Vitamin D Genes & Exposure in Relation to Kidney Cancer

Sara Karami

Dissertation Research Committee:

Katherine L. Hunting, Professor of Environmental and Occupational Health and of

Epidemiology and Biostatistics, Dissertation Director

Lee E. Moore, Epidemiological Investigator, NIH, NCI, Co-Director

Paul H. Levine, Professor of Epidemiology and Biostatistics, Committee Member

Yinglei Lai, Assistant Professor of Statistics, Committee Member

ii

© Copyright 2009 by Sara Karami All rights reserved

iii Dedication

The author wishes to thank everyone involved in the dissertation process for their guidance and support. Special thanks to Dr. Lee Moore, Dr. Katherine Hunting, Dr. Paul

Levine, Dr. Yinglei Lai, Dr. Sean Cleary, and Dr. Donte Verme.

iv Acknowledgement

The author wishes to acknowledge the National Cancer Institute, the International

Agency for Research on Cancer, and the School of Public Health and Health Services of

The George Washington University for their assistance.

v Abstract of Dissertation

Vitamin D Genes & Exposure in Relation to Kidney Cancer

Introduction : Vitamin D may have anti-carcinogenic properties that include inhibition of clonal tumor cell proliferation, induction of immune cell differentiation, and decreased angiogenesis. Within the kidney, vitamin D is metabolized to its active form. Since the incidence of renal cell carcinoma (RCC) and prevalence of vitamin D deficiency have increased over the past few decades, this study hypothesized that increased vitamin D exposure (via occupational ultraviolet exposure or dietary intake) was associated with decreased RCC risk and that genetic variations within the vitamin D pathway modified risk. Methods : Cases (N=1,097) and controls (N=1,476) in a hospital-based case-control

study in Central Europe were interviewed to collect data on demographics and lifetime

occupational histories. Genomic DNA was also collected from a subset of participants.

Results : Significant reduction in RCC risk was observed with occupational ultraviolet exposure among male participants. No association between ultraviolet exposure and RCC risk was observed among females. Analyses stratified by latitude showed a stronger reduction in risk among males at the highest latitude study site, Russia. Analyses of eight vitamin D pathway genes revealed significant associations between RCC risk and the ( VDR ) and retinoid-X-receptor-alpha (RXRA ) genes. Across VDR ,

three haplotypes within two regions were associated with increased risk. Across RXRA ,

RCC risk was higher among participants with one particular haplotype located

downstream of the coding region. Dietary analyses showed increased RCC risk with

increasing intake frequency of yogurt for foods rich in calcium, while decreased risk was

vi observed with eggs for foods rich in vitamin D. Additional evaluation of RXRA and VDR , revealed RXRA variants, 3 ′ of the coding sequence, modified associations between RCC

risk and intake frequency of vitamin D rich foods, specifically eggs. Conversely, RXRA variants in introns 1 and 4 modified associations between calcium rich foods.

Furthermore, increased RCC risk was observed with increasing occupational ultraviolet exposure among males with one specific VDR haplotype, centered on intron two.

Conclusion : Results suggest that vitamin D is associated with RCC risk. Genetic variants

across VDR and RXRA genes may be associated with RCC risk and may modify associations between RCC risk and vitamin D.

vii Table of Contents

Dedication iv

Acknowledgments v

Abstract of Dissertation vi

Table of Contents viii

List of Figures ix

List of Tables x

List of Symbols / Nomenclature xii

Chapter 1: Cancer & Vitamin D 1

Chapter 2: Study Hypothesis, Specific Aims and the Central & Eastern 33

European Renal Cell Carcinoma (CEERCC) Study

Chapter 3: Occupational Sunlight Exposure and Risk of Renal Cell Carcinoma 55

Chapter 4: Analysis of SNPs and Haplotypes in VDR Pathway Genes and Renal 81

Cancer Risk

Chapter 5: Effect of VDR pathway genes on vitamin D intake and exposure 129

and Renal Cancer Risk

Chapter 6: Past, Present and Future of Vitamin D and Cancer Risk Studies 157

References 186

Appendices 238

viii List of Figures

Figure 1: Worldwide kidney cancer incidence rates from 1998 to 2000 2

Figure 2: U.S. incidence and mortality rates for kidney cancer from 1975 4

to 2001

Figure 3: Synthesis of vitamin D in the body from diet & sunlight exposure 9

Figure 4: Genes involved in the vitamin D pathway 29

Figure 5: Power calculations to detect an OR of 1.3-1.5 fold increase in RCC 52

risk with 900 cases and 900 controls

Figure 6: Power calculations to detect an OR of 1.3-1.5 fold increase in RCC 53

risk with 1,100 cases and 1,100 controls

Figure 7: HaploWalk and Haploview analysis for the group specific 250

component ( GC ) vitamin D binding protein

Figure 8: HaploWalk and Haploview analysis for the signal transducer and 252

activator of transcription ( STAT1 ) gene

ix List of Tables

Table 1: Epidemiological studies of dietary exposures and RCC risk 5

Table 2: Dietary Intake for Vitamin D Recommended by the National 18

Academy of Science

Table 3: Common VDR variants and risk of prostate cancer 24

Table 4: Common VDR variants and risk of breast cancer 27

Table 5: General characteristics of participants in the CEERCC study 41

Table 6: Power to detect a two-fold interaction with 900 cases and 900 controls 54

Table 7: Tissue that express VDR 169

Table 8: SEER Skin Cancer (excluding Basal and Squamous) Rates from 184

2001-2005

Table 9: Risk of renal cell carcinoma and exposure to occupational sunlight 239

by body mass index

Table 10: Risk of renal cell carcinoma and exposure to occupational sunlight 240

by body mass index among male participants

Table 11: Risk of renal cell carcinoma and exposure to occupational sunlight 241

by body mass index among female participants

Table 12: Risk of renal cell carcinoma and exposure to occupational sunlight 243

by hypertensive status

Table 13: Risk of renal cell carcinoma and exposure to occupational sunlight 244

by hypertensive status among male participants

Table 14: Risk of renal cell carcinoma and exposure to occupational sunlight 245

by hypertensive status among female participants

x Table 15: Risk of renal cell carcinoma and exposure to occupational sunlight 247

by dietary intake of calcium rich foods among female participants

Table 16: Haplotype associations with genes in the vitamin D pathway 251

Table 17: Gene-gene analyses for vitamin D pathway genes and RCC risk 254

Table 18: Joint effect of vitamin D pathway genes and dietary intake 256

frequency of total vitamin D on renal cancer risk

Table 19: Joint effect of vitamin D pathway genes and dietary intake 258

frequency of total calcium on renal cancer risk

Table 20: Joint effect of vitamin D pathway genes and cumulative 260

occupational UV exposure on renal cancer risk

Table 21: Population attributable risk among participants occupationally 263

exposed to sunlight

Table 22: Population attributable risk for dietary intake frequency of total 264

vitamin D

Table23: Population attributable risk among VDR genotyped participants 265

xi List of Symbols / Nomenclature

1. 1,25(OH) 2D: 1,25 dihydroxycholecalciferol or calcitriol

2. 1,25(OH) 2: 1,25-dihydroxy

3. 19-nor-1,25-dihydroxyvitamin D 2: paricacitol

4. 25(OH): 25-hydroxy

5. 7DHC: 7-dehydrocholesterol

6. 95% CI: 95% confidence interval

7. AIPC: androgen-independent prostate cancer

8. BMI: body mass index

9. CEERCC: Central & Eastern European Renal Cell Carcinoma

10. CGF: Core Genotyping Facility

11. CYP24A1 : cytochrome P450, family 24, subfamily A, polypeptide 1 or 24-

dehydroxylase

12. DNA: deoxyribonucleic acid

13. DRIP : vitamin D receptor interacting protein

14. FDR: false discovery rate

15. GC : group specific component

16. HCC: hepatocellular carcinoma

17. IARC: International Agency for Research on Cancer

18. ICD-O: International Classification of Diseases for Oncology

19. ISCO: International Standard Classification of Occupation

20. IU: International Units

21. JEM: job exposure matrix

xii 22. LD: linkage disequilibrium

23. MAF: minor allele frequency

24. Min-P: minimum-p-value permutation

25. mRNA: messenger ribonucleic acid

26. NACE: Statistical Classification of Economic Activities of the European Community

27. NCI: National Cancer Institute

28. NHANES: National Health and Nutrition Examination Survey

29. NHL: non-Hodgkin lymphoma

30. OPA: Oligo Pool All

31. OR: odds ratio

32. PAR: Population Attributable Risk

33. PCR: polymerase chain reaction

34. PSA: prostate specific antigen

35. PTH: parathyroid hormone

36. RCC: renal cell carcinoma

37. RFLP: restriction-fragment length polymorphism

38. RR: relative risk

39. RXR : retinoid-X-receptor

40. RXRA : retinoid-X-receptor-alpha

41. SEER: Surveillance Epidemiology and End Results

42. SNP: single nucleotide polymorphism

43. STAT1 : signal transducer and activator of transcription

44. THRAP : associated protein

xiii 45. U.S.: United States

46. USRCC: United States Renal Cell Carcinoma

47. UTR: untranslated region

48. UV: ultraviolet

49. VDR : vitamin D receptor

50. VDRE : vitamin D response element

51. vitamin D 2: ergocacliferol

52. vitamin D 3: cholecaliferol

53. χ2: Chi-square

xiv Chapter 1: Cancer & Vitamin D

1.1 Renal Cell Cancer:

In 1985, kidney cancer accounted for 1.7% of all malignant diseases worldwide; today

kidney cancer accounts for approximately 2.6% of all new primary cancer cases [1-4] .

Kidney cancer incidence has been increasing steadily [5] , particularly in North America

[6] and Eastern Europe [7] where renal cell carcinoma (RCC) of the renal parenchyma accounts for more than 80% of kidney cancer diagnoses [2-4, 8] . The incidence rates of

RCC vary more than ten-fold around the world (Figure 1) [9] , with an eight-fold greater

incidence in developed countries than in developing countries [10, 11] . The highest

incidence rates (>20 cases/100,000 population) are observed in the Czech Republic [3, 8,

12] , while low rates (<8 cases/100,000 population) are reported in Asian countries [3, 8] .

1

Figure 1: Worldwide kidney cancer incidence rates from 1998 to 2000 [9]

These differences in incidence over time by country may reflect the increasing use of imaging modalities such as ultrasonography, computed tomography scans, and magnetic resonance imaging scans that are commonly used in developed nations, or these differences may reflect environmental factors [13] . It is important to note however, that imaging modalities alone do not entirely explain the increased incidence of RCC,

2 particularly since there has also been an increase in the diagnosis of advanced tumors; in fact, nearly half of patients today are diagnosed with advanced RCC [14] .

In the United States (U.S.), approximately 56% of kidney and renal pelvis cancer cases are diagnosed while the cancer is still confined to the primary site (localized stage) [15] ; furthermore, current incidence rates are estimated at 12.8 per 100,000 population (Figure

2) with a higher incidence among males than females [15] . Overall, RCC rates worldwide are approximately two times higher for males than for females [15] . Since the mid 1990’s, throughout most of the world, renal cancer mortality rates have stabilized [2,

8] . This is possibly due to improvements in 5-year survival rates which have been observed in the U.S. and in other countries [2, 14, 16, 17] . Overall, renal cell mortality rates for men are approximately twice those for women [15] .

3

Figure 2: U.S. incidence and mortality rates for kidney cancer 1975-2001 [15]

Kidney cancer is a multi-factorial disease, with both environmental and hereditary

components playing a role [18, 19] . Diet may be a factor in the pathogenesis of RCC,

though results for specific dietary components have been inconsistent, as seen in Table 1

[19-47] . Briefly summarizing, consumption of red meats has generally been associated with increased risk of RCC, [19, 21, 33, 35, 39, 42] while consumption of vegetables and in particular cruciferous vegetables has commonly been associated with decreased risk of the cancer [19-22, 30, 32, 33, 35, 36 39, 42-44].

4

Table 1: Epidemiological studies of dietary exposures and RCC risk [19-47] Author Study Design Sample Size Dietary Exposures Association (Year) Hu J Case-Control 767 RCC Cases Starch Significant increased association (2008) 1,534 Controls Vegetable, Unsaturated, & Polyunsaturated Fats Significant inverse association Linoleic Acid, Lenolenic Acid Significant inverse association Ward MH Case-Control 201 RCC Cases Nitrate in Drinking Water No association (2008) 1,244 Controls Hu J Case-Control 1,138 Kidney Alcohol Significant inverse association (2008) Cancer Cases Total Meat Significant increased association 5,039 Controls Fish, Poultry No association Hogervorst Prospective 4,382 Renal Acrylamide Significant increased association JG (2008) Cohort Cancer Cases 120,852 Followed Lee JE Pooled Analysis 1,478 RCC Cases Coffee, Tea Weak inverse association (2007) 13 Prospective 774,952 Controls Milk, Soda, Fruit & Vegetable Juice No association Cohorts Galeone C Case-Control 767 RCC Cases Vegetable Fiber Significant inverse association (2007) 1,534 Controls Fruit Fiber, Grain Fiber No association Hsu CC Case-Control 1,065 Kidney Milk, Meat, Yogurt Significant increased association. (2007) Cancer Cases Vegetables, Cruciferous Vegetables Significant inverse association 1,509 Controls Alcohol No association Bosetti C Case-Control 767 RCC Cases Flavones, Flavonls Significant inverse association (2007) 1,534 Controls Lee JE Prospective 248 RCC Cases Vitamin A & C, Carotenoi ds, Fruits & Vegetables Significant inverse association- (2006) Cohort 2,316,525 person- males yrs Vitamin E No association Bravi F Case-Control 767 RCC Cases Bread Significant increased association (2007) 1,534 Controls Milk & Yogurt, Pasta & Rice Weak increased association Coffee, Tea, Soups, Eggs, Red Meats, Fish, No association Pulses Cheese, Potatoes, Fruits, Sugars Wolk A Prospective 150 RCC Cases Fatty Fish Significant inverse association (2006) Female 940,357 person- Lean Fish No association Cohort yrs. Rashidkhani Prospective 93 RCC Cases Drinker: Wine, Hard Liquor, Beer, Snacks Weak inverse association B (2005) Female 46,572 Followed Western Diet: Sweets, Processed Meats, High No association Cohort Fat Dairy, Refined Grains, Margarine/ Butter, Fried Potatoes, Soft Drinks Healthy Diet: Fish, Fruit, Poultry, Vegetables, No association Whole Grains, Tomatoes van Dijk BA Case-Cohort 275 RCC Cases Vegetables, Fruits, Fruits & Vegetables No association (2005) 120,852 Followed Mandarin Significant increased association Hu J Prospective 1,279 RCC Cases Active Smoker, Passive Smoker Significant increased association (2005) Cohort 5,370 Followed Rashidkhani Prospective 122 RCC Cases Vegetables, Fruits, Fruits & Vegetables Weak inverse association B (2005) Female Cohort 61,000 Followed Mucci LA Case-Control 379 Kidney Dietary Acrylamide No association (2004) Cancer Cases 538 Controls Hu J Prospective 1,279 RCC Cases Vegetable, Vegetable Juice Significant inverse association (2003) Cohort 5,370 Followed Beef, Pork, Lamb Significant increased association Hamburgers, Processed Meats, Bacon, Sausage Significant increased association Cruciferous Vegetables, Dark Green Vegetables Significant inverse association- females Vitamin E, Calcium Significant increased association- females Vitamin E, Iron Significant inverse association- males Tomatoes, Fish, Chicken, Milk, Dairy, Cheese, No association Total Grains, Whole Wheat Grains Nicodemus Prospective 124 Kidney Vitamin E Significant inverse association KK (2004) Female Cancer Cases Vitamin C Significant increased association Cohort 34,637 Followed Alcohol Significant inverse association Handa K Case-Control 461 RCC Cases Beef, Unhealthy (high fat and protein) Diet Increased association- males (2003) 627 Controls Juice Diet, Dessert Diet Significant increased association- Males

5

SParker AS Case-Control 406 RCC Cases Alcohol Significant inverse association- (2002) 2,429 Controls Females Bianchi GD Case-Control 406 Kidney Tea No association (2000) Cancer Cases 2,434 Controls Augustsson Case-Control 138 Kidney Dietary Heterocyclic Amines No association K (1999) Cancer Cases 553 Controls De Steani E Case-Control 121 RCC Cases Red Meats, Barbecued Meats Significant increased association (1998) 243 Controls Dietary Heterocyclic Amines Significant increased association Yuan JM Case-Control 1,204 RCC Cases Cruciferous Vegetables, Carotenoids Significant inverse association (1998) 1,204 Controls Mellemgaard Case-Control 351 RCC Cases High Fat Diet Increased association A (1998) 340 Controls Fruits No association Cruciferous Vegetables Weak inverse association RCC: Renal Cell Carcinoma

Smoking, obesity and hypertension are recognized as primary risk factors that account for

nearly half of RCC diagnoses in the U.S. [48, 49] . The most consistently established

causal risk factor, accounting for approximately 20% to 30% of renal cell cancers, is

cigarette smoking [2, 50-52] ; a recent meta-analysis of five cohort studies and 19 case- control studies revealed that cigarette smoking was associated with a 1.54 fold and 1.22 fold increase in RCC risk among males and females, respectively [52] . Nearly every cohort and case-control study that has examined the association between obesity and

RCC risk has found an excess risk [2, 53-56] ; greater than 40% of U.S. and 30% of

European renal cell cancers have been attributed to obesity [2, 52, 54, 57] . Additionally, a recent quantitative review of published literature revealed a 3.7 fold increase in renal cancer risk among individuals with a body mass index (BMI) >40 kg/m 2 compared with individuals with a normal BMI (18.5-24.9 kg/m 2) [54] . Furthermore, another recent study that systematically reviewed epidemiologic studies published from 1985 to 2002 found between a 1.4 and 3.2 fold increase in RCC risk among hypertensive individuals compared with non-hypertensive individuals [58-66] . Even though approximately half of

RCC cases can be attributed to smoking, obesity, and hypertension, additional studies regarding diet, occupational exposures, and genetic risk factors that modify cancer

6 susceptibility may have the potential to improve our understanding of RCC etiology.

1.2 Vitamin D :

Vitamin D has long been recognized for its role in maintaining bone density and bone growth by regulating calcium and phosphorus levels in the blood by promoting the vitamin’s absorption from the intestine and reabsorption in the kidneys [67, 68]. Vitamin

D also plays a critical role in the maintenance and function of the nervous and organ

system [67] and has been shown to affect the immune system by promoting phagocytosis,

anti-tumor activity, and immunomodulatory functions [68] . Moreover, in recent years

there has been mounting scientific and public health interest in the role of vitamin D in

protecting against several types of cancers [6, 69-72] . In vitro and animal studies suggest that vitamin D and its metabolites may impede carcinogenesis by stimulating cell differentiation, inhibiting cell proliferation, inducing apoptosis, and suppressing invasiveness, angiogenesis, and metastasis [18, 73-75] .

Vitamin D is a fat soluble vitamin found in food and produced in the body after exposure to ultraviolet (UV) rays from the sun [76] . Very few foods, such as fatty fish, naturally contain significant amounts of vitamin D. Fortified foods, such as milk, butter, and cereal are common sources of vitamin D in the U.S. and Canada. However, in most countries, such as those in Central and Eastern Europe, foods are not fortified with vitamin D; individuals in these areas obtain most of their vitamin D though sunlight exposure [76] .

In general, UV exposure accounts for 90% of 1,25-dihydroxy (1,25(OH) 2) vitamin D levels, the biologically active form of vitamin D [77] .

7

UV rays from sunlight trigger vitamin D synthesis in the skin. Season, latitude, time of

day, cloud cover, smog, pollution, sunscreens, race and age all affect UV ray exposure

[78-80] . The use of sunscreen, even a weak sun protection factor of 8, can inhibit greater than 95% of vitamin D production in the skin [78, 81-83] . Furthermore, since melanin acts like a sun-block, individuals with darker pigmentation, such as African-American,

Hispanics and people with a Middle Eastern or Mediterranean heritage, require more sunlight [78, 84] . Among the elderly, vitamin D levels tend to decrease due to reduced dermal production of vitamin D and reduced outdoor activities that limit solar exposure

[79, 80] . On average, adequate amounts of vitamin D among Caucasians can be made in the skin with as little as ten to fifteen minutes of natural unprotected sun exposure at least two times per week to the face, arms, hands, or back without sunscreen [76, 78] .

Toxicity of vitamin D from sun exposure is very rare because with longer exposure to

UVB rays, equilibrium is achieved in the skin, and the vitamin simply degrades as fast as it is generated [76] . Similarly, normal food and pill vitamin D concentration levels are also too low to be toxic in adults [85] .

The active form of vitamin D is synthesized from precursors that derive from either diet or solar UV radiation exposure on the skin [78] . The ingestible forms of vitamin D are ergocalciferol (vitamin D 2) and cholecalciferol (vitamin D 3) [78, 84, 86] . Figure 3 shows how cholecalciferol can also be derived from the conversion of 7-dehydrocholesterol

(7DHC) in the skin by UVB radiation [86] . Both D 2 and D 3 are hydroxylated in the liver to the major circulating (or storage) form, 25-hydroxy (25(OH)) vitamin D and

8 subsequently in the kidney into 1,25(OH) 2 vitamin D, the active or hormonal form of the vitamin [78, 84] . In recent years, investigators have learned that hormonal vitamin D is also synthesized from 25(OH) vitamin D in some peripheral tissue, such as the skin, lymph nodes, colon, pancreas, and brain [87] . Since 25(OH) vitamin D levels are related to dietary intake and sun exposure, whereas hormonal vitamin D is homeostatically controlled, circulating 25(OH) vitamin D is considered reflective of vitamin D status in the body.

Figure 3: Synthesis of vitamin D in the body from diet & sunlight exposure [86]

9 The major circulating form of vitamin D has a half-life of approximately two to three

weeks and is the only metabolite that is used to determine whether an individual is

vitamin D deficient, sufficient, or intoxicated [88, 89] . The biologically active form of

vitamin D is not an ideal measure for vitamin D status because it has a circulating half-

life of only four to six hours and is also highly influenced by parathyroid hormone (PTH)

secretions [89] . When an individual becomes vitamin D deficient, decreased intestinal

calcium absorption results in the production and secretion of PTH. PTH regulates

calcium metabolism by increasing calcium reabsorption in the kidney, which

subsequently increases renal production of 1,25(OH) 2 vitamin D, making the metabolite

useless as a measure of vitamin D status [88, 89] .

Vitamin D deficiency, insufficiency, and toxicity are diagnosed by measuring the

concentration of 25(OH) vitamin D in serum. Vitamin D deficiency is defined as blood

serum 25(OH) vitamin D levels at or below 20 ng/ml, while vitamin D insufficiency

occurs when 25(OH) vitamin D serum levels are between 21 and 29 ng/ml [88, 89] .

Adequate vitamin D levels, defined by most physicians today, is obtained when blood serum 25(OH) vitamin D concentrations reach 30 ng/ml or higher [88, 89] . While there has never been a reported case of vitamin D intoxication from sun exposure, vitamin D toxicity is defined as 25(OH) vitamin D serum levels above 150 ng/ml that is associated with hypercalcemia, hypercalciuria and/or hyperphosphatema [89] .

1.3 Vitamin D Sunlight Exposure and Cancer Risk:

In 1942 the first ecological study investigating the association between cancer mortality

10 and solar radiation was reported by Apperly [90] . Observing higher cancer mortality risks among subjects residing at higher latitudes, Apperly stated, “the relative immunity to cancer is a direct effect of sunlight” [90, 91] . Four decades later, the relationship between solar UVB and cancer mortality rates was revisited by Cedric and Frank Garland

[92] and continues to be an issue of debate.

Renal cancer : Most studies of vitamin D UV exposure have been in relation to colon, breast, prostate, or lymphatic cancers. Only a limited number of studies, the majority being ecological, have investigated the association between vitamin D UVB exposure and renal cancer risk. While the inherent limitations resulting from the use of aggregate data on exposure and disease make it difficult to draw firm conclusions from these ecological studies, European and U.S. ecological studies investigating the association between multiple-cancer sites and sun exposure have generally reported an inverse relationship between kidney cancer incidence and mortality risk and UVB exposure [70, 93-96] . An inverse association between kidney cancer mortality rate and UVB radiation was reported in an ecological study that investigated cancer deaths from 1970 and 1994 [94] . Similar results were reported for kidney cancer mortality risk across European countries [95] ,

where potential confounding factors such as those related to diet and socioeconomics did

not affect associations. UVB irradiance was inversely associated with renal cancer

incidence rates (men p-value= 0.0003; women p-value= 0.04) in another ecological study

that examined UVB irradiance in 175 European countries. Renal cancer incidence rates

were highest in countries situated at the highest latitudes for both men and women [97] .

Further evidence of an inverse relationship between kidney cancer risk and UVB

11 exposure was reported by Boscoe and colleagues, who through the use of daily satellite-

measured solar UVB levels, investigated over three million incident cancer cases between

1998 and 2002 and cancer deaths from 1993 to 2002 in the U.S. [96].

Though limited, more convincing evidence for an association between sun exposure and

cancer risk may be drawn from case-control or cohort studies. An occupational cohort

study exploring the relationship between sunlight exposure and cancer risk among

Swedish male construction workers found increasing sun exposure to be significantly

associated with reduced kidney cancer risk [98] ; however, this study also looked at forty-

five different cancer sites and was limited in the number of subjects with kidney cancer.

Additional supporting evidence for kidney cancer risk and UV exposure was seen in a

cohort study of over 400,000 skin cancer cases and 3.5 million non-skin cancer controls,

which reported that increased vitamin D production in the skin decreased the risk of

several solid cancers, one of which was the kidney [99] . The protective effect of sun

exposure against second primary cancers, observed in 10,886 melanoma and 35,620 non-

melanoma skin cancer cases, was more prominent in sunny countries compared to less

sunny countries [99]. The authors of this study explained that although sun exposure was a risk factor for melanoma, it may also be a protective factor for internal cancers through the production of vitamin D [99] .

Occupational sunlight exposures may also contribute to adequate vitamin D levels. To date, no occupational studies have specifically investigated the association between sunlight exposure and renal cancer incidence risk. Furthermore, with the exception of

12 only one study [97] , all ecological reports on kidney cancer have explored ambient UV

exposure and multiple cancer sites. Due to the limited number of non-ecological studies,

the anti-carcinogenic potential of vitamin D, and the vital role that the kidneys play with

vitamin D activity, it is crucial that the potential association between occupational

sunlight exposure and renal carcinoma risk be explored.

Colorectal cancer : There are more epidemiological studies on sunlight vitamin D exposure and colorectal cancer than on any other cancer site, and the evidence for the sunlight-vitamin D-cancer hypothesis is seen as strongest for colon cancer [6]. Initial

support for the hypothesis came from an ecological study that reported an inverse

correlation between statewide colon cancer mortality and solar radiation [100] .

Subsequent ecological studies have noted inverse correlations between colon and rectal cancer incidence/mortality and various sunlight-related measures [96, 101, 102] ; however, not all studies have been consistent with regards to gender, race, and study location [96, 102] .

Individual sun exposure has been assessed in a few case-control studies. The only study that estimated residential and occupational sun exposure found a reduced risk in colon cancer mortality with increasing indications of sun exposure [103] . Of the two studies

that assessed individual time in the sun (based on a single year), one found no

relationship between self-reported sun exposure and risk of colon cancer in men or

women [104] , while the other reported reduced cancer risk among those diagnosed before

age sixty but not among those diagnosed after sixty years, with the association suggested

13 in men but not women [105] . To date, no cohort studies have examined the relationship between sunlight exposure and colon cancer.

Breast cancer : In several ecological analyses, breast cancer mortality and incidence rates have been inversely associated with surrogates for residential sunlight exposure, such as average annual sunlight energy by region [94, 103, 106, 107] . However, the Nurses’

Health Study, [108] which represents the single largest cohort study of women, failed to find the expected regional gradient. The only prospective breast cancer incidence study explicitly designed to assess individual sun exposure, the National Health and Nutrition

Examination Survey (NHANES) I Epidemiologic follow-up study, found that various

measures of sun exposure, such as diagnosed actinic skin damage, and self-reported

recreational and occupational sun exposure, were associated with reduced breast cancer

risk, although protective effects were largely limited to regions of high solar radiation

[109] . Regional disparities were also noted in a large U.S. case-control study based on

death certificates which assessed occupational sun exposure based on “usual”

occupations [103] . The study found a significant negative association for non-farming

outdoor jobs, particularly in the regions of highest solar radiation. The negative

association was notably characterized among both white and black women who died from

breast cancer.

Prostate cancer : In 1990, Schwartz and Hulka [110] hypothesized that vitamin D

deficiency was a risk factor for prostate cancer based on the association between prostate

cancer and several indicators of vulnerability to low vitamin D status, such as dark skin,

14 age, and northern latitude in the U.S. Subsequent ecologic data have been consistent with

the hypothesis. A study by Hanchette and Schwartz found a 50% mortality difference

from northern to southern U.S. counties for deaths in 1970-79 [111] . In 2006, Schwartz and Hanchette extended the ecologic analysis over a 45-year period (1950-94) and showed that the geographic distribution of prostate cancer mortality in Caucasians was inversely related to UV radiation [112] .

Unlike most other cancer sites, there have been several studies of prostate cancer risk in relation to individual UV exposure. The only prospective study, a study of U.S. white men [113] , found significant inverse associations between incident prostate cancer and various indicators of high residential solar radiation. The cohort study did not, however, show an association between recreational and occupational sun exposure and prostate cancer risk [114] . In a death certificate based case-control study, Freedman and associates [103] also found surrogates for residential (but not occupational) sun exposure associated with a significantly decreased risk of prostate cancer mortality.

Several prostate cancer case-control studies in the United Kingdom [115-117] that

examined more specific indicators of sun behavior found protective associations with

childhood sunburns, adult sunbathing score, and higher cumulative UV radiation

exposure (comprising recreational and occupational exposure) based on questionnaire

responses. Another large case-control study [118] of advanced prostate cancer used a

reflectometer to rank relative cumulative sun exposure by comparing skin pigmentation

in sun-exposed and sun-protected areas. Substantial reduced risks of advanced prostate

15 cancer were associated with high cumulative sun exposure based on the reflectometer

measurements and with self-reported occupational sun exposures.

Lymphoma : Ecologic studies exploring the relationship between UV exposure and lymphoma risk have been inconsistent. UV indices and surrogates such as latitude were positively related to lymphoma risk in Europe [119-121] but not in the U.S. [95, 122-

124]. Recently, three large population-based case-control studies from Australia [125] ,

Sweden and Denmark [126] , and the U.S. [127] examined personal sunlight exposure and

lymphoma risk. Each study found moderately lower risks for more time in the sun. Sun

vacations [125, 126] were also linked to lower risk. In the U.S. study, no clear pattern of

risk was seen for protective clothing (in this case, hats) and sunscreen use [127] . Because

hats block only a small measure of potential total body exposure and sunscreen use may

be related to higher time outdoors, these findings may not be informative about overall

sun exposure and risk. The Australian and Scandinavian studies addressed occupational,

as well as recreational, sun exposure [125, 126] , but neither found a protective

association with outdoor work.

Several other studies have examined occupational UV exposure and non-Hodgkin

lymphoma (NHL) risk [98, 122, 128-133] , with most exposure assessments based on

work history and job titles or tasks. The sole study to assess individual self-reported

occupational exposure [132] , only obtained dichotomous information regarding whether

the worker had been exposed to UV on the job (yes, no) and the duration of outdoor work

in years. Three studies found no association with UV [130-132]; one found a modest, but

16 non-significant, elevated risk in the high-exposure group [98] ; and third study [122] found a somewhat lower cancer risk. In addition, one study [133] observed no association in men but an elevated risk in women, apparently due to farming, which has been linked to non-sun risk factors (e.g., pesticides) [134] .

1.4 Dietary Vitamin D Intake and Cancer Risk:

Dietary sources of vitamin D are limited and include only a few natural food sources, such as fatty fish like salmon, herring or mackerel, as well as fortified foods such as milk and cereals. As aforementioned, although food sources in the U.S. are commonly fortified with vitamin D, throughout most parts of the world this is not a common practice.

Vitamin supplements also constitute another source of vitamin D intake and have rarely been examined. While dietary vitamin D intake accounts for less than 10% of vitamin D levels [135], scientific studies suggest that dietary intake of vitamin D may play an important role in determining cancer risk.

Based on numerous studies, in 1997, the National Academy of Sciences-Institute of

Medicine’s Food and Nutrition Board revised its recommended daily intake of vitamin D to 200 International Units (IU) for children, adolescents, and adults up to age 50 years,

400 IU for adults aged 51-70 years, and 600 IU for those aged 70 years or older [136].

The revised levels, shown in Table 2, reflected a new understanding of vitamin D's role in preventing bone loss and maintaining bone health; however, many experts believe the new recommendations are still inadequate for preventing other conditions associated with low vitamin D, such as high blood pressure, cancer, chronic pain, depression,

17 schizophrenia, seasonal affective disorder [137] and several autoimmune diseases including rheumatoid arthritis, scleroderma, and type 1 diabetes [81, 138] . In fact, in

2008 the American Academy of Pediatrics issued recommendations for vitamin D intake in children that were typically double that of the National Academy of Sciences and based on evidence from clinical trials [139] . Furthermore, the Canadian Pediatric Society currently recommends 2000 IU/day of vitamin D for pregnant and breastfeeding females,

400 IU/day for all babies who are exclusively breastfed, and 800 IU/day for babies residing in areas that are above 55 degrees in latitude [140] .

Studies investigating the relationship between dietary vitamin D intake and kidney cancer are limited. To date, there are no studies that have looked at vitamin D supplement intake and renal cancer risk and only one epidemiological study has explored the relationship between dietary micronutrient vitamin D intake and RCC risk [141] . A statistically significant inverse association (Odds Ratio (OR)= 0.79; 95% Confidence Interval (CI)=

0.63-0.98) between dietary vitamin D intake, based on a 78-item food frequency questionnaire and renal cancer risk was observed in an Italian case-control study of 767

18 cases and 1,534 controls [141] . Dietary vitamin D intake was more strongly associated with reduced cancer risk among females (OR= 0.60; 95% CI= 0.42-0.88) when analyses were stratified by sex [141] .

The associations between dietary vitamin D intake and other cancer sites, particularly cancers of the colon, breast, prostate, ovary and NHL, have been well explored.

Increased dietary vitamin D intake was associated with reduced risk of colon cancer in two European case-control studies [142, 143] . Yet, no relationship with colon cancer risk was observed in two other European [144, 145] or three U.S. case-control studies [104,

146, 147]. Cohort studies have also generally found no association for colorectal [148-

152], colon [153-157], or rectal [153, 156, 158] cancers. However a convincing meta-

analysis of 63 published reports observed that daily dietary intake of an additional

1,000 IU of vitamin D reduced colon cancer risk by almost 50% [159] . Furthermore,

vitamin D intake from supplements has generally been associated with decreased

colorectal cancer risk in both case-control and cohort studies [104, 146, 153, 154, 160] .

Two prospective breast cancer studies have examined risk associated with dietary and

supplemental vitamin D. The Nurses’ Health Study found an inverse association between

total and dietary vitamin D intake and breast cancer risk among pre-menopausal women

[161]. In the Cancer Prevention II Nutrition Cohort, a study of postmenopausal women,

total and dietary vitamin D intake was unrelated to overall breast cancer risk [162] . But,

inverse associations between dietary, but not total, vitamin D, and

positive tumors were observed. Vitamin D intake was also assessed in the NHANES I

19 Epidemiologic follow-up study [109] , which reported non-significant reduced cancer risk with increasing dietary vitamin D intake, though results were limited by small case numbers and low intake levels. Moreover, a recently publish meta-analysis restricted to studies that had explored higher levels (>400 IU/day) of dietary vitamin D intake and risk of breast cancer reported a statistically significant inverse association [163] . A second and larger meta-analysis, reported that dietary intake of an additional 1,000 IU of vitamin

D reduced breast and ovarian cancer risks by almost 30% [159]. Yet, no associations

with dietary vitamin D and breast cancer were observed in several case-control studies

[164-166] , some of which were small, nor were they found in retrospective studies of

dietary vitamin D consumed in high school [167, 168] , which relied on inherently

difficult recollections of diet many years earlier.

Dietary vitamin D studies have largely been unsupportive with regards to reduced

prostate cancer risk and inconsistent with regards to ovarian cancer risk [114, 169-182].

A prospective study of 47,781 men in the Health Professional Follow-Up Study reported

vitamin D intake was not associated with risk of total prostate cancer for high versus low

intake, nor with advanced prostate cancer [176] . These analyses were based on 1,369 total cases and 423 advanced prostate cancer cases. A meta-analysis of 26,769 cases in

45 observational studies found dietary intake of vitamin D was not related to prostate cancer risk [182] . Yet, beneficial effects of high levels of vitamin D intake have been reported in clinical trials of patients with advanced prostate cancer [183] .

A large pooled analysis of cohort studies with more than 2,000 epithelial ovarian cancer

20 cases evaluated associations with vitamin D. A non-significant higher cancer risk was

associated with dietary vitamin D intake, but not with total vitamin D intake (including

supplements), or with fish and cereal, key sources of dietary vitamin D, thus undermining the causal link to vitamin D [178] . Of the four case-control studies examining dietary vitamin D intake, a small Mexican study [184] found a protective effect, whereas three other much larger U.S. [179, 180] and Italian studies [181] did not, possibly due to

different food patterns related to vitamin D in the countries (fish vs. milk). Three Italian

case-control studies [185-187] found fish consumption associated with reduced ovarian

cancer risk [186] . Thus, the evidence of an association between ovarian cancer and

vitamin D is quite weak, largely due to the paucity of non-dietary studies.

Of the four studies that have examined dietary or total vitamin D intake and NHL risk

[127, 188-190] , all of which were retrospective, only one, an Italian hospital-based case-

control study [189] observed an association, which was protective. Fish, which has been

linked to lower risks of NHL in several studies [187, 188, 191] , and is identified as the

chief source of dietary vitamin D in the Italian study [189] , may explain the association.

The sole study to look at total dietary intake (including supplements) as well as food

[127] , found no association with vitamin D.

1.5 Common Vitamin D Receptor (VDR) Genes and Cancer Risk:

The conversion of vitamin D to its main biologically active form occurs in the kidney;

where the vitamin mediates its biological effect by binding to vitamin D receptors ( VDR )

[75, 192] . VDR are transcriptional factors that are part of the nuclear hormone receptor

21 family which influence the behavior of genes involved in cell regulation, growth, and

immunity [75, 192, 193] . Given the potential impact that the vitamin D gene and its pathway have on vitamin D levels, analysis of genetic variation in VDR and other vitamin

D genes may elucidate the role of vitamin D in RCC etiology.

Epidemiological studies investigating the association between variations in the vitamin D receptor ( VDR ) gene and cancer risk have primarily focused on five polymorphisms: 1) the rs2228570 or FokI (EX4+4T>C) on exon 4; 2) rs1544410 or BsmI (IVS10+283G>A) on intron 10; 3) rs731236 or TaqI (Ex11+32T>C) on exon 11; 4) rs7975232 or ApaI

(IVS10-49G>T) on intron 10; and 5) the poly(A) mononucleotide repeat at the 3 ′ untranslated region (UTR) section of the gene. Depending on the location, different polymorphisms may have different functions [194, 195] .

The FokI polymorphism is near the 5 ′ UTR of the gene within the deoxyribonucleic acid

(DNA)-binding domain. FokI is not linked to other VDR polymorphisms and alters the

transcription initiation site of the VDR protein. The protein associated with the FokI F

variant is three amino acids shorter than the f allele and functions with higher activity as a

than the wild-type ( ff ) protein [196, 197] ; evidence from previous

studies indicate the FokI f variant to be associated with decreased cancer risk [198, 199] .

The other polymorphisms are close to the 3 ′ UTR region of the gene and like many tagging single nucleotide polymorphisms (SNPs), BsmI, ApaI, TaqI, and the poly(A) monoculeotide repeat polymorphisms are not known to have functional relevance, yet are thought to be in linkage disequilibrium (L D) with functional variants in the 3 UTR,

22 which modify messenger ribonucleic acid (mRNA) stability [200] . While many allelic variants of the VDR gene occur naturally in the human population, the allelic frequencies for these SNPs appear to vary by race and ethnic group [194, 195] . The expression of

VDR variants are associated with a number of conditions, including autoimmune diseases, resistance to vitamin D therapy, susceptibility to infections, and cancer [194] .

Though limited, only a few studies have investigated the association between genetic

variations and common VDR variants and kidney cancer. In a recently published large

European case-control study, no association between RCC risk and TaqI , FokI , or BsmI

polymorphisms were observed [201] . Similarly, a Japanese case-control study of 135

RCC cases and 150 controls, found no association between the TaqI and BsmI variants

and RCC risk; however, participants with the AA ApaI allele had a significant increased

risk of renal cancer [202] . Previously, Ikuyama and collogues had reported increased risk

of rapid growth renal cancer among Japanese participants with the TT TaqI allele [203] .

By far, most epidemiological studies investigating the association between common VDR

gene variants and cancer risk have been in relation to breast and prostate cancer risk.

Numerous epidemiological studies have investigated the association between common

VDR polymorphic variants and risk of prostate cancer (Table 3), with most focusing on

the TaqI and FokI variants. In general, an observed 70% to 80% reduction in prostate

cancer risk has been reported among individuals who were carriers of the tt TaqI genotype [118, 204-207] . However, the association between this variant and prostate cancer risk has not been confirmed in all studies [118, 208-216] . Results for the FokI

23 allele have not been so consistent. Among carriers of the ff FokI genotype, African

Americans [217] had a reduced risk of prostate cancer in one small study, while this

association was only reported for aggressive tumors in two other studies [198, 218] .

Subsequently, carriers of the ff FokI variant had increased prostate cancer risk in a recent case-control study [219] , while this association was only observed for aggressive tumors in a large prospective cohort study [220] . Seven other epidemiological studies, however, reported no association with prostate cancer risk and the FokI variant [119, 207-212 ] . To date, genetic susceptibility studies investigating the association between prostate cancer risk and the poly(A) allele have all been null [211, 219, 221, 222]. Null results were also reported in the majority of studies that had investigated the association between the ApaI

[223-227] and BsmI [211, 213, 217, 219, 223, 226] variants and prostate cancer risk.

Table 3: Common VDR variants and risk of prostate cancer [118, 198, 204-227]

Author- Study Sample VDR SNPs Association Year Design Size Mikhak B Nested 704 Cases FokI f vs. F significant decreased (2007) Case- 51,529 risk of aggressive tumor Control Controls BsmI b vs. B significant decreased risk of aggressive tumor Li H Prospective 1,066 Cases FokI ff vs. FF significant (2007) Cohort 14,916 increased risk of aggressive Controls tumor Cicek MS Case- 439 Cases BsmI, poly(A), No association (2006) Control 479 TaqI A vs a significant decreased Controls ApaI risk FokI FF vs ff significant decreased risk

24 Anderson Case- 137 Cases TaqI No association P Control 124 (2006) Controls Huang SP Case- 416 Cases FokI FF vs. ff significant (2006) Control 502 increased risk of Controls aggressive tumor Chaimuan Case- 28 Cases BsmI, ApaI, No association graj S Control 30 Controls TaqI (2006) John EM Case- 405 Cases FokI, TaqI No association (2005) Control 455 Controls Mishra Case- 128 Cases FokI No association DK Control 147 (2005) Controls Maistro S Case- 165 Cases ApaI, TaqI No association (2004) Control 200 Controls Oakley- Case- 354 Cases BsmI, ApaI, No association Girvan Control 292 TaqI F vs. f significant increased I (2004) Controls FokI risk in African Americans Yang Y Case- 80 Cases FokI No association (2004) Control 90 Controls Tayeb MT Case- 28 Cases TaqI TT vs. tt significant (2004) Control 56 Controls FokI increased risk FF vs. ff increased risk Cheteri Case- 559 Cases BsmI bb vs. BB significant MB Control 523 increased risk of localized (2004) Controls FokI, poly(A), tumor BsmI No association Huang SP Case- 160 Cases ApaI, TaqI No association (2004) Control 205 BsmI BB vs. bb significant Controls decreased risk Bodiwala Case- 368 Cases FokI, TaqI No association D Control 243 (2004) Controls

25 Suzuki K Case- 81 Cases BsmI, TaqI, No association (2003) Control 105 ApaI Controls Tayeb MT Cohort 21 Cases TaqI No association (2003) 379 Controls Medeiros Case- 163 Cases TaqI T vs. t significant increased R Control 211 risk 2002 Controls Hamasaki Case- 110 Cases TaqI TT vs. tt significant T Control 90 Controls increased risk 2002 Gsur A Case- 190 Cases TaqI TT vs. tt increased risk (2002) Control 190 Controls Chokkalin Case- 191 Cases BsmI, FokI No association gam AP Control 304 (2001) Controls Hamasaki Case- 115 Cases TaqI TT vs. tt significant T Control 131 increased risk in metastatic (2001) Controls disease Bousema Case- 98 Cases TaqI No association JT Control 61 Controls (2000) Habuchi T Case- 222 Cases TaqI, ApaI No association (2000) Control 128 BsmI bb vs. BB significant Controls increased risk Blazer Case- 77 Cases poly(A), TaqI No association DG Control 183 (2000) Controls Watanobe Case- 100 Cases poly(A), TaqI No association M Control 202 (1999) Controls

26 Over the past decade, at least fourteen studies have investigated the association between

common VDR variants and breast cancer risk (Table 4). Recent findings reported an inverse association with breast cancer risk for women homozygous for the aa ApaI [228,

229] , BB BsmI [230-233] , and short poly(A) [232, 234] alleles, although these findings have not been universal [229, 230, 235-237] . Associations for FokI and TaqI have generally been null for breast cancer risk [77, 229, 232, 234, 236-239] . However, a large nested case-control study within the prospective Nurses’ Health Study reported increased breast cancer risk among women who were carriers of the ff FokI allele [240] . In two

studies, VDR genotypes were not associated with overall cancer risk, yet were associated

with metastatic status, suggesting that common VDR allelic variants may influence tumor

progression [233, 239] .

Table 4: Common VDR variants and risk of breast cancer [77, 98, 228–240]

Author-Year Study Sample Size VDR SNPs Association Design Abbas S Case- 1,408 Cases FokI No association (2008) control 2,612 TaqI No association Controls Trabert B Case- 1,631 Cases BsmI bb vs. BB significant (2007) Control 1,435 increased risk Controls poly(A) No association John EM Case- 1,788 Cases FokI No association (2007) Control 2,129 TaqI No association Controls McCullough Nested 500 Cases BsmI, TaqI, No association ML Case- 500 Controls ApaI, FokI, (2007) Control poly(A) Chen WY Nested 1,234 Cases FokI ff vs. FF significant (2005) Case- 1676 Controls increased risk Control BsmI No association Lowel LC Case- 179 Cases BsmI bb vs. BB significant (2005) Control 179 Controls increase risk

27 Guy M, Case- 398 Cases poly(A) LL vs. SS significant (2004) Control 427 Controls increased risk BsmI BB vs. bb significant increased risk FokI No association

Sillanpaa P Case- 483 Cases ApaI aa vs. AA significant (2004) Control 483 Controls decreased risk TaqI T vs t non-significant decreased risk Hou MF Case- 80 Cases ApaI AA vs. aa significant (2002) Control 169 Controls increased risk TaqI No association BsmI B vs. b increased risk Bretherton- Case- 181 Cases BsmI bb vs. BB significant Watt D Control 241 Control increased risk (2001) poly(A) L vs. S significant increased risk FokI No association Ingles SA Case-Cohort 143 Cases BsmI B vs. b significant (2000) 300 Controls increased risk poly(A) SS vs. LL significant increased risk Curran JE Case- 135 Cases ApaI aa vs. AA significant (1999) Control 110 Controls increased risk TaqI, FokI No association

1.6 Vitamin D Pathway Genes and Cancer:

In view of the fact that 1,25(OH) 2 vitamin D, the active form of the vitamin, exerts its activity through the intercellular vitamin D receptor, most epidemiological studies have primarily focused on the VDR gene [241] . However as shown in Figure 4, numerous genes are involved in the vitamin D pathway, which individually or as a whole may influence vitamin D metabolism, transport, binding, function and/or expression and therefore may also be important in modifying the mechanisms through which vitamin D influences cancer risk.

28

Figure 4: Genes involved in the vitamin D pathway [202, 203, 242-266]

The group specific component (GC) vitamin D binding protein serves as the major carrier of vitamin D and its metabolites in plasma to target tissue [242, 243] . Renal uptake and activation of 25(OH) vitamin D occurs primarily in the proximal tubules of the kidney, where reports have shown that 25(OH) vitamin D is specifically targeted to the proximal tubule through GC action [242, 244-246] . To date only two epidemiological studies have investigated the role of GC and RCC, both with conflicting results [247, 248] . Moreover, no human epidemiological studies have investigated the role of RCC and 24- dehydroxylase ( CYP24A1 ), an enzyme involved in the metabolism of vitamin D in target tissue; CYP24A1 degrades and regulates 1,25(OH) 2 vitamin D levels [249-251] . In the

kidney, inactivation of vitamin D is attributed to CYP24A1 , which is induced

transcriptionally by 1,25(OH) 2 vitamin D whose action is mediated by binding to VDR

[252] .

29

The biological activity of vitamin D is mediated by a high-affinity receptor, VDR which

acts as a ligand-activated transcription factor that forms a heterodimer with retinoid-X-

receptors ( RXR ) [253, 254] ; this VDR-RXR heterodimer complex is directed to the vitamin D-responsive element ( VDRE ) in the promoter region of 1,25-regulated genes

[72, 253, 254] . Different polymorphisms in the VDR gene have been speculated to result

in variation of VDR expression and therefore to further result in changes to circulating

levels of active vitamin D [203] . A number of epidemiological studies have reported that

increased binding of vitamin D to VDR is associated with decreased RCC risk and that

active levels of vitamin D are significantly lower in RCC patients [202, 203, 255, 256] .

While results have been mixed concerning associations between RXR gene

polymorphisms with cancer risk on different cancers [257, 258] , only two

epidemiological studies to date have examined the association between RXR expression

patterns and RCC [257, 259] . The first study of 63 RCC patients found participants with

atypical RXR-alpha (RXRA ) subcellular distribution had a significantly lower survival

compared with participants with predominantly nuclear localization of RXRA [257] . The

second study, which included 68 RCC cases, found no association between RXRA

expression levels and renal cancer risk [259] .

Moreover, interactions have been reported between VDR , RXR and the vitamin D

receptor interacting protein ( DRIP ) complex, which includes the thyroid hormone

receptor associated protein ( THRAP ) genes [260, 261] . THRAP’ s involvement in RCC

carcinogenesis and its interactions with vitamin D pathway genes affect VDR

30 transcription, making this gene important in elucidating our understanding in RCC

etiology [262] . To date only two epidemiological studies have investigated the role of

THRAP and RCC, both reporting aberrant expression of the gene in RCC patients [262,

263]. Studies regarding the interaction between VDR and the signal transducer and activator of transcription ( STAT1 ) gene have also shown inhibition of the VDR-RXR complex to VDRE , thus impeding transcription [264-266] .

Given the limited number of epidemiological studies and the importance that these genes play in vitamin D metabolism, binding, and function, it is central that epidemiological studies with thorough coverage and fine mapping techniques be conducted to help clarify how genetic variations in the vitamin D pathway may affect cancer risk, particularly RCC risk.

1.7 Public Health Impact:

Mass media approaches have been widely used in public health programs to address behavioral risk factors, particularly strategies aimed at reducing sun exposure in efforts to prevent skin cancer. While it is well accepted that reduced sun exposure can decrease the risk of cataracts, skin cancers, and melanoma [267] , UV prevention efforts have also been partly to blame for the epidemic rates of vitamin D deficiency [81, 268-271] . Throughout the U.S. and Europe, epidemic rates of vitamin D deficiency/insufficiency continue to be a problem for all age groups because of inadequate vitamin D intake and decreased sun exposure [272] . Over the past 25 years, the prevalence of vitamin D deficiency has increased throughout many parts of the world [273]. A recent report on vitamin D levels

31 throughout the world reported that over one billion people worldwide are currently

vitamin D deficient [88].

Reduced UV exposure resulting in decreased vitamin D levels has been associated with

increased risk of auto-immune and bone diseases, as well as different cancers [274-276] .

Therefore, epidemiological UV studies are needed to determine exactly how much sun exposure and dietary vitamin D intake is required for adequate vitamin D levels. Public health messages that accurately detail the benefits and risks of sun exposure will empower the public to select individual levels and patterns of sun exposure that enhance vitamin D levels, while avoiding the risks of excess exposure. Alternatively, the solution may be in the fortification of foods, a policy driven intervention that would both increase vitamin D levels among consumers while reducing the risk of other diseases associated with increasing UV exposure.

Even with recent mounting evidence regarding the beneficial role of vitamin D, epidemiological studies of the vitamin are quite limited for many sites, such as that for the kidney, and far from consistent. Given the potential cancer prevention promise, it is important to investigate the relationship between vitamin D and cancer risk. Renal cell carcinoma is of particular interest since the kidney is a major organ for vitamin D metabolism and activity, and calcium homeostasis and since incidence rates of RCC have increased steadily over the past two decades in the U.S. [6, 81, 277] and globally [277].

32 Chapter 2: Study Hypothesis, Specific Aims and the Central & Eastern European

Renal Cell Carcinoma (CEERCC) Study

2.1 Hypothesis and Specific Aims:

The relationship between vitamin D and renal cancer still remains unexplored. Globally, epidemic rates of vitamin D deficiency continue to be a problem for all age groups due to inadequate vitamin D intake and decreased sun exposure [88, 272, 273] . Additionally,

renal cancer incidence continues to increase worldwide with half of all diagnoses

unexplained [6, 81, 277] . Currently, it is recognized that vitamin D deficiency is associated with numerous cancers and may be associated with kidney cancer since the kidney is a major organ for vitamin D metabolism and activity, and calcium homeostasis

[6, 69-72] . The anti-carcinogenic properties of vitamin D, which include inhibition of clonal tumor cell proliferation, hematopoieses, induction of immune cell differentiation, and apoptosis, further support the value of investigating the relationship between this vitamin and renal cancer [18, 68, 73-75] . Lastly, the lack of non-ecological studies that have investigated the relationship between renal cancer risk and vitamin D, either through ultraviolet (UV) B exposure, dietary intake, or genetic variation of genes involved in the vitamin D pathway, may have the potential to improve our understanding regarding the role of vitamin D in renal cell carcinoma (RCC) etiology.

Therefore, the current study hypothesized that increased vitamin D exposure (via occupational sunlight exposure or dietary intake) was associated with decreased RCC risk and that genetic variation within the vitamin D pathway modified this risk.

33 Demographic, occupational, and genomic data from the Central and Eastern European

Renal Cell Carcinoma (CEERCC) Study was used to test this study hypothesis.

Specifically, the aims of this study included:

(1) Investigating whether occupational UV sunlight exposure was associated with

lower RCC risk by examining occupational histories of subjects and estimating

potential sunlight exposures through the use of a job exposure matrix (JEM).

(2) Exploring potential gene-environment interactions between occupational sunlight

exposure and polymorphisms in vitamin D pathway genes to determine if genetic

variants modified associations between occupational UV exposure and RCC risk.

(3) Examining the association between genetic variants in the vitamin D pathway

(CYP24A1, GC, RXRA, RXRB, STAT1, THRAP4, THRAP5, and VDR) and RCC

risk, both overall and within particular case subgroups.

(4) Assessing potential gene-environment interactions between dietary intake of

vitamin D and calcium rich foods and polymorphisms in vitamin D pathway

genes to determine if genetic variants modified associations between RCC risk

and dietary vitamin D and calcium intake.

2.2 The Central and Eastern European Renal Cell Carcinoma (CEERCC) Study:

The CEERCC Study is funded by the National Cancer Institute (NCI) and the

34 International Agency for Research on Cancer (IARC). CEERCC is a hospital-based

case-control study of renal cell cancer conducted between 1999 and 2003 in seven centers

in four Central and Eastern European countries (Moscow, Russia; Bucharest, Romania;

Lodz, Poland; and Prague, Olomouc, Ceske-Budejovice, and Brno, Czech Republic).

This region has among the highest incidence of RCC in the world [3, 12]. Each center followed an identical protocol and was responsible for recruiting a consecutive group of newly diagnosed cases of kidney cancer as well as a comparable group of hospital controls.

Cases were residents of one of the above named study areas for at least one year and were newly diagnosed with kidney cancer (International Classification of Diseases for

Oncology (ICD-O) codes C64 and C65) at participating hospitals. Cases between 20 and

88 years of age were ascertained through a rapid reporting system, whereby participating physicians and other professional staff members visited the urology, surgery, radiology, or pathology departments of participating hospitals to regularly identify patients admitted for kidney cancer work-up. All patients suspected of having kidney cancer were approached for interview, but only histologically confirmed cases were included in the final analysis. The majority of renal tumors were clear cell carcinomas (83.4%) while other subtypes included papillary (7%), chromophobe (2.4%) oncocytoma (2.3%) oncocytic neoplasms (0.2%) transitional cell carcinomas (1.1%) and unclassified (3.6%).

Controls in all centers were chosen among subjects admitted as inpatients or out-patients in the same hospital as the cases, with non–tobacco-related conditions, including

35 infections (1.1%), hematologic (3.2%), endocrine (2.0%), psychiatric (1.4%), neurologic

(11.2%), ophthalmologic or otologic (14.5%), cardiovascular (9.6%), pulmonary (3.9%), gastrointestinal (18.7%), dermatologic (2.8%), orthopedic or rheumatologic (8.9%), genitourinary (benign prostatic hyperplasia) (3.8%), obstetric or perinatal (0.1%), injury

or poisoning (3.0%), and other (15.9%). Some of the controls were also recruited in

parallel for studies of lung [278] and head and neck cancer [279] . No single disease made up more than 20% of the control group. The entire group of controls was frequency-matched to multiple case series by sex, age (+/- 3 years), study center, and referral (or residence) area. Although controls had to be free from cancer at time of enrollment, previous history of cancer was not an exclusion criterion in either cases or controls.

Both cases and controls underwent an identical interview with the same questionnaire.

Cases were interviewed within three months of diagnosis. Standardized lifestyle and food frequency questionnaires were piloted in all centers prior to use and were administered in-person by trained personnel to elicit information on demographic characteristics, education, exposure to tobacco smoke, alcohol consumption, dietary practices, anthropometry, medical history, family history, and occupational history.

Written consent for participation was obtained from all study subjects, and ethical approval was obtained for all study centers as well as at the IARC and the NCI.

Individuals for whom information was missing on diet (N= 6) or alcohol consumption

(N= 11) were excluded. Also excluded were those with missing covariates (age, sex,

36 tobacco use, hypertension medication use, body mass index (BMI), and education (N=

61)). All participants were of Caucasian descent. The response rates across study centers among eligible subjects who were requested to participate ranged from 90.0% to 98.6% for cases and from 90.3% to 96.1% for controls. The final number of participants for analysis in this study included 1,097 renal cancer cases and 1,476 controls.

This study is unique in that detailed information on occupational exposure histories was collected and assessed by trained industrial hygienists. Lifetime occupational information was collected during interviews through the use of a general occupational questionnaire and job-specific questionnaires. For jobs held for at least 12 months duration, data were collected on the title, the detailed tasks, the company, and the year of beginning and ending employment. Job titles were coded by local industrial hygienists or occupational physicians, blinded to case-control status, according to the International

Standard Classification of Occupation 1968 version (ISCO-68) [280, 281] ; industries were similarly coded according to the Statistical Classification of Economic Activities of the European Community, 1999 version (NACE-99) [281, 282] . Both the ISCO and

NACE coding systems are regularly utilized in occupational epidemiological studies conducted throughout Europe for occupation and industry classifications [283] .

Moreover, intensity, frequency, and confidence of 74 different occupational exposures were then assessed for each subject’s jobs by occupational health experts and industrial hygienists.

A food frequency questionnaire was comprised of 23 food items, which the study

37 investigators selected by consensus during the planning stage of the study and further validated during the pilot stage by asking participants to name common food items not

already specified. Frequency of consumption was assessed for each item as never, less

than once per month, less than once per week, one to two times per week, three to five

times per week, and daily. A standardized questionnaire was used in each of the study

centers that was translated from a common English version and then back-translated into

English to ensure the validity of the initial translation. The questionnaire was repeated for

two different time periods: 1) the year prior to interview, and 2) the year prior to political

and market changes in 1989 (1991 in Russia). Data for the year prior to interview and the

year prior to political change were then extrapolated to represent lifetime average dietary

intake by multiplying the score for each time period by the number of years the

participant was alive during the time period, then summing the time period scores and

dividing by the total age of the individual.

This study is extraordinary in that it is the largest kidney cancer study conducted to date

with biologic samples. A subset of 925 (84.2%) newly diagnosed and histologically

confirmed RCC (ICD-O-2 codes C64) cases and 1,192 (80.7%) controls provided

genomic deoxyribonucleic acid (DNA), which was successfully genotyped for the

selected genotyping panel. Blood samples were stored at -80°C and shipped to the NCI

on dry ice.

DNA for genotyping assays was extracted from buffy coat and whole blood samples

using phenol-chloroform extraction. DNA from RCC cases and controls were

38 randomized on polymerase chain reaction (PCR) plates for genotyping. Analyses for

genotyping were blinded, while genotyping of a randomly selected 5% duplicate samples was conducted for quality control. Genotyping of genes was conducted at IARC and at

NCI’s Core Genotyping Facility (CGF). Methods for all genotype assays can be found at: http://snp500cancer.nci.nih.gov/home.cfm [284] . Genomic analysis of single

nucleotide polymorphisms (SNPs), selected based on evidence of functional relevance or

those leading to amino acid sequence changes were conducted using TaqMan Assays.

Highly specific and sensitive, TaqMan PCR provides optimized assays for genotyping

SNPs [285] . TaqMan SNP Genotyping Assays are ideal for amplifying and detecting specific SNP alleles in purified genomic DNA samples, allowing researchers to genotype study participants for a specific SNP.

The recently completed HapMap project, as well as the development of highly multiplexed technologies, have resulted in vast improvements in technologies used for genotyping. During the summer of 2007, we selected genes for a GoldenGate©

(Illumina©) Oligo Pool All (OPA) assay ( http://www.illumina.com ) [286] . This assay

was developed by using tag SNPs in the SNP500Cancer project

(http://snp500cancer.nci.nih.gov ) [284] , with previous re-sequence analysis and plausible

evidence that the genes were related to carcinogenic processes [284] . Genes were

selected to comprehensively evaluate common variation in different pathway genes in

kidney cancer, and also to extend interesting findings from our SNP based analyses

described above.

39 The majority of SNPs were selected using a tag SNP method [287] with at least 80% genomic coverage of the genomic region of interest, while some non-synonymous SNPs were selected for their putative functional significance. Genes with a variant allele frequency of at least 5% in the Central European population as reported by NCI’s

SNP500Cancer Database [284] , and a validated assay at the NCI’s CGF

(http://cgf.nci.nih.gov/home.cfm ) [288] were selected for this study.

Smoking status was categorized into never smoker, former smoker, or current smoker.

Specifically, participants still smoking anytime during the 24 months prior to the interview were classified as current smokers; participants who had ceased smoking for greater than two years prior to interview were classified as former smokers. Through the use of lifestyle questionnaires, hypertension status was defined by ever use of

antihypertensive medication. BMI was calculated as weight in kilograms divided by the

square of height in meters (kg/m 2). Education was categorized into primary (elementary unfinished and finished), secondary and apprenticeships, and higher education (high school, university, or higher). Family history of cancer was categorized as having a first-

degree relative with kidney cancer, any other cancer, or no cancer.

2.3 Previous Results Published from the CEERCC Study:

Previous reports from the CEERCC Study have evaluated the associations between

certain lifestyle factors and kidney cancer risk. A description of study participants is

provided in Table 5. Cases and controls were comparable in age. However, cases were

more often female and more likely to have excess BMI (>30 kg/m 2), hypertension, and a

40 first-degree relative with cancer [289] . Compared to those in the lowest BMI category

(<25 kg/m 2), risks for those with BMI 30-35 kg/m 2 and 35+ kg/m 2 had higher RCC risks,

1.39 (95% Confidence Interval (CI)= 1.09-1.75) and 1.33 (95% CI= 0.91-1.96), respectively. Increased risk was observed for subjects with self-reported hypertension compared to those without (Odds Ratio (OR)= 1.22; 95% CI= 1.03-1.46). Table 5 also shows that cases were less likely to smoke, however, after adjusting for age, BMI, hypertension, center and sex, this association disappeared [12, 52] . A food frequency questionnaire that included 23 food items was used to determine dietary intake of specific food group items among cases and controls. Increased risks were associated with consumption of red meat, milk, and yogurt, and reduced risks were associated with consumption of vegetables, most notably cruciferous vegetables [19] . Sensitivity analyses were conducted and determined that associations of RCC with dietary consumption patterns were similar before and after political and market changes in 1989

(1991 in Russia) in adjusted multivariate logistic regression models.

Table 5: General characteristics of participants in the CEERCC study [12, 19,

170]

Variables Cases Controls N % N % p-value

Participants 1,097 42.6 1,476 57.4

Sex Males 648 59.1 952 64.5 Females 449 40.9 524 35.5 0.01

Age at Interview <45 86 7.8 122 8.3 45-54 278 25.4 379 25.6

41 55-64 335 30.5 460 31.2 65-74 353 32.2 452 30.6 75+ 45 4.1 63 4.3 0.61

Mean Age (std) 59.6 years (10.3) 59.3 years (10.3)

Center Romania-Burcharest 95 8.7 160 10.8 Poland-Lodz 99 8.7 198 13.4 Russia-Moscow 317 28.9 463 31.4 *Czech Republic 586 53.4 655 44.4 <0.001

BMI at Interview <25 327 29.8 532 36.2 25-29.9 476 43.4 620 42.1 30+ 293 26.8 319 21.7 <0.001

Tobacco Status Never 510 46.6 599 40.6 Ever 584 53.4 876 59.4 0.003

Hypertension No 600 54.7 906 61.4 Yes 496 45.3 569 38.6 0.001

Family History of Cancer No 1 st degree relative with cancer 733 66.8 1074 72.8 1st degree relative with cancer 364 33.2 402 27.2 0.001

*Brno, Olomouc, Prague, Ceske-Budejovice

2.4 Research Design and Methods:

As aforementioned, the research hypothesis and study aims of the present dissertation study were investigated by analyzing data and genomic DNA from the CEERCC Study.

The potential anti-carcinogenic effect of vitamin D on RCC risk was examined as an add- on to the results already reported in the CEERCC Study. Results of the present dissertation are described in three separate manuscripts:

42 2.5 Manuscript One & Specific Aim One- Occupational Sunlight Exposure:

Occupational information from the CEERCC Study was used to create a JEM for occupational sunlight exposure. Based on ISCO coded job and NACE coded industry titles assessed by occupational health experts and industrial hygienists, I created a JEM that categorized confidence, frequency, duration, and intensity of occupational sunlight exposure for all occupations (N= 10,617) held for at least 12 months duration for each study participant. My assignment of confidence, frequency, duration, and intensity of occupational sunlight exposure categories was based on previous guidelines utilized by

Dosemeci and colleagues [131], who evaluated job titles among Europeans to assess occupational sunlight exposure and non-Hodgkin Lymphoma (NHL) risk through the use of a JEM [131]. New exposure variables were created from this JEM. These included:

• frequency of exposure, which was coded in three categories and represented the

percentage of time in a 8 hour work day during which exposure was possible:

<30% (0 to <2.5 hours), 30% -69.9% (2.5 to <5.6 hours), and >70% (5.6 hours or

more);

• confidence of occupational sunlight exposure was also coded in three categories

as “possible” (<40%), “probable” (40% -90%), or “certain” (>90%) and

represented the degree of confidence in our assignment for frequency of exposure;

and

• intensity of exposure, which was coded in two categories, “high” for participants

suspected to be exposed to strong UV light reflected from the sea, and for

agricultural workers and outdoor occupations in a rural setting, or “low” for

participants suspected to be exposed to weak UV light. Intensity of exposure (on

43 average) was assumed to be twice that for jobs rated “high” compared to jobs

rated “low”.

My assessment of confidence, frequency, and intensity of occupational sunlight exposure for each job held by each participant was reassessed by two occupational health experts at

NCI. For the assigned categories in which there was a disagreement, a third health expert was consulted and a final decision was made regarding the assigned categories.

Accordingly, confidence of occupational sunlight exposure was then recoded to a lower value. Inter-rater agreement for the frequency, confidence, and intensity of occupational

UV exposure was also calculated using a kappa statistic.

Four measures of occupational sunlight exposure were assessed: cumulative exposure (in low exposure-unit-years) across all jobs, frequency-adjusted duration of exposure in years across all jobs, frequency-adjusted duration of exposure in years among participants who held only low intensity jobs, and frequency-adjusted duration of exposure in years among participants who held any high intensity jobs.

• Cumulative exposure was calculated as duration (years) x frequency midpoint x

intensity of exposure for each job and summed over jobs.

• Frequency-adjusted duration of exposure in years was calculated as duration

(years) x frequency midpoint for each job and summed over jobs.

• Frequency-adjusted duration of exposure in years among participants who held

only low intensity jobs excluded participants who held only high intensity or both

high and low intensity jobs; and

44 • Frequency-adjusted duration of exposure in years among participants who held

any high intensity jobs excluded participants who held only low intensity jobs.

Categorical exposure metrics were used to evaluate exposure-response relationships with occupational exposure based on tertiles of exposure levels among all controls, all male controls, and all female controls. However due to small numbers, frequency-adjusted duration of exposure among participants who held any high intensity jobs was evaluated dichotomously by comparing the first two tertiles of duration to the third tertile.

Subgroup analyses were examined by sex, median age, and restricted to jobs assigned a high confidence (probable or certain) of occupational sunlight exposure. Subgroup analyses were also performed by latitude/study center (Russia (55.8° N, 37.6°E), Poland

(51.6°N, 19.5°E), Czech Republic (49.2°-50.1°N, 14.4°-17.3°E), and Romania (44.4° N,

26.1°E)) as another estimate of sunlight intensity. Exploratory analyses of associations between UV exposure and other relevant variables, such as by BMI, self-reported hypertension, and smoking status, were also evaluated. Logistic regression and linear trend were calculated using continuous variables.

In addition to being the first epidemiological case-control study to examine the association between occupational sunlight exposure and RCC risk, this study is unique in that: (1) detailed information on occupational exposure histories was collected and assessed by trained industrial hygienists, (2) high participation rates were observed among both cases and controls, and (3) sufficient statistical power for analyses was

45 obtainable due to the large sample size of the study.

2.6 Manuscript Two & Specific Aim Three- Vitamin D Pathway Genes:

During the summer of 2007, I selected genes for a GoldenGate© (Illumina©) OPA assay

to comprehensively evaluate common variation in vitamin D pathway genes in kidney

cancer. Through the use of existing DNA extracted from whole blood and buffy coats,

987 (90.0%) RCC cases and 1,298 (87.9%) provided blood samples and suitable genomic

DNA that was available for genotyping at the NCI’s CGF. Genotyping was conducted

with a GoldenGate ® Illumina assay ( http://www.illumina.com ), which was developed

by using tag SNPs [286] . One-hundred-thirty-nine SNPs across eight target genes

(CYP24A1, GC, RXRA, RXRB, STAT1, THRAP4, TRAP5, and VDR) were selected for their involvement in renal maintenance and vitamin D metabolism, binding, transport, expression, and/or function. These genes may therefore affect the mechanisms through which we hypothesized that vitamin D influenced cancer risk.

Preliminary analyses of these genes showed that about two-thirds (63%) of the SNPs were located in introns, 11% in exons, 12% in promoter regions, and 15% in 3 ′ of stop codon. For SNPs located in exons, 7% were intergenic, 7% were in 5 ′ untranslated regions (UTRs), 13% were in 3 ′ UTRs, 20% were downstream, and 53% were in coding

regions (with the majority of changes being synonymous versus non-synonymous). The

median (range) minor allele frequency (MAF) among controls was 0.25 (0.03–0.52). The

number of SNPs per gene also varied, ranging from 1 to 35 SNPs.

46 Main effects for these SNPs in eight target genes in the vitamin D pathway were analyzed

among genotyped participants. Subgroup analyses by age, sex, smoking status, BMI,

hypertension, and family history of cancer were also evaluated. Haplotype and multiple

comparison test analyses were conducted to elucidate whether genetic variations in these

genes in the vitamin D pathway modified RCC risk.

Several methods were used to test the association between genes and RCC risk. First,

Hardy-Weinberg equilibrium was tested by the goodness of fit χ2 test. Pairwise linkage

disequilibrium (LD) between SNPs was estimated based on D ′ and r 2 values using

Haploview ( http://www.broad.mit.edu/mpg/haploview/index.php ) [290] . A dominant model was used therefore, polymorphisms of SNPs with a frequency of <5% among

controls was combined. Unconditional logistic regression was used to calculate the main

effects of each SNP. Analyses for only one SNP were evaluated among SNPs with a high

correlation (r ΄ >0.85) in order to avoid redundant analyses. Next, global p-values were

used to evaluate several tests including the minimum-p value permutation (Min-P) test,

Simes global test and Bonferroni correction, which are described in more detail below.

Haplotype analyses for genes with more than one SNP were evaluated in blocks using R

(version 2.4.0; http://www.r-project.org ) [291] , adjusted for age, gender, center and

smoking status. An additional haplotype-based method (sliding window analysis of three

consecutive SNPs) was also used to identify chromosomal regions of interest that

remained significant at a false discovery rate (FDR) level of 10%. Subsequently,

unadjusted and adjusted risk estimates were generated for regions with a high level of

signal.

47

This study is unique in that it is the largest RCC study with biological samples to date, providing sufficient statistical power to detect relatively small associations between RCC risk and genetic variations across polymorphisms in vitamin D pathway genes.

Furthermore, this is the first study to comprehensively evaluate vitamin D pathway genes in relation to renal cancer risk.

2.7 Manuscript Three & Specific Aims Two and Four- Effect of Vitamin D Pathway

Genes in Relation to Vitamin D Intake and Exposure and Renal Cell Carcinoma Risk:

Vitamin D pathway genes were assessed to determine whether SNPs variants across these genes modified associations between dietary intake of vitamin D and calcium rich foods and occupational UV exposure and RCC risk. As aforementioned, 987 (90.0%) RCC cases and 1,298 (87.9%) controls provided blood samples and suitable genomic DNA that was available for genotyping at the NCI’s CGF. A GoldenGate ® Illumina assay was used to genotype SNPs across vitamin D pathway genes [286] .

First, vitamin D pathway genes were assessed in relation to RCC risk and dietary intake of vitamin D and calcium rich foods. Dietary vitamin D and calcium intake were based on information obtained from the food frequency questionnaires from the CEERCC

Study. Consumption of food-specific items was categorized into: never, low (<1 a month), medium ( ≥1 a month but <1 a week), and high (>1 a week). Using these data, intake frequencies of liver, egg, and fish (salt-water and fresh-water) consumption were used to create a new dietary exposure variable for total vitamin D intake. Since calcium

48 intake has been reported to influence vitamin D levels, intake frequencies of cheese, milk and yogurt were used to create a new total calcium intake exposure variable. Also, the assessment of vitamin D pathway genes in relation to RCC risk and occupational UV exposure was evaluated using the occupational sunlight job-exposure matrix data created for manuscript one.

Hardy-Weinberg equilibrium was tested by the goodness of fit χ2 test. Pairwise LD between SNPs was estimated based on D ′ and r 2 values using Haploview [290] . Analyses for only one SNP were evaluated among SNPs with a high correlation (r ΄ >0.85) in order to avoid redundant analyses. A dominant model was used for analysis. Unconditional logistic regression was used to calculate the main effects of each SNP. All regression models were adjusted for age (continuous), sex, study center, smoking status (ever, never), BMI, and self-reported hypertensive status (no, yes). Potential interactions were tested using a multiplicative model, where interactions between SNPs and dietary or occupational exposures were tested comparing regression models with and without interaction terms using a likelihood ratio test (LRT).

This is the first study to comprehensively evaluate the association between renal cell carcinoma risk, vitamin D pathway genes, vitamin D intake, calcium intake and occupational UV exposure.

2.8 General Statistical Methods:

Unconditional logistic regression was used to calculate the odds ratio (OR) and 95 percent confidence interval (95% CI) of kidney cancer associated with each exposure of

49 interest. Different models were used in examining the risk of kidney cancer. The

inclusion of potential adjustment variables for each model was determined by each

variable’s ability to affect OR values by 10% or more.

• Occupational sunlight exposure assessment models were considered for

adjustment of: age (continuous), study center, sex, smoking status (never, ever),

self-reported hypertension (no, yes), BMI (categorical: >25, 25-29.9, >30kg/m 2),

and/or dietary intake of vitamin D rich foods.

• Dietary vitamin D intake assessment models were considered for adjustment of:

age (continuous), sex, study center, smoking status (ever, never), self-reported

hypertension (no, yes), BMI (categorical: >25, 25-29.9, >30 kg/m 2), and/or

dietary intake of calcium rich foods.

• Genotyping assessment models were considered for adjustment of: age

(continuous), sex, study center, smoking status (never, ever), self-reported

hypertension (no, yes), BMI (categorical: >25, 25-29.9, >30 kg/m 2), dietary

vitamin D intake and/or occupational sunlight exposure variables.

To check for lack of fit for each of the regression models, Hosmer-Lemeshow goodness of fit test statistics were calculated. All analyses were conducted in STATA 9.0 unless otherwise specified (STATA Corporation, College Station, TX).

2.9 Multiple Comparison Issues:

To address the inherent multiple statistical comparisons problem, controlling for type I error rates, the Simes Global Test, and the Min-P test were evaluated. The Simes Global test evaluates each SNP variant or haplotype while providing a more powerful alternative

50 to the Bonferroni test [292] . This test uses a non-iterative procedure to adjust the

minimum observed p-value for multiplicity [292, 293] . While the Min-P test corrects for

multiple testing, this test also accounts for correlations between SNPs within a gene

[292] . However, inference of the Min-P test is based on the permutation distribution of

the minimum of the ordered p-values [292] . Lastly, an FDR test was used to identify

associations unlikely to be due to chance. FDR is the expected proportion of type I errors

among all significant results [294] .

2.10 Power:

The prevalence of potential risk factors and the sample size of this study are already known. The large sample size for this study provided sufficient statistical power (>80%)

to detect relatively small associations between RCC risk and potential risk factors such as

by genotype, occupational sunlight exposure, and dietary intake of vitamin D rich foods.

Based on calculations from the NCI Power Program V.3.0.0 [295] , with greater than 900

cases and 900 controls genotyped, adequate statistical power was available to detect a 1.5

fold increase in renal cancer risk, with an alpha of 5%. However, the power to detect a

1.4 fold or smaller increase in risk was limited since the prevalence of the “at risk allele”

needed to be greater than 18% (Figure 5). For example, the prevalence of the “at risk”

allele for the TaqI VDR polymorphism for Caucasians from the SNP500Cancer project database is 25.8% [284] ; thus, as illustrated in Figure 5 there is adequate statistical power to detect an association as small as 1.4, however the power to detect a 1.3 fold increase in risk is limited.

51 Figure 5: Power calculations to detect an OR of 1.3-1.5 fold increase in RCC

risk with 900 cases and 900 controls

Occupational sunlight exposure and dietary vitamin D data were collected from approximately 1,100 cases and 1,100 controls. According to Figure 6, using an alpha of

5%, relatively small (1.3, 1.4 and 1.5 fold increase) associations between occupational sunlight exposure and/or dietary vitamin D intake and RCC risk could be detected with a power of 80% or greater; yet, the power to detect a 1.3 fold increase in risk was limited for exposures whose prevalence fell below 26%.

52 Figure 6: Power calculations to detect an OR of 1.3-1.5 fold increase in RCC

risk with 1,100 cases and 1,100 controls

The power to detect interactions for this study was limited (Table 6). Using an additive

or multiplicative model, the prevalence of the “at risk” allele must have been greater than

10% in order to detect a two-fold interaction; additionally, a semi-high exposure

prevalence (>20% for additive model and >15% for multiplicative model) was required

to achieve 80% power. For example, as described above, if the prevalence of the “at

risk” allele for the TaqI VDR polymorphism is 25.8% among Caucasians, then the prevalence for dietary vitamin D intake must be greater than 30% or 15% to have sufficient power to detect an interaction when using an additive or multiplicative model, respectively. Thus, the power to detect interaction in this study is limited and was considered during interpretation of results.

53 Table 6: Power to detect a two-fold interaction with 900 cases and 900 controls

“At Risk” Exposure Additive Multiplicative Allele Prevalence Prevalence Model Model 50% 70% 98% 77% 50% >99% 91% 20% 90% 89% 10% 67% 73% 20% 70% 85% 77% 50% 90% 89% 20% 71% 84% 10% 45% 64% 10% 70% 59% 59% 50% 67% 73% 20% 45% 63% 10% 26% 42% 5% 70% 41% 43% 50% 46% 55% 20% 29% 45% 10% 17% 27%

54 Chapter 3: Occupational Sunlight Exposure and Risk of Renal Cell Carcinoma

Karami S 1, Boffetta P 2, Stewart P 3, Rothman N 1, Hunting KL 12 , Zaridze D 4, Navritalova

M7, Mates D 9, Dosemeci M 1, Janout V 5, Kollarova H 5, Bencko V 6, Szeszenia-

Dabrowska N 8, Holcatova I 6, Mukeria A 4, Gromiec J 11 , Chanock S 10 , Berndt SI 1, Brennan

P2, Chow W-H1, Moore LE 1

1Division of Cancer Epidemiology and Genetics, National Cancer Institute, NIH, DHHS,

Bethesda, MD, USA

2International Agency for Research on Cancer, Lyon, France

3Stewart Exposure Assessments, LLC, Arlington, VA, USA, Formerly of the Division of

Cancer Epidemiology and Genetics, National Cancer Institute, Bethesda, MD, USA

4Institute of Carcinogenesis, Cancer Research Centre, Moscow, Russia

5Department of Preventive Medicine, Faculty of Medicine, Palacky University, Olomouc,

Czech Republic

6Institute of Hygiene and Epidemiology, Charles University, First Faculty of Medicine,

Prague, Czech Republic

7Department of Cancer Epidemiology and Genetics, Masaryk Memorial Cancer Institute,

Brno, Czech Republic

8Department of Epidemiology, Institute of Occupational Medicine, Lodz, Poland

9Institue of Public Health, Bucharest, Romania

10 Core Genotyping Facility at the Advanced Technology Center of the National Cancer

Institute, NIH, Department of Health and Human Services

11 Nofer Institute of Occupational Medicine, Department of Chemical Hazards, Lodz,

55 Poland

12 Department of Epidemiology & Biostatistics, George Washington University,

Washington, DC, USA.

Running Title: Sunlight and kidney cancer risk

Key words: RCC, kidney cancer, vitamin D, UV, UVB, sunlight

56 Abstract:

Recent findings indicate that vitamin D obtained from ultraviolet (UV) exposure may

reduce the risk of a number of different cancers. Anti-carcinogenic properties of vitamin

D include inhibition of clonal tumor cell proliferation, induction of immune cell

differentiation and apoptosis, and decreased angiogenesis. Vitamin D is metabolized to

its active form within the kidney, the major organ for vitamin D metabolism and activity,

and calcium homeostasis. Since both the incidence of renal cell cancer and prevalence of vitamin D deficiency have increased over the past few decades, the present study sought to explore whether occupational UV exposure was associated with renal cell carcinoma

(RCC) risk. A hospital-based case-control study of 1,097 RCC cases and 1,476 controls was conducted in four Central and Eastern European countries. Demographic and occupational information was collected to examine the association between occupational

UV exposure and RCC risk. A significant (38%) reduction in RCC risk was observed with occupational UV exposure among male participants. No association between UV exposure and RCC risk was observed among female participants. When analyses were stratified by latitude, a stronger (73%) reduction in RCC risk was observed between UV exposure and RCC risk among males residing at the highest latitudes. To our knowledge, this is the first occupational case-control study to specifically investigate the association between sunlight exposure and RCC risk. The results of this study suggest that among males there is an inverse association between occupational UV exposure and renal cancer risk.

57 Introduction:

The incidence of renal cell carcinoma (RCC), the main form of kidney cancer, has

increased both in the United States (U.S.) and globally over the past two decades. [1][2]

Reasons for this increase remain speculative; however, reduced vitamin D could be a

contributing factor. Recent large scale epidemiological studies, many of them ecological,

have shown linear inverse associations between solar ultraviolet (UV) B exposure and

incidence and/or mortality rates for breast, colorectal, ovarian, and prostate cancer and

Hodgkin and non-Hodgkin lymphoma. [3][4] Anti-carcinogenic properties of vitamin D

include inhibition of clonal tumor cell proliferation, induction of immune cell

differentiation and apoptosis, and decreased angiogenesis. [5][6] Yet, throughout the U.S.

and Europe, widespread vitamin D deficiency/insufficiency continue to be a problem for

all age groups because of inadequate vitamin D intake and decreased sun exposure,

possibly due to a sedentary lifestyle. [4]

Vitamin D is a fat soluble vitamin found in food and synthesized in the skin upon

exposure to solar UVB rays. [7] Fortified foods, such as milk, butter, and cereal are

common sources of vitamin D in the U.S. and Canada. However, in most countries, very

few foods naturally contain significant amounts of vitamin D; the majority of individuals

obtain most of their vitamin D though sunlight exposure. [7] In general, UVB exposure

accounts for approximately 90% of 1,25-dihydroxyvitamin D (1,25(OH) 2D3) levels, the biologically active form of vitamin D. [4][8] After vitamin D is produced in the skin or consumed in food, the vitamin is hydroxylated in the liver and subsequently the kidney to form 1,25(OH) 2D3. [7-11]

58

Although both the incidence of renal cell cancer and prevalence of vitamin D deficiency have increased over the past two decades, the relationship between UVB exposure and kidney cancer has not been explored. [1][9][10] Given the widespread public health interest in vitamin D and its potential cancer prevention promise, it is important to investigate the relationship between vitamin D and cancer risk, particularly renal cell carcinoma since the kidney is the major organ for vitamin D metabolism and activity, and calcium homeostasis. [9-11] In the present study, we investigated the association between estimated exposure to occupational sunlight and RCC risk in one of the largest, multi- centered renal case-control studies conducted to date in Central and Eastern Europe, an area with one of the highest rates of RCC in the world. [2]

Material and Methods:

Study Population

From 1999 through 2003, a hospital-based case-control study of RCC was conducted in seven centers in four countries of Central and Eastern Europe (Moscow, Russia;

Bucharest, Romania; Lodz, Poland; and Prague, Olomouc, Ceske-Budejovice, and Brno,

Czech Republic). Cases, aged 20-88 years, included patients with newly diagnosed histologically confirmed RCC (IDC-O-2 codes C64) who had lived in the study areas for at least one year and were interviewed within three months of diagnosis. RCC tumors were histologically confirmed at the National Cancer Institute (NCI) by a world expert in renal tumor pathology. Frequency-matched to cases on age (+/- 3 years), sex, and place of residence, controls were selected from patients admitted to participating hospitals for

59 diagnoses unrelated to smoking or urological disorders with the exception of benign

prostatic hyperplasia. No single disease made up more than 20% of the control group.

Some controls were also recruited in parallel for studies of lung and head and neck

cancers. [12][13] All participants were of Caucasian descent. The final number of

participants for analysis in this study included 1,097 renal cancer cases and 1,476

controls. The response rates across study centers for study participation ranged from

90.0% to 98.6% for cases and from 90.3% to 96.1 % for controls. All subjects provided

written informed consent. This study was approved by the institutional review boards of all participating centers.

Occupational Exposure Assessment

Interviewers were trained in each center to perform face-to-face interviews of cases and controls during hospitalization using standard questionnaires. The questionnaire covered basic demographic characteristics, family history of cancer, history of tobacco consumption, and dietary habits. Lifetime occupational information for jobs of >12

months duration was also ascertained through the use of a general occupational

questionnaire. Data collected for each job included title, detailed tasks, and type of

employer, as well as year of beginning and ending employment.

Job titles were coded by local industrial hygienists or occupational health experts, blinded

to case-control status, according to the International Standard Classification of

Occupation 1968 version (ISCO-68). [14][15] Industries were similarly coded according

to the Statistical Classification of Economic Activities of the European Community, 1999

60 version (NACE-99). [15][16] Both the ISCO and NACE coding systems are regularly utilized in occupational epidemiological studies conducted throughout Europe for occupation and industry classifications. [17]

Based on the ISCO coded job and NACE coded industry titles, a job exposure matrix

(JEM) was created to categorize frequency, duration, confidence, and intensity of occupational sunlight exposure for each study participant. Frequency of exposure was estimated by the percentage of time in an 8 hour day during which exposure was possible: <30% (0 to <2.5 hours), 30-69.9% (2.5 to <5.6 hours), and >70% (>5.6 hours or more) of the time. To compute sunlight exposure across jobs that had different frequencies of exposure, frequency weights of (0.15, 0.50, and 0.85) were assigned to the three categories, corresponding to the midpoint of the range of each category. Level of confidence of occupational sunlight exposure for each job was coded in three categories as “possible” (<40%), “probable” (40-90%), or “certain” (>90%), based on the likely tasks and location (indoor, outdoor) of the job, representing the degree of confidence in our assignment for frequency of exposure. Intensity of exposure was coded as “high” for participants suspected to be exposed to strong ultraviolet light reflected from the sea and for agricultural workers and outdoor occupations in a rural setting or “low” for participants suspected to be exposed to weak ultraviolet light (all other jobs). Intensity of sunlight exposure was assumed to be twice that for jobs rated “high” (two units) compared to jobs rated “low” (one unit).

To assure high quality exposure assessment, the assignments of probability, frequency,

61 and intensity of occupational sunlight exposure categories were reassessed by two

industrial hygiene experts (MD, PS) at the NCI for each job held by each participant.

Inter-rater agreement for the probability, frequency, and intensity of occupational

sunlight exposure was calculated using Cohen’s kappa statistic. [18]

Statistical Analysis

Several measures of occupational sunlight exposure were assessed:

 cumulative exposure (low exposure-unit-years) across all jobs, calculated as

duration (years) x frequency midpoint x intensity of exposure (units) for each job

and summed over jobs;

 frequency-adjusted duration of exposure in years across all jobs, calculated as

duration (years) x frequency midpoint for each job and summed over jobs;

 frequency-adjusted duration of exposure in years among participants who held

only low intensity jobs, (excluding participants who held only high intensity or

both high and low intensity jobs); and

 frequency-adjusted duration of exposure in years among participants who held

any high intensity jobs (excluding participants who held only low intensity jobs).

Categorical exposure metrics were used to evaluate exposure-response relationships with occupational exposure based on tertiles of exposure levels among all controls, all male controls, and all female controls. However due to small numbers, frequency-adjusted duration of exposure among participants who held any high intensity jobs was evaluated dichotomously by comparing the first two (lowest) tertiles of duration to the third

(highest) tertile. Subgroup analyses were examined by sex, median age, and restricted to

62 jobs assigned a high confidence (probable or certain) in the occupational sunlight exposure assessment. Since the exposure metrics for nearly all (>98%) occupations were assigned a high confidence, analyses are presented only for this high confidence subset.

Subgroup analyses were also performed by latitude/study center (Russia (55.8° N,

37.6°E), Poland (51.6°N, 19.5°E), Czech Republic (49.2°-50.1°N, 14.4°-17.3°E), and

Romania (44.4° N, 26.1°E)) as another estimate of sunlight intensity. Exploratory analyses of associations between ultraviolet exposure and other relevant variables, such as body mass index (BMI), self-reported hypertension, and smoking status, were also evaluated for potential confounding.

Odds ratios (OR) and 95% confidence intervals (95% CI) were calculated to estimate

RCC risk and association with occupational sunlight exposure, using unconditional logistic regression models adjusting for sex, age, center, smoking status (never, ever),

BMI, and self-reported hypertensive status (no, yes). To check for lack of fit for each model, Hosmer-Lemeshow goodness of fit test statistics were calculated. Logistic regression was performed and tests for linear trends using continuous variables were calculated to estimate RCC risk by occupational exposure. All analyses were conducted in STATA 9.0 unless otherwise specified (STATA Corporation, College Station, TX).

Results:

Inter-rater agreement for assessment of intensity of occupational sunlight exposure, using

Cohen’s kappa statistic, was very good (kappa= 0.82). The agreement for frequency of occupational sunlight exposure was good (kappa= 0.73). Confidence in the exposure

63 assessment, coded in three categories as “possible,” “probable,” or “certain” had a fair inter-rater agreement (kappa= 0.48); however, when probability of exposure was assessed as low (<40%) versus high (> 40%), inter-rater agreement was better (kappa= 0.65).

A description of study participants and known RCC risk factors is provided in Table 1.

Cases and controls were comparable in age and education level, but cases were more likely to be female and were more likely to have an excess BMI (>30 kg/m 2), hypertension, and a first-degree relative with cancer. The association with smoking was not observed after adjustment for age, BMI, hypertension, study center, and sex. [19]

No significant patterns in association for occupational sunlight exposure were seen for all participants. Significant linear inverse associations with RCC were, however, observed among males occupationally exposed to sunlight (Table 2). Decreased RCC risk and significant exposure-response relationships were observed among male participants in the highest tertile for cumulative exposure (OR= 0.76; 95% CI= 0.58-1.00; p-trend= 0.05), frequency-adjusted duration of exposure (OR= 0.76; 95% CI= 0.58-0.99; p-trend= 0.04), and frequency-adjusted duration of exposure among subjects who only held low intensity jobs (OR= 0.62; 95% CI= 0.45-0.85; p-trend= 0.003). A non-significant decrease in

RCC risk was observed among male participants who held any high intensity jobs (p- trend= 0.60). No significant associations were observed between RCC risk and occupational sunlight exposure for female subjects. The Hosmer-Lemeshow test supported the fit of the model for all exposure-response matrices examining the association between RCC risk and occupational sunlight exposure among all, male, and

64 female participants.

As another estimate of sunlight intensity, analyses were stratified by latitude/study center.

No statistically significant odds ratios were found when the data were stratified by study center, except among Russian males who showed strong inverse associations (Table 3).

Specifically, significant exposure-response relationships were seen for cumulative exposure (p-trend <0.001), frequency-adjusted duration of exposure (p-trend <0.001), and duration of exposure among subjects who only held low intensity jobs (p-trend <0.001).

These associations were more pronounced than seen for all males combined. No significant association was observed for any high intensity jobs and RCC risk. No significant association was observed among females or among males from the centers located at lower latitudes. Logistic regression analyses for all exposure-response matrices examining the association between RCC risk and occupational sunlight exposure among Russian male and Russian female participants was also supported by the Hosmer-

Lemeshow goodness of fit test.

Discussion :

To our knowledge, this is the first case-control study to show evidence of an inverse association between occupational UV exposure and RCC risk. The inverse association between occupational UV exposure and RCC risk was observed only among males.

Reduced RCC risk among males was significantly and linearly observed for increasing cumulative exposure (p-trend= 0.05), frequency-adjusted duration of exposure (p-trend=

0.04), and frequency-adjusted duration of exposure for only low intensity jobs (p-trend=

65 0.003). No significant association between occupational UV exposure and RCC risk was observed among female participants; hypertension, smoking, and excess BMI also showed no association between UV exposure and RCC risk among females. When analyses were stratified by latitude of study center as another estimate of UV intensity, a stronger inverse association between UV exposure and RCC risk was observed among males from Moscow, the study center expected to have the lowest intensity UV exposure compared to the other centers located at lower latitudes.

The results of our study are supported by other epidemiological kidney cancer studies. In general, ecological studies investigating the association between cancer risk and sun exposure have reported an inverse relationship between kidney cancer incidence and mortality risk and UVB exposure. [20-24] Recently, Mohr and colleagues examined the association between UVB exposure and renal cancer risk in 175 countries using latitude and solar UVB irradiance. [24] The highest renal cancer incidence rates were found in countries situated at the highest latitudes (p-value <0.01 in both men and women). [24]

Furthermore, an occupational cohort study exploring the relationship between UV exposure and cancer risk among Swedish male construction workers found that participants exposed to the highest level of occupational sunlight had reduced kidney cancer risk (OR= 0.7; 95% CI= 0.4-1.00). [25] Similar inverse associations have been reported between occupational sunlight exposure and other cancers, such as diffuse large

B-cell lymphoma (OR= 0.72; 95% CI= 0.54-0.97), follicular lymphoma (p-trend= 0.04), and rectal cancer (OR= 0.62; 95% CI= 0.42-0.93). [26][27]

66 One explanation for the inverse association between UV and cancer is the hypothesis that sunlight exposure increases production of vitamin D in the skin that travels through the blood to other sites in the body. This is particularly important with regards to kidney cancer, because the kidney is the major organ responsible for vitamin D metabolism and activity, and calcium homeostasis. [9-11] Vitamin D and its analogues are thought to suppress tumor activity by inhibiting clonal tumor cell proliferation and the G1 cell cycle arrest phase, inducing immune cell differentiation and apoptosis, and decreasing angiogenesis. [5][6][28][29]

In this study, we observed significant inverse trends only among males with low intensity occupations and not among males with high intensity UV exposure. This finding was corroborated when we stratified our analysis by study center latitude as an estimate of

UV intensity. Significant trends were observed among Russian males expected to have the lowest intensity UV exposure. The lack of an apparent effect among subjects with high UV intensity exposure can be explained considering that high intensity UV exposure can result in sufficient amounts of vitamin D in a short amount of time, whereas longer exposure time is necessary if UV intensity is lower. After saturation, an equilibrium is reached and furthermore, UV exposure no longer results in the formation of additional vitamin D and actually starts to degrade the vitamin. [30][31] For example, most healthy

Caucasians can generate sufficient amounts of vitamin D with as little as 10 to 20 minutes of high intensity sun exposure to unprotected skin. [30] One possibility is that, in lower latitudes with higher intensity sunlight exposures, there may not be sufficient contrast in vitamin levels across groups with varying frequency of exposures because

67 most people are already exposed beyond this level of saturation through everyday, non- occupational activities. In other words, occupational sunlight exposure may be relatively more important in higher latitudes with lower intensity UVB because residents there are less likely to reach the threshold of adequate sunlight exposure outside of work. [30]

Alternatively, the lack of association among high intensity exposure occupations, (such as farmers and gardeners) may be influenced by other carcinogenic co-exposures in this group. We have already reported increased kidney cancer risk among pesticide exposed individuals that was highest among glutathione S-transferase mu 1/theta 1

(GSTM1 /GST1 ) active subjects in this study. [31]

The association between occupational UV exposure and RCC risk was observed among males and not females in this study. Gender differences related to UV sensitivity and cancer risk have been demonstrated in other studies but it is not known whether they are real and due to due to hormonal differences between sexes or caused by differential misclassification due to behavior differences. [32][33] Males and females may differ biologically in their response to UV exposure. Laboratory studies suggest that there are gender related hormonal differences that may play a role in responses to acute UVB exposure as well as UV-induced tumor development. [34][35] Several studies have observed gender differences related to UV exposure and cancer risk; however, the results have not been consistent. A recent European case-control study that found an inverse association between UV radiation and lymphoma risk and a significant interaction between sex and skin sensitivity (tendency to sunburn) was observed with Non-Hodgkin

Lymphoma (NHL) risk. Specifically, female participants experienced higher risks of

68 NHL with increasing skin sensitivity (p-trend <0.001; p-interaction= 0.02) compared with male participants. [26] Similarly, elevated NHL risk was reported among females but not among males with outdoor occupations in a large English cohort study that included 401 NHL male and 27 NHL female cases. [36] Gender differences were also observed in an Australian occupational case-control study that found an inverse association between occupational UV exposure and risk of glioma in females (OR= 0.54;

95% CI= 0.27-1.07) while there was a positive association among males (OR= 1.60; 95%

CI= 0.95-2.69). [37] Alternatively, gender differences may also be due to misclassification of occupational exposure due to behavioral differences, such as a higher tendency for females to use sunscreen on a regular basis and the ability of males to work outdoors while shirtless. [38] Occupational exposure may be more precise for males than for females. Historically, males have had occupations that require them to spend more time outdoors in the sun compared with females. [34][39][40] It is noteworthy that in this study, occupational exposure levels estimated among females were considerably less than among males and we did not ask questions specifically related to recreational exposures among either sex. Lastly, we cannot completely rule out the possibility that these differences are due to chance.

Strengths of our study include a large sample size, a high participation rate, inclusion of only histologically confirmed cancers, and the use of both the ISCO and NACE coding systems to assign individual-specific exposure information. Furthermore, inter-rater agreement scores confirmed consistent exposure assessment for frequency, intensity, and confidence of occupational sunlight exposure. There are several limitations that are

69 inherent to our study. First, we did not obtain data regarding recreational exposure to UV light, history of sunburns, and sunbathing activities, which have been associated with other cancers. Failure to measure all sources of sunlight exposure may have led to misclassification of sunlight exposure and biased our results towards the null. Secondly, while participants were primarily of Central European descent, information on hair color, eye color, tanning ability, use of sunscreen or personal protective equipment, such as hats, gloves, long pants, etc. were not ascertained. While we were able to control for known RCC risk factors, such as hypertension, smoking, and BMI, other potential risk modifiers (i.e. diet, genetics) were not considered. Additional limitations of our study include non-differential inaccurate or incomplete recall of all occupational histories, non- differential exposure misclassification, and the use of hospital-based controls. While this study had sufficient statistical power to detect relatively small associations, only 39 cases and 30 controls had only high intensity jobs, thus, limiting the precision of our associations within this subgroup.

To our knowledge, this is the first and largest occupational case-control study to investigate the association between UV exposure and RCC risk. Our findings, supported by most UV/cancer studies, demonstrated that occupational UV exposure is associated with reduced renal cancer risk among males. Additional studies that consider recreational UV exposure and behavioral differences, in an effort to reduce exposure misclassification, are warranted, particularly among females.

70 References:

[1] Kimball S, Fuleihan Gel-H, Vieth R. Vitamin D: a growing perspective. Crit Rev Clin

Lab Sci 2008;45:339-414.

[2] International Agency for Research on Cancer. GLOBACAN. Available at: http://www-dep.iarc.fr/ Last Accessed on June 12, 2008.

[3] Grant WB, Garland CF, Gorham ED. An estimate of cancer mortality rate reductions in Europe and the US with 1,000 IU of oral vitamin D per day. Recent Results Cancer

Res 2007;174:225-34.

[4] Calvo MS, Whiting SJ, Barton CN. Vitamin D intake: a global perspective of current status. J Nutr 2005;135:310-6.

[5] Ordonez-Moran P, Larriba MJ, Pendas-Franco N, Aguilera O, Gonzalez-Sancho JM,

Munoz A. Vitamin D and cancer: an update of in vitro and in vivo data. Front Biosci

2005;10:2723-49.

[6] Trump DL, Hershberger PA, Bernardi RJ, Ahmed S, Muindi J, Fakih M, Yu WD,

Johnson CS. Anti-tumor activity of calcitriol: pre-clinical and clinical studies. J Steroid

Biochem Mol Biol 2004;89-90:519-26.

[7] Dietary Supplement Fact Sheet: Vitamin D. National Institutes of Health. Archived

71 from the original on 2007-09-10. Available at: http://www.webcitation.org/5R15u0LB5

Last Accessed on December 25, 2008.

[8] John EM, Schwartz GG, Koo J, Wang W, Ingles SA. Sun exposure, vitamin D receptor gene polymorphisms, and breast cancer risk in a multiethnic population. Am J

Epidemiol 2007;166:1409-19.

[9] Norman AW. Sunlight, season, skin pigmentation, vitamin D, and 25-hydroxyvitamin

D: integral components of the vitamin D endocrine system. Am J Clin Nutr

1998;67:1108-10.

[10] Matsuoka LY, Wortsman J, Haddad JG, Kolm P, Hollis BW. Racial pigmentation and the cutaneous synthesis of vitamin D. Arch Dermatol 1991;127:536-8.

[11] Deeb KK, Trump DL, Johnson CS. Vitamin D signaling pathways in cancer: potential for anticancer therapeutics. Nat Rev Cancer 2007;7:684-700.

[12] Scélo G, Constantinescu V, Csiki I, Zaridze D, Szeszenia-Dabrowska N, Rudnai P,

Lissowska J, Fabiánová E, Cassidy A, Slamova A, Foretova L, Janout V, et al.

Occupational exposure to vinyl chloride, acrylonitrile and styrene and lung cancer risk.

Cancer Causes Control 2004;15:445-52.

[13] Hashibe M, Boffetta P, Zaridze D, Shangina O, Szeszenia-Dabrowska N, Mates D,

72 Fabiánová E, Rudnai P, Brennan P. Contribution of tobacco and alcohol to the high rates

of squamous cell carcinoma of the supraglottis and glottis in Central Europe. Am J

Epidemiol 2007;165:814-20.

[14] ILO. International standard classification of occupations (ISCO). Geneva, 1968

(Rev. Ed.)

[15] Durusoy R, Boffetta P, Mannetje A, Zaridze D, Szeszenia-Dabrowska N, Rudnai P,

Lissowska J, Fabiánová E, Cassidy A, Mates D, Bencko V, Salajka F, et al. Lung cancer risk and occupational exposure to meat and live animals. Int J Cancer 2006;118:2543-7.

[16] Eurostat. NACE Rev, 1: statistical classification of economic activities in the

European community. Luxembourg, 1996.

[17] Mannetje A, Kromhout H. The use of occupation and industry classifications in general population studies. Int J Epidemiol 2003;32:419-28.

[18] Carpenter CR. Kappa statistic. CMAJ 2005;173:15-6.

[19] Hunt JD, van der Hel OL, McMillan GP, Boffetta P, Brennan P. Renal cell carcinoma in relation to cigarette smoking: meta-analysis of 24 studies. Int J Cancer

2005;114:101-8.

73 [20] Boscoe FP, Schymura MJ. Solar ultraviolet-B exposure and cancer incidence and

mortality in the United States, 1993-2002. BMC Cancer 2006;6:264.

[21] Grant WB. The effect of solar UVB doses and vitamin D production, skin cancer

action spectra, and smoking in explaining links between skin cancers and solid tumours.

Eur J Cancer 2008;44:12-5.

[22] Grant WB, Garland CF. Evidence supporting the role of vitamin D in reducing the

risk of cancer. J Intern Med 2002;252:178-9.

[23] Grant WB. An estimate of premature cancer mortality in the U.S. due to inadequate

doses of solar ultraviolet-B radiation. Cancer 2002;94:1867-75.

[24] Mohr SB, Gorham ED, Garland CF, Grant WB, Garland FC. Are low ultraviolet B

and high animal protein intake associated with risk of renal cancer? Int J Cancer

2006;119:2705-9.

[25] Håkansson N, Floderus B, Gustavsson P, Feychting M, Hallin N. Occupational

sunlight exposure and cancer incidence among Swedish construction workers. Epidemiol

2001;12:552-7.

[26] Boffetta P, van der Hel O, Kricker A, Nieters A, de Sanjosé S, Maynadié M, Cocco

PL, Staines A, Becker N, Font R, Mannetje A, Goumas C, et al. Exposure to ultraviolet

74 radiation and risk of malignant lymphoma and multiple myeloma- a multicentre

European case-control study. Int J Epidemiol 2008;37:1080-94.

[27] Slattery ML, Neuhausen SL, Hoffman M, Caan B, Curtin K, Ma KN, Samowitz W.

Dietary calcium, vitamin D, VDR genotypes and colorectal cancer. Int J Cancer

2004;111:750-6.

[28] John EM, Schwartz GG, Koo J, Van Den Berg D, Ingles SA. Sun exposure, vitamin

D receptor gene polymorphisms, and risk of advanced prostate cancer. Cancer Res

2005;65:5470-9.

[29] Mullin GE, Dobs A. Vitamin d and its role in cancer and immunity: a prescription for sunlight. Nutr Clin Pract 2007;22:305-22.

[30] Vieth R. Vitamin D supplementation, 25-hydroxyvitamin D concentrations, and safety. Am J Clin Nutr 1999;69:842-56.

[31] Karami S, Boffetta P, Rothman N, Hung RJ, Stewart T, Zaridze D, Navritalova M,

Mates D, Janout V, Kollarova H, Bencko V, Szeszenia-Dabrowska N, et al. Renal cell carcinoma, occupational pesticide exposure and modification by glutathione S-transferase polymorphisms. Carcinogenesis 2008;29:1567-71.

[32] Broekmans WM, Vink AA, Boelsma E, Klöpping-Ketelaars WA, Tijburg LB, van't

75 Veer P, van Poppel G, Kardinaal AF. Determinants of skin sensitivity to solar irradiation.

Eur J Clin Nutr 2003;57:1222-9.

[33] Calvo MS, Whiting SJ. Prevalence of vitamin D insufficiency in Canada and the

United States: importance to health status and efficacy of current food fortification and dietary supplement use. Nutr Rev 2003;61:107-13.

[34] Oberyszyn TM. Non-melanoma skin cancer: importance of gender, immunosuppressive status and vitamin D. Cancer Lett 2008;261:127-36.

[35] Zouboulis CC, Chen WC, Thornton MJ, Qin K, Rosenfield R. Sexual hormones in human skin. Horm Metab Res 2007;39:85-95.

[36] Newton R, Roman E, Fear N, Carpenter L. Non-Hodgkin's lymphoma and solar ultraviolet radiation. Data are inconsistent. BMJ 1996;313:298.

[37] Karipidis KK, Benke G, Sim MR, Kauppinen T, Giles G. Occupational exposure to ionizing and non-ionizing radiation and risk of glioma. Occup Med (Lond) 2007;57:518-

24.

[38] Peacey V, Steptoe A, Sanderman R, Wardle J. Ten-year changes in sun protection behaviors and beliefs of young adults in 13 European countries. Prev Med 2006;43:460-

5.

76

[39] Glanz K, Buller DB, Saraiya M. Reducing ultraviolet radiation exposure among outdoor workers: state of the evidence and recommendations. Environ Health 2007;6:22.

[40] Hall HI, May DS, Lew RA, Koh HK, Nadel M. Sun protection behaviors of the U.S. white population. Prev Med 1997;26:401-7.

77 Table 1. General characteristics of participants in the Central and Eastern European Renal Cell Carcinoma Study Variables Cases Controls ‡ N%N% P-value Participants 1,097 100.0 1,476 100.0 Sex Males 648 59.1 952 64.5 Females 449 40.9 524 35.5 0.01 Age at Interview <45 86 7.8 122 8.3 45-54 278 25.3 379 25.7 55-64 335 30.5 460 31.2 65-74 353 32.2 452 30.6 75+ 45 4.1 63 4.3 0.61 Mean Age (std) 59.6 years (10.3) 59.3 years (10.3) Center Romania-Bucharest 95 8.7 160 10.8 Poland-Lodz 99 8.7 198 13.4 Russia-Moscow 317 28.9 463 31.4 *Czech Republic 586 53.4 655 44.4 <0.001 Body Mass Index at Interview <25 327 29.8 532 36.2 25-29.9 476 43.4 620 42.1 30+ 293 26.7 319 21.7 <0.001 Tobacco Status Never 510 46.6 599 40.7 Ever 584 53.4 874 59.3 0.003 Hypertension No 600 54.7 906 61.4 Yes 496 45.3 569 38.6 0.001 § Education Level Higher Education 331 31.32 350 24.48 Secondary Education 611 57.81 962 67.27 Primary Education 115 10.88 118 8.252 0.07 Familial History of Cancer No 1st degree relative with cancer 733 66.8 1074 72.8 1st degree relative with cancer 364 33.2 402 27.2 0.001 * Brno, Olomouc, Prague, Ceske-Budejovice ‡ P-values from Chi-square test § Higher education include high school, university, or higher; secondary education includes middle school and apprenticeships; primary education includes finished and unfinished elementary

78 P-trend 4-4.80) 0.92 95% CI for with Subject sure for Subjects with for Subjects sure OR 5 1.17 (0.83-1.65) 9 1.18 (0.84-1.66) 5 1.21 (0.85-1.71) % N ure (low exposure-unit-yrs) % Among Female Participants Female Among justed Duration of Duration justed (yrs) Exposure 6 10.5 7 13.2 1.00 N 134 31.0 187 37.0 1.00 135 31.3 189 37.4 1.00 132 35.2 185 40.9 1.00 Cases Controls 4.50 4.50 4.50 5.70 Only Low Intensity Jobs (yrs) >5.70 167 38.7 175 34.7 1.22(0.88-1.71) 0.24 any Highany Jobs (yrs) Intensity Sunlight >5.80 163 37.7 169 33.5 1.22(0.87-1.70) 0.26 >5.70 116 30.9 129 28.5 1.16(0.81-1.67) 0.41 < < < < 0.04 0.05 0.003 P-trend 95% CI ubject with ubject of Duration Exposure Frequency-Adjusted Subjects with of Duration Expo Frequency-Adjusted OR % 95 25.9 0.62 (0.45-0.85) N 306 33.0 0.76 (0.58-1.00) 326 35.2 0.76 (0.58-0.99) 131 74.9 0.86(0.48-1.52) 0.60 >5.70 51 89.5 46 86.8 1.08(0.2 s (yrs) s (yrs) s Among Male Participants Among % 35 37.6 352 38.011 0.83(0.64-1.08) 33.8 298 32.1 >4.50-5.80 0.92(0.71-1.20) 135 31.375 35.9 149 29. 263 35.0 >4.50-5.70 130 30.1 0.84(0.64-1.11) 141 27. >4.50-5.70 127 33.9 138 30. N to small numbers (Cases N= (Cases 39,numbers to 31) Controls N= small 37 26.8 44 25.1 1.00 197 31.5 269 29.0217 1.00 34.7 303 32.7 1.00 216 44.4 294 39.1 1.00 and (no,self-reported hypertension yes) Cases Controls ve Exposureve (low exposure-unit-yrs) Expos Cumulative ncy-Adjusted Durationncy-Adjusted (yrs) of Exposure Frequency-Ad tionalsunlight 6.00 6.30 6.30 13.50 Sunlight < < < < P-trend s withs Duration for of Exposure Frequency-Adjusted with Duration for of Exposure Frequency-Adjusted S 95% CI with only high intensity jobs (yrs) not shown due jobs shown not (yrs) intensity high only with ontrols, all male ontrols,controls and controls,male female all all OR er, smoking status status never),(ever, index, body mass er,smoking % N Among all Participants all Among % N 32 16.4 42 18.4 1.00 352 33.3 476 33.2 1.00 354 33.5 484 33.8 1.00 350 40.6 474 39.4 1.00 Cases Controls 5.25 5.25 5.25 9.30 Table 2. Risk of cellRisk renal2.Table and carcinomaexposure tooccupa Only Low Jobs (yrs) Intensity < Only Low Intensity Job >5.25-9.55 363>9.55 34.3 476 33.2 342 32.4>5.25-9.30 1.05(0.85-1.31) 375 480>9.30 35.5 33.5 472 1.01(0.81-1.26) 33.0 328 >6.00-16.10 31.0 0.96 1.11(0.90-1.37) 2 >16.10 476>5.25-9.30 33.2 347>9.30 40.3 1.00(0.80-1.25) 193 440 30.9 >6.30-13.50 0.96 36.5 165 2 19.1 >13.50 1.09(0.87-1.36) 290 24.1 197 31.5 0.84(0.64-1.09) >6.30-13.50 0.27 1 >13.50 96 19.7 1 Frequency-Adjusted Duration of Exposure Duration forFrequency-Adjusted Subject any Highany Intensity Jobs (yrs)< any Job High Intensity Frequency-Adjusted Duration of Exposure Duration (yrs)Frequency-Adjusted Freque of Exposure Duration forFrequency-Adjusted Subject < >9.30 163 83.6 186 81.6 1.36(0.76-2.42) 0.30 >13.50 101 73.2 Sunlight (low Exposure Cumulative exposure-unit-yrs) Cumulati < Tertile all based onc among values levels exposure All values adjusted values All for cent age sex, (continuous), Frequeny-adjusted Frequeny-adjusted duration of for exposure subjects

79 P-trend 95% CI Russia OR re for Subject with for Subject re Only Low ure for Subject with forure Subject any High 2.9 NA % N 9.1 42 32.3 1.13(0.58-2.19) % 30.6 45 31.328.5 1.18(0.62-2.26) 42 29.2 1.17(0.61-2.26) Among Russian Female Participants Russian Female Among 0 0.0 1 7.1 1.00 N 37 25.7 4437 30.6 25.7 1.00 44 30.637 26.2 1.00 43 33.1 1.00 Cases Controls ve Exposure (low Exposure ve exposure-unit-yrs) ncy ofDuration Adjusted Exposure (yrs) to small numbers (Cases N=2,Controls N=4) numbers small to rted hypertension rted(no, yes) hypertension 4.50 4.50 4.50 5.70 Intensity JobsIntensity (yrs) < < < < Intensity JobsIntensity (yrs) Sunlight >5.80 63>5.70 43.8 55 66 38.2 45.8>5.70 1.50(0.79-2.82) 58 0.20 40.3 63 44.7 1.49(0.80-2.79) 0.20 45 34.6 1.79(0.93-3.43) 0.09 P-trend <0.001 <0.001 <0.001 tooccupational sunlight amongsubjects Moscow, in with Only Low ofDuration Adjusted Exposu Frequency with any High ofDuration Adjusted Expos Frequency 95% CI OR with Only High Intensity Jobs (yrs) not shown due Jobsnot (yrs) shown Intensity High Only with ols tatus (ever,index, tatus self-repo and never), body mass % N Among Among Russian Participants Male % 8 44.4 17 33.3 1.00 N 69 45.1 7173 25.0 47.7 1.00 81 28.572 53.3 1.00 79 33.9 1.00 Cases Controls 6.00 6.30 6.30 13.50 Table 3. Risk of renalof Table 3.carcinomacell exposure Risk and Intensity JobsIntensity (yrs) < Intensity JobsIntensity (yrs) < Frequency Adjusted Duration ofDuration Adjusted ExposureFrequency (yrs)< ofDuration Adjusted ExposureFrequency for Subject Freque ofDuration Adjusted ExposureFrequency for Subject Sunlight (low Exposure exposure-unit-yrs)Cumulative Cumulati < >6.00-16.10>16.10 56 36.6 28 114>6.30-13.50 18.3 40.1>13.50 51 33.3 99 0.48(0.29-0.79) 29 34.9 102 19.0 35.9>6.30-13.50 0.27(0.15-0.47) >13.50 44 101 0.55(0.34-0.91) 32.6 >4.50-5.80 35.6 19 44 87 0.29(0.17-0.52) 14.1 37.3 >4.50-5.70 67 0.54(0.32-0.92) 28.8 41 0.28(0.15-0.54) >4.50-5.70 41 2 >13.50 10 55.6 34 66.7 0.71(0.19-2.70) 0.62 >5.70 3 100.0 13 9 All values adjusted for age (continuous), smoking s adjusted values All for smoking age (continuous), Frequeny exposureAdjusted Duration Frequeny forof Subjects Tertile contr based values on exposure levels among

80 CHAPTER 4: Analysis of SNPs and Haplotypes in VDR Pathway Genes and Renal

Cancer Risk

Karami S 1, Brennan P 2, Stewart P 10 , Rosenberg PS 1, Navratilova M 6, Mates D 8, Zaridze

D3, Janout V 4, Kollarova H 4, Bencko V 5, Matveev V 3, Szeszenia-Dabrowska N 7,

Holcatova I 5, Yeager M 9, Chanock S 9, Menashe I 1, Rothman N 1, Chow W-H1, Boffetta

P2, Moore LE 1

1Division of Cancer Epidemiology and Genetics, National Cancer Institute, NIH, DHHS,

Bethesda, MD, USA

2International Agency for Research on Cancer, Lyon, France

3Institute of Carcinogenesis, Cancer Research Centre, Moscow, Russia

4Department of Preventive Medicine, Faculty of Medicine, Palacky University, Olomouc,

Czech Republic

5Institute of Hygiene and Epidemiology, Charles University, First Faculty of Medicine,

Prague, Czech Republic

6Department of Cancer Epidemiology and Genetics, Masaryk Memorial Cancer Institute,

Brno, Czech Republic

7Department of Epidemiology, Institute of Occupational Medicine, Lodz, Poland

8Institue of Public Health, Bucharest, Romania

9Core Genotyping Facility at the Advanced Technology Center of the National Cancer

Institute, NIH, Department of Health and Human Services

81 10 Stewart Exposure Assessments, LLC, Arlington, VA, USA, Formerly of the Division of

Cancer Epidemiology and Genetics, National Cancer Institute, Bethesda, MD, USA

Running Title: Vitamin D Pathway and RCC Risk

Key words: Renal cancer, vitamin D, VDR , RXR ,

82 Abstract:

Vitamin D is activated in the kidney. Analysis of variation in the vitamin D receptor

(VDR) and its pathway may provide insight into the role of vitamin D in renal cell

carcinoma (RCC) etiology. RCC cases (N= 987) and controls (N= 1,298) were genotyped

to investigate the relationship between RCC risk and variation in eight target genes.

Minimum-p-value permutation (Min-P) tests were used to identify genes associated with

RCC risk. HaploWalk analysis was used to identify regions with a False

Discovery Rate <10%, where haplotype relative risks were computed. Min-P values

showed VDR (p-value= 0.02) and retinoid-X-receptor-alpha (RXRA ) (p-value= 0.10) genes were strongly associated with risk. Two chromosomal regions of interest within

VDR were identified using HaploWalk. Region1 identified two haplotypes within intron 2 that increased RCC risk approximately 25%. Region2 identified a haplotype

(rs12717991, rs2239179) centered around intron 4 that increased risk among participants with the AG (OR= 1.31, 95% CI= 1.09-1.57) haplotype compared to participants with the homozygous haplotype, AA . Global p-values for both these regions were 0.04. Across the

RXRA gene, one haplotype located downstream, 3 of the coding sequence (rs748964,

rs3118523), increased RCC risk 35% among individuals with the variant haplotype

compared to those with the most common haplotype (Global p-value= 0.03). This is the

first study to comprehensively evaluate genetic variation in VDR and its pathway in relation to RCC risk. Results suggest that genes in the vitamin D pathway may modify

RCC risk. Replication studies are warranted to confirm these findings.

83 Introduction:

Within the kidney, vitamin D is metabolized to 1,25-dihydroxycholecalciferol

(1,25(OH) 2D3), the active form of vitamin D [1, 2] . The anti-carcinogenic properties of

vitamin D include inhibition of clonal tumor cell proliferation, induction of immune cell

differentiation and apoptosis, and decreased angiogenesis [3, 4] . Vitamin D activity is mediated through binding of 1,25(OH) 2D3 to the vitamin D receptor ( VDR ), which can

regulate transcription of other genes involved in cell regulation, growth, and immunity

[4-7] . VDR modulates the expression of genes by forming a heterodimer complex with

retinoid-X-receptors ( RXR ) [8, 9] . This VDR-RXR complex is directed to the vitamin D response elements ( VDREs ) in the promoter region of a broad spectrum of 1,25 regulated genes to initiate transcription [8-10] .

Since 1,25(OH) 2D3, the most biologically active form of vitamin D, exerts its activity through intercellular VDR , most studies of genetic susceptibility to vitamin D related diseases have focused on investigating variation within the VDR gene [7, 11-14] .

However, the concert of genes that interacts with VDR is vast; therefore, an analysis of genetic variation in VDR and other genes in its pathway may provide insight into the role of vitamin D in renal cell carcinoma (RCC) etiology. Given the importance of vitamin D metabolism in the kidney and the lack of studies that have evaluated the relationship between vitamin D, the VDR gene and RCC, we investigated whether variation in VDR and other vitamin D pathway genes modified RCC risk.

In this study, we selected tagging SNPs with 80-90% genomic coverage across the

84 genomic regions of interest to comprehensively assess variation across the chromosomal

region of each gene, thus in turn reducing the potential for false negative findings.

Specifically, we investigated the relationship between RCC risk and 139 single

nucleotide polymorphisms (SNPs) in eight target genes ( VDR, RXRA, RXRB, CYP24A1,

GC, STAT1, THRAP4 and TRAP5) of the expanded vitamin D pathway among cases and controls residing in Central and Eastern Europe, an area with one of the highest rates of

RCC worldwide [15] . To our knowledge, this represents the largest and most detailed

investigation of vitamin D pathway genetic variation and RCC risk conducted to date.

Materials and Methods:

Study Subjects. Details of the Central and Eastern Europe Renal Cell Carcinoma

(CEERCC) study have been previously reported [16] . Briefly, the CEERCC study is a hospital-based case-control study of renal cell cancer coordinated by the United States

(U.S.) National Cancer Institute (NCI) and the International Agency for Research on

Cancer (IARC). Participants were recruited between 1999 and 2003 in seven centers in four Central and Eastern European countries (Moscow, Russia; Bucharest, Romania;

Lodz, Poland; and Prague, Olomouc, Ceske-Budejovice, and Brno, Czech Republic).

Each center followed an identical protocol and was responsible for recruiting a consecutive group of newly diagnosed cases of kidney cancer as well as a comparable group of hospital controls. Controls in all centers were chosen among subjects admitted as inpatients or out-patients in the same hospital as the cases, with conditions unrelated to tobacco or genitourinary disorders, including infections (1.1%), hematologic (3.2%), endocrine (2.0%), psychiatric (1.4%), neurologic (11.2%), ophthalmologic or otologic

85 (14.5%), cardiovascular (9.6%), pulmonary (3.9%), gastrointestinal (18.7%), dermatologic (2.8%), orthopedic or rheumatologic (8.9%), genitourinary (benign prostatic hyperplasia) (3.8%), obstetric or perinatal (0.1%), injury or poisoning (3.0%), and other

(15.9%). No single disease made up more than 20% of the control group. All subjects were residents of the study areas for at least one year at the time of recruitment.

Cases between 20 and 88 years of age were ascertained through a rapid reporting system.

Participating physician and other professional staff members visiting the urology, surgery, radiology, or pathology departments of participating hospitals regularly identified patients admitted for kidney cancer work-up. All patients suspected of having kidney cancer were approached for interview, but only histologically confirmed cases were included in the final analysis. All tumors were centrally reviewed to confirm diagnosis; each RCC case was histologically confirmed by an expert in renal tumor pathology. Controls were frequency-matched to cases on age (±3 years), sex, and study center. Some controls were previously recruited for an earlier study of lung [17] and head and neck cancer [18]. A total of 1,097 incident, histologically confirmed RCC cases and 1,476 controls were included in this study. Suitable quantity and quality of genomic

DNA was obtained from a subset of 987 (90.0%) RCC cases and 1,298 (87.9%) controls.

The response rates across study centers among eligible subjects who were requested to participate ranged from 90.0% to 98.6% for cases and from 90.3% to 96.1% for controls.

Interview. Cases and controls were interviewed with the same questionnaire [19] .

Information on demographic characteristics, medical histories, and lifestyle factors was

86 obtained through in-person interviews by trained personnel using standardized lifestyle

and food frequency questionnaires. Cases were interviewed within three months of

diagnosis. Written consent for participation was obtained from all study subjects, and

ethical approval was obtained for all study centers as well as at the IARC and the NCI.

Laboratory Procedures. Genomic DNA was extracted from whole blood buffy coat

using a standard phenol chloroform method at the NCI laboratory. Genotyping was

conducted with a GoldenGate ® Oligo Pool All (OPA) assay by Illumina®

(www.illumina.com ). DNA samples from RCC cases and controls were coded and

randomized on polymerase chain reaction (PCR) plates for genotyping analyses. A

random 5% duplicate sample was selected and genotyped for quality control.

For this study, eight candidate genes ( VDR, RXRA, RXRB, CYP24A1, GC, STAT1,

THRAP4 and THRAP5 ) were selected for their involvement in vitamin D metabolism, transport, binding, function and/or expression which may affect the mechanisms through which vitamin D may influence cancer risk. Since it is well known that VDR elicits a

transcriptional response by forming a heterodimer complex with RXR , in this study both

VDR and RXR were the primary genes suspected a priori to modify renal cancer risk.

The group specific component (GC) vitamin D binding protein was selected for its involvement as the major carrier of vitamin D and its metabolites in plasma to target tissue [20, 21]. Cytochrome P450, family 24, subfamily A, polypeptide 1 ( CYP24A1 ),

an enzyme involved in the metabolism of vitamin D in target tissue was selected for its

involvement in degrading and regulating 1,25(OH) 2D3 levels [22-24]. Lastly, the signal

87 transducer and activator of transcription 1 ( STAT1) and the thyroid hormone receptor

associated protein 4 and 5 ( THRAP4 and THRAP5 ) genes have been speculated to induce

or suppress transcription by interacting with the VDR-RXR complex [25-27] .

Tag SNPs were selected to provide 80% to 90% genomic coverage across the genomic

regions of interest, while some non-synonymous SNPs were selected for their putative

functional significance. To ensure thorough coverage of the targeted genes, additional

coverage of regions both upstream (20kb) and downstream (10kb) of the target gene were

also analyzed. Therefore, a total of 139 SNPs in eight vitamin D-related pathway target

genes were finally selected for analysis with a variant allele frequency of at least 5% as

reported by NCI’s SNP500Cancer Database ( http://snp500cancer.nci.nih.gov ) [28] and a validated assay at the NCI Core Genotyping Facility (CGF)

(http://cgf.nci.nih.gov/home.cfm ). The genotyping failure rate for the selected SNPs was

4.8% (7/146). The concordance rate between duplicate DNA samples ranged from 93% to 100% and completion rates ranged from 98% to 100%. The genotype frequencies among controls did not differ from the expected Hardy-Weinberg equilibrium proportions

(p >0.05).

Statistical Analysis. The distributions of selected characteristics and known RCC risk

factors (sex, age, smoking habits, hypertension, body mass index (BMI), family history

of cancer, and country of residence) were compared between cases and controls using the

Chi-square test. Characteristics associated with renal cancer were further evaluated to

determine if associations between SNPs and RCC were modified.

88

The associations between individual SNPs and RCC risk were estimated by calculating

odds ratios (ORs) and 95% confidence intervals (95% CI) using unconditional logistic

regression. To avoid redundant analyses, only one SNP was evaluated when high

correlations were observed between two SNPs (r ΄ >0.85). Of the 139 SNPs genotyped, two pairs of SNPs in the VDR gene were found to be highly correlated (r ΄ = 0.95 for

rs154410 and rs731236; r ΄ = 0.89 for rs2239185 and rs2248098) and therefore only one

SNP was used per pair. Risk estimates were calculated for the heterozygous and homozygous variant genotypes relative to the common homozygous genotype assuming an additive model. When the frequency of the homozygous variant allele was less than

5% among controls, a dominant model was used to determine risk estimates for the presence and absence of the variant allele. All regression models were adjusted for age

(continuous), sex, study center, and smoking status (ever, never). Additional risk factors

(i.e. hypertension, BMI, and family history of cancer) were also included in models, however, these variables did not affect risk estimates by at least 10%. To check for lack of fit for each model, Hosmer-Lemeshow goodness of fit test statistics were calculated.

Associations between SNP variants and RCC were assessed using an additive model (i.e. linear test of trend for the number of copies of the variant allele (0, 1, 2)) or a dominant model (i.e. Wald chi-square test for the presence or absence of the variant allele (0, 1)).

To assess global significance of association with genes, a minimum-p-value permutation

(Min-P) test was used for analysis because it corrects for multiple testing while also accounting for correlations between SNPs within a gene [29] . Genes that had a

89 significant Min-P test at a cut off level of 10% were selected for further evaluation for

haplotypes and RCC risk. Additionally, the Bonferroni and Simes Global tests were used

to examine the global significance of association with genes.

HaploWalk, a haplotype-based method, was used to evaluate candidate genes with N= K

SNPs. The HaploWalk procedure considered a 3 SNP window for each SNP from SNP 2 through SNP K-1. For each window, haplotype frequencies in cases and controls were reconstructed using the EM algorithm [30] , and a Wald test was used to screen for association with case-control status [29] . In the initial screening phase, no adjustment was made for potential confounders. The Wald test used a threshold value of 5%, such that haplotypes in cases and controls that had an estimated frequency below the threshold in controls were pooled into an ‘other’ category for testing. To account for multiple testing across the K SNPs, the K-2 p-values (one for each window) were adjusted for multiple comparisons using the False Discovery Rate (FDR)-controlling procedure of

Benjamini and Hochberg [31] . Haplotype windows with an FDR-adjusted p-value of

0.10 or below were promoted to the second stage. In stage two, adjacent SNPs that were significant, along with one SNP on either side, were amalgamated into a candidate block.

The significance of haplotypes in the candidate block was assessed with adjustment for age, sex, center, and smoking status using the haplo.stats procedure assuming a co- dominant model [32] . Adjusted haplotype relative risks and global p-values were also

calculated. When the candidate block was significant at the 5% level in the adjusted

analysis this was considered a noteworthy finding.

90 Haplotype blocks identified with HaploWalk were analyzed in relation to RCC risk using

Haplostats (R version 2.4.0; http://www.r-project.org ) [33] , adjusted for age, sex, study center, and smoking status. Additionally, we used standard methods to identify haplotypes using Haploview program version 3.32 [34] . Linkage disequilibrium (LD) between markers was assessed by calculating pairwise Lewontin’s D` and r 2 using

Haploview among population controls [34] .

Associations between common haplotypes (>5% frequency) and RCC risk were evaluated by computing ORs and 95% CIs using the most common haplotype as the referent category.

All analyses were conducted in STATA 9.0 unless otherwise specified (STATA

Corporation, College Station, TX).

Results:

A description of study participants and known RCC risk factors is provided in Table 1 .

Cases and controls were comparable in age, but cases were more likely to have excess

BMI (>30 kg/m 2), hypertension and a first-degree relative with cancer. The association

with smoking was no longer observed after adjustment for age, BMI, hypertension, study

center, and sex [35] .

Table 2 lists results for global gene-based tests of association with case/control status

using the minimum-p-value permutation test. VDR (p-value= 0.02) and RXRA (p-value=

91 0.10) genes were most strongly associated with RCC risk. The minimum p-value for all tagging SNPs within the gene, the Simes Global test, the Bonferroni correction test, and the minimum false discover rate (FDR) adjusted p-values for the three SNP haplowalk sliding window for all eight genes examined in this study are also presented.

HaploWalk methods identified three significant VDR haplotypes in two regions that increased RCC risk ( Figure 1A ). Haplotype analyses (shown in Table 3 ) revealed two

strong signals within the VDR gene. For the first haplotype, within intron 2, subjects with

the TTG (OR= 1.29; 95% CI= 1.10-1.52) and CCG (OR= 1.25; 95% CI= 1.04-1.51)

haplotype were at a significantly increased risk compared to patients with the most

common referent haplotype, CCA . Increased risk appears to be driven by the A to G

change in the third SNP. The second haplotype, centered around intron 4 of the VDR

gene, was also associated with increased RCC risk among participants with the AG (OR=

1.31; 95% CI= 1.09-1.57) haplotype compared to participants with most common referent

haplotype, AA . The R adjusted global p-values for both these regions were 0.04. For the

RXRA gene ( Figure 1B) , a si ngle haplotype located downstream, 3 of the coding sequence, was shown to be associated with increased RCC risk among individuals with the GG (OR= 1.35; 95% CI= 1.11-1.66) haplotype compared to subjects with the most common haplotype, CA . The R adjusted global p-value for this region was 0.03.

Results for individual analyses of these SNPs are provided in Supplemental Table 1.

Thirteen SNPs in five VDR pathway genes were significantly associated with RCC risk using an additive model (number of significant tag SNPs): VDR (5), RXRA (5),

92 CYP24A1 (1), GC (1) , and STAT1 (1). No statistically significant interactions between

these SNPs and potential confounders (i.e. BMI, age, sex, smoking status, hypertension,

and a family history of cancer) were detected (data not shown). Logistic regression

analyses between RCC risk and each of the 139 SNPs examined were also supported by

the Hosmer-Lemeshow goodness of fit test.

Discussion:

In this analysis of eight genes and 139 tagging SNPs in the VDR pathway, we identified three regions within two genes that were significantly associated with RCC risk. Both genes ( VDR and RXRA ) for which we hypothesized a priori associations with RCC risk were globally significant at the gene level, and contained regions with modest associations (approximately 1.3 fold). In the VDR gene, three haplotypes within two regions (intron 2 and intron 4) were significantly associated with increased risk. Across the RXRA gene, RCC risk was higher among those with one particular haplotype located downstream 3 of the coding region .

Few previous studies of RCC and genetic variations in the VDR pathway have been conducted and this is the first study to our knowledge to comprehensively evaluate genetic variation in VDR and other pathway genes in relation to RCC risk. A few polymorphisms in the VDR gene, such as poly(A) , TaqI , and BsmI , have been speculated to result in variation of VDR expression and are hypothesized to subsequently result in changes to circulating levels of active vitamin D [36-39] . Altered VDR protein expression

has been reported in a number of tumor types including breast, malignant gliomas,

93 prostate, and colon cancer [40] . Moreover, previous epidemiological studies have reported that increased binding of vitamin D to VDR is associated with decreased RCC risk and that active levels of vitamin D in serum are significantly lower in RCC patients compared to population controls [38, 39, 41] . Together these results strongly indicate that vitamin D levels influence RCC risk. While results have been mixed concerning associations between genetic variations in the RXR gene and cancer risk [42-51] , in vitro animal and human studies have shown that elevated levels of RXRA increase the antiproliferative effects of 1,25(OH) 2D3 [46-48] .

Only a handful of studies have comprehensively analyzed genetic variation in the VDR gene in relation to other cancer types. A case-control study of 630 incident prostate cancer patients comprehensively analyzed genetic variation in the VDR gene. In this study, twenty-two SNPs across the VDR gene were selected and genotyped in order to capture a high percentage of variation in the VDR gene [49] . A two-fold increase in prostate cancer risk was observed for two VDR loci (rs2107301 and rs2238135), which were located within introns 2 and 4, respectively [49] . Though information is limited regarding variation in the VDR gene and RCC risk, genetic susceptibility studies for several common VDR gene polymorphisms, usually ( BsmI , FokI , TaqI, and ApaI ), have been shown to modify RCC risk [38, 41] . A recent case-control study of 135 RCC patients found that participants with the AA genotype at the ApaI site had a significant increase in RCC risk (OR= 2.59; p-value= 0.01) compared to participants with the Aa or aa genotype; however, the AA genotype was an independent prognostic factor for cause- specific survival (Relative Risk= 3.3; p-value= 0.04) [41] . A two-fold increase in cancer

94 risk was also reported among participants with the TT genotype at the TaqI site in a second case-control study of 102 RCC patients compared to subjects with the Tt or tt genotype (p-value= 0.001) [38] .

In vitro and animal studies suggest that vitamin D and its metabolites may impede carcinogenesis by stimulating cell differentiation, inhibiting cell proliferation typically characterized by the G0/G1 cell cycle arrest, inducing apoptosis, and suppressing invasiveness, angiogenesis, and metastasis [3-6, 50] . In laboratory studies, vitamin D has been shown to significantly reduce tumor growth for a variety of cancers including colon, breast, prostate and lung [40] . The use of vitamin D or its analogs as chemopreventive agents in humans has been very limited. However, beneficial effects were observed in a clinical trial for patients with inoperable advanced hepatocellular carcinoma (HCC). A daily dose of 10 mg of Seocalcitol for up to one year resulted in a reduction in tumor dimensions among HCC patients [51] . Beneficial effects were also observed among 37 patients with metastatic androgen-independent prostate cancer (AIPC) who were treated with oral calcitriol/docetaxel. High dose treatment of calcitriol/docetaxel showed prostate-specific antigen (PSA) reductions of at least 50%. Additionally, time to progression and survival were promising when compared with other phase II studies of single-agent docetaxel in AIPC [52] .

The VDR variants associated with RCC risk in this study were intronic and were not located near intron-exon boundaries that may produce splicing errors. These alleles seem to have no effect on the expression levels or activity of translated VDR protein and it is

95 hypothesized that they are correlated with other variants within coding regions with functional relevance. In a recently published genetic study, several candidate binding sites in intronic human VDR genes were identified through in silico approaches where

VDR was shown to be directly up-regulated by the transcriptional coactivator p53 [53] . Furthermore, synonymous SNPs are often disregarded in many genetic susceptibility studies based on the assumption that they do not alter amino acid sequence, which would subsequently affect protein structure and function, nor protein expression [54]. Yet, the list of synonymous mutations associated with human diseases is growing [54-56]. Recent genetic studies have shown that synonymous mutations are not random and may play a significant role in disease etiology since synonymous SNPs can affect protein expression and function by altering mRNA stability [54, 57-59]. In our study, the association with synonymous polymorphisms and RCC may be attributed to alterations in the stability of mRNA or due to linkage disequilibrium with other functional variation within the VDR gene or another closely linked gene. Studies solely focusing on genetic variations within exonic regions of DNA that lead to changes in protein sequence may be simplistic and may not represent the larger picture that is evolving between genetic variation and disease [60].

In addition to being the first genetic susceptibility study to comprehensively examine the association between VDR and other pathway genes in relation to RCC risk, an additional strength of our study includes the use of HapMap to tag genes of interest using high (80-

90%) genomic c overage both 5 and 3 of the target genes. In this study we also observed high participation rates, used newly diagnosed cases, included only histologically

96 confirmed cancers, and collected biologic materials from a high proportion of subjects.

The large sample size of this study provided sufficient statistical power to detect relatively small associations between genotypes and risk. Analysis of genes with significant global p-values reduced the risk of Type I errors in our study, while the use of two different Haplotype-based methods (HaploWalk and HaploStats) reduced the possibility of finding significant results based on chance. Inherent limitations stemming

from uncontrolled confounding, such as diet or occupational exposures, may have biased

results. Furthermore, while hospital-based case-control studies have potential limitations

due to the lack of population controls, these studies can improve response rates for the

intense collection of biological specimens and therefore reduce the chances of bias in the

assessment of gene-environment interactions [61]. Lastly, we did not employ a direct

marker for vitamin D among these study subjects. While there is general agreement that

the serum 25(OH)D level is the best indicator of current vitamin D status, the

biochemical marker has a half-life of only 3 weeks [62]. Therefore, a single

measurement of 25(OH)D may not reflect long-term vitamin D status, particularly for

case-control studies where cases have just recently been diagnosed.

In conclusion, among participants in the Central and Eastern European Renal Cell

Carcinoma Study two genes ( VDR and RXRA ) were significantly associated with RCC

risk. Investigation of genetic variation in VDR and other vitamin D pathway genes in

RCC patients will have the potential to improve our understanding of RCC etiology;

therefore, additional studies with large sample sizes and similarly thorough genomic

coverage of VDR pathway genes that include both synonymous and non-synonymous

97 mutations are needed to confirm results. Our a priori hypothesis was supported and

genetic variations in VDR and RXRA did modify RCC risk. Additional studies of genetic

variations of these two genes and RCC risk are warranted.

References:

[1] Norman AW. Sunlight, season, skin pigmentation, vitamin D, and 25-hydroxyvitamin

D: integral components of the vitamin D endocrine system. Am J Clin Nutr

1998;67:1108–10.

[2] Matsuoka LY, Wortsman J, Haddad JG, Kolm P, Hollis BW. Racial pigmentation and

the cutaneous synthesis of vitamin D. Arch Dermatol 1991;127:536–8.

[3] Trump DL, Hershberger PA, Bernardi RJ, et al. Anti-tumor activity of calcitriol: pre-

clinical and clinical studies. J Steroid Biochem Mol Biol 2004;89:519-26.

[4] Ordonez-Moran P, Larriba MJ, Pendas-Franco N, Aguilera O, Gonzalez-Sancho JM,

Munoz A. Vitamin D and cancer: an update of in vitro and in vivo data. Front Biosci

2005;10:2723-49.

[5] Valdivielso JM, Fernandez E. Vitamin D receptor polymorphisms and diseases. Clin

Chim Acta 2006;371:1-12.

[6] Walters MR. Newly Identified actions of the vitamin D endocrine system. Endocr

98 Rev 1992;13:719-64.

[7] Slattery ML. Vitamin D receptor gene (VDR) associations with cancer. Nutr Rev

2007;65(8 Pt 2):S102-4.

[8] Lin R, White JH. The pleiotropic actions of vitamin D. Bio Essays 2003;26:21-8.

[9] Brown AJ, Dusso A, Slatopolsky E. Vitamin D. Am J Physiol Renal Physiol

1999;227:157-75.

[10] Holick MF. Vitamin D: Its role in cancer prevention and treatment. Prog Biophys

Mol Biol 2006;92:49-59.

[11] Maalej A, Petit-Teixeira E, Michou L, Rebai A, Cornelis F, Ayadi H. Association

study of VDR gene with rheumatoid arthritis in the French population. Genes Immun

2005;6(8):707-11.

[12] Fukazawa T, Yabe I, Kikuchi S, et al. Association of vitamin D receptor gene

polymorphism with multiple sclerosis in Japanese. J Neurol Sci 1999;166(1):47-52.

[13] Handoko HY, Nancarrow DJ, Mowry BJ, McGrath JJ. Polymorphisms in the

vitamin D receptor and their associations with risk of schizophrenia and selected

anthropometric measures. Am J Hum Biol 2006;18(3):415-7.

99

[14] Onen IH, Ekmekci A, Eroglu M, Konac E, Yesil S, Biri H. Association of Genetic

Polymorphisms in Vitamin D Receptor Gene and Susceptibility to Sporadic Prostate

Cancer. Exp Biol Med (Maywood).2008;233(120:1608-14.

[15] International Agency for Research on Cancer. GLOBACAN. Available at:

http://www-dep.iarc.fr/ Last Accessed on June 12, 2008.

[16] Brennan P, McKay J, Moore L, et al. Uncommon CHEK2 mis-sense variant and

reduced risk of tobacco-related cancers: case control study. Hum Mol Genet

2007;16(15):1794-801.

[17] Scelo G, Constantinecu V, Csika I, et al. Occupational exposure to vinyl chloride,

acrylonitrile and styrene and lung cancer risk. Cancer Causes Control 2004;15:445-

52.

[18] Hashibe M, Boffetta P, Zaridze D, et al. Contribution of tobacco and alcohol to the

high rates of squamous cell carcinoma of the supraglottis and glottis in Central

Europe. Am J Epidemiol 2007;165:814-20.

[19] Hsu CC, Chow WH, Boffetta P, et al. Dietary risk factors of renal cell carcinoma in

eastern and central Europe. Am J Epidemiol 2007;166:62-70.

100 [20] Yang F, Bergeron JM, Linehan LA, Lalley PA, Sakaguchi AY, Bowman BH.

Mapping and conservation of the group-specific component gene in mouse.

Genomics 1990;7(4):509-16.

[21] Verboven C, Rabijns A, De Maeyer M, Van Baelen H, Bouillon R, De Ranter C. A

structural basis for the unique binding features of the human vitamin D-binding

protein. Nat Struct Biol 2002;9(2):131-6.

[22] Sawada N, Kusudo T, Sakaki T, et al. Novel metabolism of 1 alpha,25-

dihydroxyvitamin D3 with C24-C25 bond cleavage catalyzed by human CYP24A1.

Biochemistry 2004;43(15):4530-7.

[23] Väisänen S, Dunlop TW, Sinkkonen L, Frank C, Carlberg C. Spatio-temporal

activation of chromatin on the human CYP24 gene promoter in the presence of

1alpha,25-Dihydroxyvitamin D3. J Mol Biol 2005;350(1):65-77.

[24] Masuda S, Strugnell SA, Knutson JC, St-Arnaud R, Jones G. Evidence for the

activation of 1alpha-hydroxyvitamin D2 by 25-hydroxyvitamin D-24-hydroxylase:

delineation of pathways involving 1alpha,24-dihydroxyvitamin D2 and 1alpha,25-

dihydroxyvitamin D2. Biochim Biophys Acta 2006;1761(2):221-34.

[25] Puzianowska-Kuznicka M, Nauman A, Madej A, Tanski Z, Cheng S, Nauman J.

Expression of thyroid hormone receptors is disturbed in human renal clear cell

101 carcinoma. Cancer Lett 2000;155(2):145-52.

[26] Vidal M, Ramana CV, Dusso AS. Stat1-vitamin D receptor interactions antagonize

1,25-dihydroxyvitamin D transcriptional activity and enhance -mediated

transcription. Mol Cell Biol 2002;22(8):2777-87.

[27] Matsuzaki J, Tsuji T, Zhang Y, et al. 1alpha,25-Dihydroxyvitamin D3

downmodulates the functional differentiation of Th1 cytokine-conditioned bone

marrow-derived dendritic cells beneficial for cytotoxic T lymphocyte generation.

Cancer Sci 2006;97(2):139-47.

[28] Packer BR, Yeager M, Burdett L, et al. SNP500Cancer: a public resource for

sequence validation, assay development, and frequency analysis for genetic variation

in candidate genes. Nucleic Acids Res 2006;34(Database issue):D617-21.

[29] Chen BE, Sakoda LC, Hsing AW, Rosenberg PS. Resampling-Based Hypothesis

Testing Procedures for Genetic Case-Control Association Studies. Gentic

Epidemiology 2006;30:495-507.

[30] Excoffier L, Slatkin M. Maximum-likelihood estimation of molecular haplotype

frequencies in a diploid population. Mol Biol Evol 1995;12:921-7.

[31] Benjamini Y, Hochberg Y. Controlling the false discovery rate: a practical and

102 powerful approach to multiple testing. Journal of the Royal Statistical Society Series

B (methodological) 1995;57:289-300.

[32] Lake SL, Lyon H, Tantisira K, et al. Estimation and tests of haplotype-environment

interaction when linkage phase is ambiguous. Hum Hered 2003;55:56-65.

[33] R Project for Statistical Computing. Available at: http://www.r-project.org/ Last

Accessed on March 14, 2008.

[34] Barrett JC, Fry B, Maller J, Daly MJ. Haploview: analysis and visualization of LD

and haplotype maps. Bioinformatics 2005;21(2):263-5.

[35] Hunt JD, van der Hel OL, McMillan GP, Boffetta P, Brennan P. Renal cell

carcinoma in relation to cigarette smoking: meta-analysis of 24 studies. Int J Cancer

2005;114:101-8.

[36] Ma J, Stampfer MJ, Gann PH, et al. Vitamin D receptor polymorphisms, circulating

vitamin D metabolites, and risk of prostate cancer in United States physicians. Cancer

Epidemiol Biomarkers Prev 1998;7(5):385-90.

[37] Morrison NA, Qi JC, Tokita A, et al. Prediction of bone density from vitamin D

receptor alleles. Nature 1994;367(6460):284-7.

103 [38] Ikuyama T, Hamasaki T, Inatomi H, Katoh T, Muratani T, Matsumoto T.

Association of vitamin D receptor gene polymorphism with renal cell carcinoma in

Japanese. Endocr J 2002;49(4):433-8.

[39] Fujiok T, Suzuki Y, okamoto T. prevention of renal cell carcinoma by active vitamin

D3. World J Surg 2000; 24:1015-20

[40] Madej A, Puzianowska-Kuznicka M, Tanski Z, Nauman J, Nauman A. Vitamin D

Receptor Binding to DNA Is altered without the change in its expression in human

renal clear cell canacer. Nephron Exp Nephrol 2003;93:150-157.

[41] Obara W, Suzuki Y, Kato K, Tanji S, Konda R, Fujioka T. Vitamin D receptor gene

polymorphisms are associated with increased risk and progression of renal cell

carcinoma in a Japanese population. Int J Urol 2007;14(6):483-7.

[42] Martinet N, Alla F, Farré G, et al. and

alterations in lung cancer precursor lesions. Cancer Res 2000;60(11):2869-75.

[43] Lawrence JA, Merino MJ, Simpson JF, Manrow RE, Page DL, Steeg PS. A high-risk

lesion for invasive breast cancer, ductal carcinoma in situ, exhibits frequent

overexpression of retinoid X receptor. Cancer Epidemiol Biomarkers Prev

1998;7(1):29-35.

104 [44] Buentig N, Stoerkel S, Richter E, Dallmann I, Reitz M, Atzpodien J. Predictive

impact of retinoid X receptor-alpha-expression in renal-cell carcinoma. Cancer

Biother Radiopharm 2004;19(3):331-42.

[45] Kong G, Kim HT, Wu K, et al. The retinoid X receptor-selective retinoid, LGD1069,

down-regulates cyclooxygenase-2 expression in human breast cells through

transcription factor crosstalk: implications for molecular-based chemoprevention.

Cancer Res 2005;65(8):3462-9.

[46] Almasan A, Mangelsdorf DJ, Ong ES, Wahl GM, Evans RM. Chromosomal

localization of the human retinoid X receptors. Genomics 1994;20(3):397-403.

[47] Tokar EJ, Webber MM. Cholecalciferol (vitamin D3) inhibits growth and invasion

by up-regulating nuclear receptors and 25-hydroxylase (CYP27A1) in human prostate

cancer cells. Clin Exp Metastasis 2005;22(3):275-84.

[48] Prüfer K, Schröder C, Hegyi K, Barsony J. Degradation of RXRs influences

sensitivity of rat osteosarcoma cells to the antiproliferative effects of calcitriol. Mol

Endocrinol 2002;16(5):961-76

[49] Holick CN, Stanford JL, Kwon EM, Ostrander EA, Nejentsev S, Peters U.

Comprehensive association analysis of the vitamin D pathway genes, VDR,

CYP27B1, and CYP24A1, in prostate cancer. Cancer Epidemiol Biomarkers Prev

105 2007;16(10):1990-9.

[50] Bouillon R, Eelen G, Verlinden L, Mathieu C, Carmeliet G, Verstuyf A. Vitamin D

and cancer. J Steroid Biochem Mol Biol 2006 Dec;102(1-5):156-62..

[51] Dalhoff K, Dancey J, Astrup L, et al. A phase II study of the vitamin D analogue

Seocalcitol in patients with inoperable hepatocellular carcinoma. Br J Cancer 2003

Jul 21;89(2):252-7.

[52] Beer TM, Eilers KM, Garzotto M, Egorin MJ, Lowe BA, Henner WD. Weekly high-

dose calcitriol and docetaxel in metastatic androgen-independent prostate cancer. J

Clin Oncol 2003 ;21(1):123-8.

[53] Maruyama R, Aoki F, Toyota M, et al. Comparitive genome analysis identifies the

vitamin D receptor gene as a direct target of p53-mediated transcriptional activation.

Cancer Res 2006;66(9):4574-83.

[54] Sauna ZE, Kimchi-Sarfaty C, Ambudkar SV, Gottesman MM. Silent Polymorphisms

Speak: How They Affect Pharmacogenomics and the Treatment of Cancer. Cancer

Res 2007; 67:9609-12.

[55] Chamary JV, Parmley JL, Hurst LD. Hearing silence: non-neutral evolution at

synonymous sites in mammals. Nat Rev Genet 2006;7:98–108.

106

[56] Laws SM, Hone E, Gandy S, Martins RN. Expanding the association between the

APOE gene and the risk of Alzheimer's disease: possible roles for APOE promoter

polymorphisms and alterations in APOE transcription. J Neurochem 2003;84:1215–

36.

[57] Nackley AG, Shabalina SA, Tchivileva IE, et al. Human catechol-O-

methyltransferase haplotypes modulate protein expression by altering mRNA

secondary structure. Science 2006;314:1930–3.

[58] Capon F, Allen MH, Ameen M, et al. A synonymous SNP of the corneodesmosin

gene leads to increased mRNA stability and demonstrates association with psoriasis

across diverse ethnic groups. Hum Mol Genet 2004;13:2361–8.

[59] Sauna ZE, Kimchi-Sarfaty C, Ambudkar SV, Gottesman MM. The sounds of

silence: synonymous mutations affect function. Pharmacogenomics 2007;8:527–32.

[60] Nejentsev S, Godfrey L, Snook H, et al. Comparative high-resolution analysis of

linkage disequilibrium and tag single nucleotide polymorphisms between populations

in the vitamin D receptor gene. Hum Mol Genet 2004;13(15):1633-9.

[61] Wacholder S, Chatterjee N, Hartge P. Joint effect of genes and environment

distorted by selection biases: implications for hospital-based case-control studies.

107 Cancer Epidemiol Biomarkers Prev 2002;11:885-889.

[62] Core Genotyping Facility. Access at: http://cgf.nci.nih.gov/home.cfm Last accessed

on March 14, 2008.

108 Figure 1: HaploWalk and Haploview analyses.

A, two chromosomal regions of interest were identified in the Vitamin D (1, 25-

dihydroxycholecalciferol D 3) Receptor gene that were significant at an FDR of 10%.

Region one included SNPs 3'-, rs4760648, rs2853564, rs2254210-5'; region two included

SNPs 3'-rs12717991, rs2239179-5'.

B, a single region of interest was identified in the retinoid-X-receptor, alpha gene that was significant at an FDR of 10%. This region included SNPs 5'-rs748964, rs3118523-3'.

109

110

Legend

111

Table 1: General characteristics of participants in the Central & Eastern European renal cell carcinoma study

Genotyped Participants Variables Cases Controls N % N % ŧ P-value Participants 987 43.2 1,298 56.8

Sex Males 589 59.7 838 64.6 Females 398 40.3 460 35.4 0.02 Age at Interview <45 76 7.7 108 8.3 45-54 253 25.6 333 25.7 55-64 303 30.7 405 31.2 65-74 313 31.7 397 30.6 75+ 42 4.3 55 4.2 0.58 Mean Age (std) 59.5 years (10.3) 59.3 years (10.3) Center Romania-Bucharest 91 9.2 132 10.2 Poland-Lodz 81 8.7 197 15.2 Russia-Moscow 288 29.2 368 28.4 *Czech Republic 527 53.4 601 46.3 <0.001 BMI at Interview <25 288 29.2 458 35.3 25-29.9 429 43.5 556 42.8 30+ 270 27.4 284 21.9 <0.001 Tobacco Status Never 454 46.1 528 40.7 Ever 530 53.9 768 59.3 0.01 Hypertension No 53954.7 80061.7 Yes 447 45.3 497 38.3 0.001 Familial History of Cancer No 1st degree relative with cancer 654 66.3 932 71.8 1st degree relative with cancer 333 33.7 366 28.2 0.004 * Brno, Olomouc, Prague, Ceske-Budejovice ‡ P-values from Chi-square test

112 Haplowalk † † Bonferroni Correction * Adjusted * Simes Test Minimum ± * ± Adjusted p-trend *Adjusted Min-P Test Tag SNPs Tag Number of ith renal cancer risk 20q13 35 0.630 0.035 0.895 1.873 0.729 9q34.3 18 0.100 0.011 0.106 0.155 0.007 2q32.2 21 0.138 0.022 0.304 0.330 0.085 6p21.3 8 0.202 0.093 0.541 0.734 0.262 19p13.3 11 0.384 0.113 0.693 1.056 0.192 17q21.1 6 0.936 0.580 0.983 3.789 0.927 12q13.11 29 0.024 0.002 0.023 0.029 0.045 4q12-q13 11 0.246 0.037 0.399 0.449 0.090 Location and g g 3 rting rting D by by D ein e- ith gh the the gh and cell iating iating the ; ;

hways. mo- or mo- 3 cium cium liding window analysis liding window , and haplowalk, and minimump-values forassociations w targeted gene using additive model additivetargeted using model gene tatus (ever, never) Mediates the biological effects retinoids; of ho as bindsheterodimers, receptorto D vitamin gene the regulates transcription. nucleus wherenucleus they actas transcription activators. forms homo- or heterodimers that translocatehomo- forms to the Phosphorylated by receptorPhosphorylated by associated kinases; prot receptor regulatesa of variety other metabolic pat is principally involved in mineral metabolism principallythrouis metabolism inmineral involved them to target tissues. them Binds vitamin D and its transpoplasma D vitamin metabolites,Binds interactsSelectively receptors, D vitamin med with transcription genes. of hormone-sensitive the action of vitamin D by binding and controlling andaction binding by the D vitamin of regulates levels of vitamin D3; plays aplays regulates levels in roleD3; vitamin of cal endocrine D vitamin homeostasis and system. Forms homodimers with the with retinoichomodimers acid, Forms thyroid receptors, D vitamin increasing and hormone, bindin and transcriptionand responseon elements. Encodes the nuclearEncodes the receptor hormone D forvitamin Initiates degradation of 1,25-dihydroxyvitamin D Initiates degradationof 1,25-dihydroxyvitamin Induces gene expression and selectivelyInduces interacts w receptors Dthe action vitamin mediating vitamin of controlling and binding the transcription hormon of sensitive genes. ) (DAQB- ) Function Gene Target Chromosome Alias CP24, ) ) DRIP92, MED16, DRIP92, ARC100, CRSP100, CRSP100, ARC100, DBP ) NR111) : Cytochrome P450, Family 2, 2, P450, Family Cytochrome : : Thyroid Hormone Receptor: Thyroid Hormone : Thyroid Hormone Receptor: Thyroid Hormone :Transducer Signal Activator and Target GeneTarget ( Name ) : Retinoid: X Receptor, alpha : Retinoid: X Receptor, beta ) : Vitamin D 1,25- dihydroxyvitamin dihydroxyvitamin Vitamin1,25- : D : Group-Specific Component Vitamin : Component Group-Specific Receptor ( Adjusted s for center,age, sex, study and smoking 3 Minimum FDR adjusted p-value for 3- SNP haplowalk s SNP for adjusted FDR 3- Minimum p-value DKFZp686B04100, ISGF-3, STAT91 ISGF-3, DKFZp686B04100, FLJ16020, FLJ16733, MGC102720, FLJ16733, MGC102720, FLJ16020, Adjusted minimum p-value for all tagging forSNPs in a all p-value tagging Adjusted minimum STAT1 Transcriptionof 91kDa 1, ( RXRA ( NR2B1 RXRB THRAP5 TRAP95) CYP24A1 Polypeptide( A, 1 Subfamily P450- MGC126274,CYP24, MGC126273, CC24 Associated Protein( 5 GC ProteinBinding D ( 314F24.5, DAUDI6, H-2RIIBP, H-2RIIBP, 314F24.5, DAUDI6, RCoR-1 NR2B2, MGC1831, Table 2. Vitamin Table 2. D pathway gene-based global, trend VDR D THRAP4 DRIP100, MGC8748, KIAA0130, CRSP4, TRAP100 MED24, Associated Protein( 4 ± † † *,

113 Table 3. Haplotype associations with genes in the vitamin D pathway Adjusted Haplotypes Cases(%) Controls(%) *OR *(LCI-UCI) *P-value Global P † VDR Region 1 3'-rs4760648, rs2853564, rs2254210-5' C-C-A 37.8 42.7 1.00 T-T-G 32.3 28.9 1.29 (1.10-1.52) 0.002 C-C-G 20.3 18.4 1.25 (1.04-1.51) 0.020 C-T-G 6.5 7.1 1.04 (0.78-1.38) 0.777 0.042 Region 2 3'-rs12717991, rs2239179-5' A-A 37.9 41.6 1.00 G-G 37.0 37.1 1.07 (0.92-1.25) 0.360 A-G 23.6 19.7 1.31 (1.09-1.57) 0.004 0.041 ‡ RXRA Region 1 5'-rs748964, rs3118523-3' C-A 78.9 82.8 1.00 C-A 6.5 5.9 1.11 (0.85-1.46) 0.441 G-G 14.3 11.1 1.35 (1.11-1.66) 0.003 0.027 *Adjusted for age (continuous), sex, study center, and smoking habit (ever, never) † VDR chr12 region1: 46466932-46559981; region 2: 46545393-46544033; ‡ RXRA chr9 region1: 136475342-136473910

114 Table S1. SNP-based analyses of genes in the extended vitamin D pathway Cases Controls SNP N % N % OR LCI - UCI P-value P-trend

rs1555439 (-15343G>T) (PFDN4-01) GG 508 65.5 679 65.9 1.00 GT/TT 268 34.5 352 34.1 1.04 0.85 - 1.27 0.73

0.749 rs2585421 (-16904A>G) (CYP24A1-54) AA 550 70.8 774 74.8 1.00 AG/GG 227 29.2 261 25.2 1.18 0.95 - 1.46 0.13

0.174 rs2252928 (-16551T>A) (CYP24A1-85) TT 609 78.4 816 79.1 1.00 AT/AA 168 21.6 215 20.9 1.09 0.86 - 1.37 0.49

0.360 rs6023012 (-16454T>C) (CYP24A1-83) TT 297 38.6 414 40.3 1.00 CT 361 46.9 487 47.4 1.01 0.82 - 1.24 0.92 CC 111 14.4 127 12.4 1.20 0.89 - 1.62 0.24 0.332 rs765058 (-14373T>C) (CYP24A1-84) TT 411 53.3 571 55.3 1.00 CT 294 38.1 396 38.4 1.01 0.83 - 1.24 0.90 CC 66 8.6 65 6.3 1.47 1.01 - 2.14 0.04 0.146 rs2585424 (-8460C>A) (CYP24A1-86) CC 640 82.7 894 86.4 1.00 AC/AA 134 17.3 141 13.6 1.34 1.03 - 1.74 0.03

0.070 rs2208588 (-6852A>T) (CYP24A1-81) AA 228 29.3 334 32.3 1.00 AT 381 49.0 500 48.3 1.09 0.87 - 1.35 0.47 TT 168 21.6 201 19.4 1.22 0.93 - 1.60 0.16 0.163 rs2426498 (-6562G>C) (CYP24A1-52) GG 598 77.0 807 78.0 1.00 CG/CC 179 23.0 228 22.0 1.02 0.81 - 1.29 0.85

0.894 rs2248359 (-1399G>A) (CYP24A1-01) GG 268 34.5 404 39.1 1.00 AG 374 48.2 473 45.7 1.19 0.96 - 1.46 0.11 AA 134 17.3 157 15.2 1.33 1.00 - 1.76 0.05 0.035 rs6022999 (IVS3+103T>C)

115 (CYP24A1-82) TT 472 60.8 651 63.0 1.00 CT/CC 304 39.2 383 37.0 1.14 0.93 - 1.39 0.20

0.211 rs13038432 (IVS3+814T>C) (CYP24A1-50) TT 680 89.6 886 86.9 1.00 CT/CC 79 10.4 133 13.1 0.81 0.60 - 1.10 0.19

0.233 rs4809960 (IVS4+58A>G) (CYP24A1-67) AA 422 54.4 605 58.6 1.00 AG 305 39.3 360 34.8 1.18 0.96 - 1.44 0.12 GG 49 6.3 68 6.6 1.00 0.67 - 1.49 0.99 0.314 rs4809959 (IVS4+272T>C) (CYP24A1-79) TT 238 30.6 277 26.8 1.00 CT 369 47.5 511 49.5 0.84 0.67 - 1.05 0.12 CC 170 21.9 245 23.7 0.80 0.61 - 1.05 0.11 0.094 rs2181874 (IVS4+1653C>T) (CYP24A1-07) CC 450 58.0 619 60.0 1.00 CT 274 35.3 351 34.0 1.08 0.88 - 1.32 0.48 TT 52 6.7 61 5.9 1.19 0.80 - 1.77 0.40 0.315 rs2762941 (IVS4-1280C>T) (CYP24A1-63) CC 344 44.7 462 44.9 1.00 CT 339 44.1 436 42.4 1.08 0.88 - 1.33 0.44 TT 86 11.2 130 12.6 0.97 0.71 - 1.33 0.85 0.848 rs3787557 (IVS4-763A>G) (CYP24A1-66) AA 558 71.8 740 71.5 1.00 AG/GG 219 28.2 295 28.5 0.97 0.78 - 1.19 0.74

0.773 rs2244719 (IVS4-486G>A) (CYP24A1-51) GG 192 25.2 266 26.0 1.00 AG 394 51.6 504 49.2 1.03 0.81 - 1.30 0.83 AA 177 23.2 254 24.8 0.93 0.71 - 1.23 0.61 0.623 rs3787554 (IVS4-308C>T) (CYP24A1-64) CC 623 80.8 850 82.4 1.00 CT/TT 148 19.2 182 17.6 1.07 0.83 - 1.36 0.61

0.477 rs3886163 (IVS6-792C>T) (CYP24A1-78) CC 626 80.6 828 80.0 1.00 CT/TT 151 19.4 207 20.0 0.96 0.75 - 1.22 0.71

0.810 rs912505 (IVS7-1179T>C) (CYP24A1-76) TT 474 61.7 643 62.4 1.00 CT/CC 294 38.3 387 37.6 1.01 0.83 - 1.23 0.94

116

0.617 rs1570669 (IVS9+198T>C) (CYP24A1-20) TT 307 39.5 400 38.6 1.00 TC 361 46.5 494 47.7 0.92 0.75 - 1.13 0.42 CC 109 14.0 141 13.6 1.00 0.74 - 1.35 0.995 0.760 rs927650 (IVS11 +967A>G) (CYP24A1-77) AA 236 30.4 295 28.6 1.00 AG 376 48.5 522 50.5 0.88 0.70 - 1.10 0.25 GG 164 21.1 216 20.9 0.96 0.73 - 1.25 0.74 0.641 rs6097807 (*4262T>C) (CYP24A1-70) TT 436 56.1 593 57.3 1.00 CT 300 38.6 371 35.8 1.06 0.87 - 1.30 0.57 CC 41 5.3 71 6.9 0.76 0.51 - 1.15 0.20 0.643 rs6068810 (*4366A>C) (CYP24A1-69) AA 687 88.4 927 89.6 1.00 AC/CC 90 11.6 108 10.4 1.15 0.85 - 1.57 0.36

0.420 rs6097801 (Ex12+2555C>T) (CYP24A1-80) CC 552 71.0 728 70.3 1.00 CT/TT 225 29.0 307 29.7 0.93 0.75 - 1.15 0.49

0.526 rs8124792 (*6910T>C) (CYP24A1-75) TT 696 89.6 922 89.2 1.00 CT/CC 81 10.4 112 10.8 0.94 0.69 - 1.28 0.67

0.679 rs2762929 (*7522G>A) (CYP24A1-60) GG 302 39.0 419 40.5 1.00 AG 361 46.6 468 45.3 1.03 0.84 - 1.27 0.76 AA 112 14.5 147 14.2 1.05 0.78 - 1.41 0.73 0.696 rs6022985 (*8104G>C) (CYP24A1-68) GG 351 45.5 491 48.0 1.00 CG 362 46.9 437 42.8 1.13 0.93 - 1.39 0.22 CC 59 7.6 94 9.2 0.85 0.59 - 1.22 0.39 0.931 rs2031343 (*9854C>T) (CYP24A1-87) CC 624 80.8 863 83.7 1.00 CT/TT 148 19.2 168 16.3 1.24 0.97 - 1.60 0.09

0.155 rs2762927 (*10751T>G) (CYP24A1-59) TT 340 43.9 446 43.2 1.00 GT 340 43.9 462 44.7 0.94 0.77 - 1.15 0.55 GG 95 12.3 125 12.1 0.99 0.73 - 1.36 0.97 0.771

117 rs2585413 (*11237T>C) (CYP24A1-53) TT 395 51.2 519 50.4 1.00 CT 307 39.8 429 41.7 0.92 0.76 - 1.13 0.44 CC 70 9.1 81 7.9 1.14 0.80 - 1.62 0.48 0.958 rs9305467 (IVS19-5122G>A) (SFRS15-07) GG 685 88.2 893 86.4 1.00 AG/AA 92 11.8 141 13.6 0.89 0.67 - 1.19 0.43

0.439 rs2833477 (IVS19-4197A>G) (SFRS15-04) AA 601 77.3 804 77.7 1.00 AG/GG 176 22.7 231 22.3 1.02 0.81 - 1.28 0.86

0.807 rs2833476 (IVS19-2651A>G) (SFRS15-03) AA 672 86.5 905 87.4 1.00 AG/GG 105 13.5 130 12.6 1.11 0.84 - 1.48 0.46

0.611 rs202449 (IVS19-1616A>T) (SFRS15-06) AA 543 70.6 690 67.1 1.00 AT/TT 226 29.4 338 32.9 0.87 0.71 - 1.07 0.20

0.339

rs16847050 (-17595G>A) (GC-15) GG 577 74.5 806 77.9 1.00 AG/AA 198 25.5 229 22.1 1.24 0.99 - 1.56 0.06

0.037 rs3733359 (Ex1-97C>T) (GC-12) CC 716 92.1 947 91.5 1.00 CT/TT 61 7.9 88 8.5 0.92 0.65 - 1.31 0.66

0.692 rs222029 (IVS1+4716C>T) (GC-10) CC 556 72.6 764 74.5 1.00 CT/TT 210 27.4 262 25.5 1.12 0.90 - 1.39 0.31

0.286 rs1352843 (IVS1+6724A>G) (GC-05) AA 622 80.1 858 82.9 1.00 AG/GG 155 19.9 177 17.1 1.21 0.95 - 1.55 0.13

0.209 rs16847015 (IVS1-3707G>T) (CG-07) GG 740 95.4 969 93.7 1.00 GT/TT 36 4.6 65 6.3 0.73 0.47 - 1.12 0.15

0.161 rs222035 (IVS8+755A>C)

118 (GC-11) AA 248 32.2 328 31.8 1.00 AC 391 50.7 520 50.5 0.93 0.75 - 1.16 0.52 CC 132 17.1 182 17.7 0.89 0.67 - 1.18 0.40 0.381 rs1491709 (IVS11-1644C>T) (GC-06) CC 710 91.5 931 90.1 1.00 CT/TT 66 8.5 102 9.9 0.82 0.59 - 1.15 0.25

0.303 rs705117 (IVS12-528G>A) (GC-13) GG 591 76.5 811 78.7 1.00 AG/AA 182 23.5 219 21.3 1.11 0.88 - 1.39 0.38

0.260 rs17467825 (Ex13+1894T>C) (GC-09) TT 376 49.3 485 47.4 1.00 CT 323 42.3 446 43.6 0.90 0.73 - 1.10 0.29 CC 64 8.4 93 9.1 0.85 0.60 - 1.22 0.39 0.235 rs17383291 (Ex13+2001A>C) (GC-08) AA 657 85.2 852 82.8 1.00 AC/CC 114 14.8 177 17.2 0.80 0.62 - 1.04 0.10

0.063 rs1491711 (*9640G>C) (GC-14) GG 300 38.7 425 41.1 1.00 CG 372 48.0 474 45.8 1.16 0.94 - 1.42 0.17 CC 103 13.3 135 13.1 1.18 0.87 - 1.61 0.28 0.163

rs3132288 (Ex2C>T) (LOC 642985 -01 ) CC 262 33.9 332 32.3 1.00 CT 363 47.0 491 47.7 0.91 0.73 - 1.13 0.40 TT 147 19.0 206 20.0 0.91 0.69 - 1.19 0.48 0.427 rs9409929 (*12052A>G) (RXRA-49) AA 302 38.9 415 40.1 1.00 AG 362 46.6 460 44.4 1.04 0.85 - 1.28 0.69 GG 113 14.5 160 15.5 0.95 0.71 - 1.27 0.75 0.900 rs1007971 (*7453G>C) (RXRA-12) GG 484 62.4 697 68.3 1.00 CG/CC 292 37.6 323 31.7 1.27 1.04 - 1.55 0.02

0.053 rs3118523 (*7060A>G) (RXRA-28) AA 482 62.2 714 69.9 1.00 AG/GG 293 37.8 308 30.1 1.40 1.14 - 1.71 0.001

0.011 rs748964 (*5628C>G) (RXRA-51) CC 563 72.5 811 78.8 1.00

119 CG/GG 214 27.5 218 21.2 1.42 1.13 - 1.77 0.002

0.006 rs877954 (IVS9+355A>G) (RXRA-48) AA 349 44.9 448 43.5 1.00 AG 327 42.1 466 45.3 0.92 0.75 - 1.13 0.44 GG 101 13.0 115 11.2 1.13 0.83 - 1.54 0.44 0.800 rs3132294 (IVS8+278T>C) (RXRA-33) TT 442 56.9 583 56.6 1.00 CT 275 35.4 394 38.3 0.96 0.78 - 1.17 0.67 CC 60 7.7 53 5.1 1.49 1.00 - 2.23 0.05 0.312 rs6537944 (IVS5-694T>C) (RXRA-53) TT 688 89.1 888 86.7 1.00 CT/CC 84 10.9 136 13.3 0.78 0.58 - 1.05 0.10

0.148 rs4240705 (IVS5-2122G>A) (RXRA-38) GG 328 42.3 425 41.3 1.00 AG 336 43.3 479 46.5 0.92 0.75 - 1.13 0.40 AA 112 14.4 126 12.2 1.15 0.86 - 1.56 0.35 0.688 rs3118536 (IVS4-542A>C) (RXRA-31) AA 518 66.7 725 70.3 1.00 AC/CC 259 33.3 306 29.7 1.22 1.00 - 1.50 0.05

0.023 rs3132296 (IVS4+1666C>T) (RXRA-34) CC 359 46.2 488 47.5 1.00 CT 322 41.4 438 42.6 1.03 0.84 - 1.26 0.79 TT 96 12.4 102 9.9 1.31 0.96 - 1.81 0.09 0.165 rs10776909 (IVS1-4732T>C) (RXRA-15) TT 475 61.2 663 64.1 1.00 CT/CC 301 38.8 372 35.9 1.17 0.96 - 1.42 0.13

0.053 rs11103473 (IVS1-5849T>A) (RXRA-19) TT 316 40.7 433 41.9 1.00 AT 345 44.4 470 45.5 1.04 0.85 - 1.28 0.71 AA 116 14.9 130 12.6 1.27 0.94 - 1.71 0.12 0.165 rs7039190 (IVS1-26774A>C) (RXRA-54) AA 692 89.1 905 87.4 1.00 AC/CC 85 10.9 130 12.6 0.87 0.65 - 1.18 0.37

0.418 rs11185662 (IVS1-31359T>C) (RXRA-52) TT 453 58.4 601 58.4 1.00 CT 275 35.4 375 36.4 0.97 0.79 - 1.18 0.74 CC 48 6.2 53 5.2 1.30 0.85 - 1.99 0.22

120 0.583 rs7871655 (IVS1+34281G>C) (RXRA-55) GG 432 55.6 590 57.5 1.00 CG/CC 345 44.4 436 42.5 1.06 0.87 - 1.29 0.55

0.269 rs881658 (IVS1+11350C>T) (RXRA-56) CC 352 45.4 468 45.3 1.00 CT 346 44.6 466 45.1 0.97 0.79 - 1.18 0.74 TT 78 10.1 99 9.6 1.05 0.75 - 1.47 0.77 0.971 rs4917348 (-4332A>G) (RXRA-58) AA 516 66.4 698 67.6 1.00 AG/GG 261 33.6 335 32.4 1.06 0.86 - 1.30 0.59

0.520

rs9277936 (*3071A>T) (RING1-07) AA 377 48.5 543 52.5 1.00 AT 335 43.1 412 39.8 1.16 0.95 - 1.41 0.15 TT 65 8.4 80 7.7 1.13 0.79 - 1.62 0.51 0.202 rs2854028 (Ex6-91C>T) (RING1-10) CC 403 51.9 588 56.9 1.00 CT 324 41.7 385 37.2 1.22 1.00 - 1.49 0.05 TT 50 6.4 61 5.9 1.15 0.77 - 1.72 0.51 0.093 rs421446 (*351G>A) (HSD17B8-01) GG 428 55.5 584 56.8 1.00 AG 296 38.4 382 37.1 1.05 0.86 - 1.29 0.62 AA 47 6.1 63 6.1 0.95 0.63 - 1.43 0.79 0.876 rs1547387 (Ex3-8C>G) (SLC39A7-01) CC 583 75.0 779 75.6 1.00 CG/GG 194 25.0 252 24.4 1.04 0.83 - 1.29 0.75

0.586 rs6531 (Ex7+29C>T) (RXRB-13) CC 446 57.4 562 54.5 1.00 CT 287 36.9 400 38.8 0.90 0.74 - 1.10 0.30 TT 44 5.7 70 6.8 0.82 0.55 - 1.23 0.34 0.198 rs2269346 (IVS1+1038G>A) (COL11A2-02) GG 715 92.0 941 91.2 1.00 AG/AA 62 8.0 91 8.8 0.91 0.64 - 1.29 0.59

0.597 rs2855459 (IVS4-61C>T) (COL11A2-07) CC 612 78.9 830 80.7 1.00 CT/TT 164 21.1 199 19.3 1.05 0.83 - 1.33 0.69

0.577

121 rs9277934 (Ex6+28G>A) (COL11A2-05) GG 324 41.8 466 45.0 1.00 AG 368 47.5 448 43.3 1.17 0.96 - 1.44 0.12 AA 83 10.7 121 11.7 0.97 0.70 - 1.34 0.86 0.562

rs925847 (IVS21+144G>A) (STAT4-36) GG 417 53.7 620 60.0 1.00 AG 301 38.8 338 32.7 1.32 1.07 - 1.61 0.01 AA 58 7.5 75 7.3 1.10 0.75 - 1.60 0.63 0.056 rs3024896 (IVS21-474G>A) (STAT4-24) GG 526 68.6 747 73.5 1.00 AG/AA 241 31.4 270 26.5 1.23 1.00 - 1.52 0.05

0.187 rs3024936 (IVS23+322C>G) (STAT4-28) CC 735 94.6 978 94.5 1.00 CG 42 5.4 57 5.5 1.06 0.70 - 1.61 0.80

0.798 rs6740131 (*4983T>C) (STAT4-45) TT 470 60.5 634 61.3 1.00 CT/CC 307 39.5 400 38.7 1.07 0.88 - 1.30 0.51

0.605 rs7558921 (*6341G>C) (STAT4-46) GG 593 76.3 811 78.4 1.00 CG/CC 184 23.7 224 21.6 1.16 0.93 - 1.46 0.20

0.198 rs6751855 ( -10041T>C) (STAT1-23) TT 277 36.7 424 41.9 1.00 CT 316 41.9 411 40.6 1.15 0.92 - 1.42 0.21 CC 161 21.4 178 17.6 1.36 1.04 - 1.78 0.02 0.022 rs1467199 ( -5772G>C) (STAT1-22) GG 496 64.0 694 67.3 1.00 CG/CC 279 36.0 337 32.7 1.15 0.94 - 1.41 0.17

0.124 rs13029532 (IVS2-1171G>T) (STAT1-14) GG 631 81.2 861 83.2 1.00 GT/TT 146 18.8 174 16.8 1.16 0.90 - 1.48 0.25

0.202 rs13029247 (IVS5-769A>G) (STAT1-24) AA 418 54.6 586 57.1 1.00 AG 287 37.5 367 35.7 1.12 0.91 - 1.37 0.28 GG 61 8.0 74 7.2 1.11 0.76 - 1.61 0.59 0.315 rs12693591 (IVS9-557G>T)

122 (STAT1-11) GG 592 76.3 791 76.4 1.00 GT/TT 184 23.7 244 23.6 0.99 0.79 - 1.24 0.95

0.922 rs7562024 (IVS11+433A>G) (STAT1-20) AA 290 37.4 431 41.8 1.00 AG 360 46.5 453 43.9 1.18 0.96 - 1.45 0.12 GG 125 16.1 148 14.3 1.21 0.91 - 1.62 0.19 0.107 rs13005843 (IVS11-326G>A) (STAT1-12) GG 676 87.1 898 87.0 1.00 AG/AA 100 12.9 134 13.0 1.01 0.76 - 1.34 0.96

0.990 rs2280232 (IVS14-380G>T) (STAT1-25) GG 428 55.1 612 59.2 1.00 GT 303 39.0 363 35.1 1.23 1.00 - 1.50 0.05 TT 46 5.9 59 5.7 1.08 0.71 - 1.64 0.71 0.120 rs1547550 (IVS18-330G>C) (STAT1-16) GG 332 42.7 471 45.6 1.00 CG 352 45.3 433 41.9 1.12 0.91 - 1.38 0.27 CC 93 12.0 129 12.5 1.00 0.74 - 1.37 0.98 0.619 rs13010343 (IVS21+137C>T) (STAT1-13) CC 596 76.7 812 78.5 1.00 CT/TT 181 23.3 223 21.5 1.12 0.89 - 1.41 0.33

0.460 rs2066804 (IVS21-8C>T) (STAT1-01) CC 484 62.5 649 62.9 1.00 CT 248 32.0 330 32.0 1.01 0.82 - 1.24 0.93 TT 42 5.4 53 5.1 0.99 0.64 - 1.53 0.97 0.976 rs17749316 (IVS22-96G>C) (STAT1-17) GG 656 84.4 848 82.0 1.00 CG/CC 121 15.6 186 18.0 0.87 0.68 - 1.13 0.30

0.189 rs16824035 (IVS24+1922G>A) (STAT1-27) GG 559 72.0 760 73.5 1.00 AG/AA 217 28.0 274 26.5 1.10 0.89 - 1.37 0.36

0.563 rs3771300 (IVS24-153A>C) (STA1-18) AA 215 27.7 312 30.2 1.00 AC 390 50.3 513 49.6 1.16 0.93 - 1.45 0.19 CC 171 22.0 209 20.2 1.17 0.89 - 1.53 0.27 0.232 rs12468579 (*3164T>C) (STAT1-26) TT 290 37.3 392 37.9 1.00 CT 363 46.7 482 46.7 1.05 0.85 - 1.30 0.64

123 CC 124 16.0 159 15.4 1.01 0.76 - 1.35 0.95 0.841 rs883844 (IVS17-1787T>C) (GLS-02) TT 349 45.6 496 48.3 1.00 CT 322 42.0 415 40.4 1.11 0.91 1.37 0.31 CC 95 12.4 115 11.2 1.12 0.82 1.54 0.47 0.313

rs1568400 (IVS1+1763T>C) (THRA-02) TT 489 63.0 647 62.6 1.00 CT/CC 287 37.0 387 37.4 0.99 0.82 - 1.21 0.95

0.892 rs7502514 (IVS8+58T>C) (THRAP4-11) TT 294 37.8 383 37.2 1.00 CT 368 47.4 497 48.3 0.97 0.79 - 1.20 0.79 CC 115 14.8 149 14.5 1.03 0.77 - 1.38 0.83 0.938 rs2302775 (IVS17+56A>G) (THRAP4-10) AA 492 63.4 651 63.0 1.00 AG/GG 284 36.6 383 37.0 0.96 0.79 - 1.17 0.71

0.856 rs9916158 (IVS18+180C>A) (THRAP4-09) CC 265 34.1 364 35.2 1.00 CA 395 50.9 510 49.3 1.04 0.84 - 1.29 0.70 AA 116 14.9 160 15.5 0.98 0.73 - 1.31 0.87 0.990 rs9913632 (IVS22+776A>G) (THRAP4-07) AA 686 88.3 926 89.6 1.00 AG/GG 91 11.7 108 10.4 1.13 0.83 - 1.53 0.43

0.580 rs2827 (Ex4-330C>T) (CSF3-20) CC 570 73.5 752 72.9 1.00 CT/TT 206 26.5 280 27.1 0.95 0.77 - 1.18 0.67

0.651

rs168405 (IVS1+2345C>A) (C19ORF22-01) CC 348 44.8 454 44.0 1.00 CA 351 45.2 454 44.0 1.06 0.87 - 1.30 0.55 AA 78 10.0 123 11.9 0.85 0.62 - 1.18 0.33 0.662 rs2306718 (IVS5+83C>T) (C19ORF22-02) CC 440 56.6 588 57.2 1.00 CT 298 38.4 361 35.1 1.11 0.91 - 1.36 0.31 TT 39 5.0 79 7.7 0.68 0.45 - 1.03 0.068 0.582 rs2965286 (*937G>A) (C19ORF22-03) GG 493 63.5 644 62.4 1.00 AG/AA 283 36.5 388 37.6 0.99 0.81 - 1.20 0.90

124

0.722 rs2930902 (IVS4-927T>C) (THRAP5-10) TT 520 66.9 684 66.5 1.00 CT/CC 257 33.1 344 33.5 1.02 0.83 - 1.25 0.87

0.994 rs1060442 (Ex5-49T>C) (THRAP5-06) TT 248 31.9 347 33.8 1.00 CT 387 49.8 500 48.6 1.06 0.86 - 1.32 0.57 CC 142 18.3 181 17.6 1.15 0.87 - 1.53 0.32 0.316 rs2965294 (IVS6+121G>C) (THRAP5-11) GG 272 35.1 327 31.7 1.00 CG 374 48.2 522 50.7 0.86 0.70 - 1.07 0.18 CC 130 16.8 181 17.6 0.87 0.66 - 1.16 0.35 0.254 rs1617214 (IVS8-257C>G) (THRAP5-08) CC 446 57.5 551 55.5 1.00 CG 282 36.4 381 38.4 0.91 0.75 - 1.12 0.39 GG 47 6.1 60 6.0 1.00 0.66 - 1.51 0.998 0.572 rs2241623 (IVS9+370G>A) (THRAP5-09) GG 352 45.3 470 45.7 1.00 AG 353 45.4 444 43.2 1.05 0.86 - 1.28 0.65 AA 72 9.3 114 11.1 0.82 0.59 - 1.15 0.25 0.540 rs13090 (Ex16+137G>A) (THRAP5-07) GG 441 56.8 569 55.1 1.00 AG 282 36.3 394 38.2 0.93 0.76 - 1.14 0.47 AA 53 6.8 69 6.7 0.99 0.67 - 1.46 0.96 0.627 rs17684161 (*2694G>A) (THRAP5-05) GG 597 76.8 768 74.4 1.00 AG/AA 180 23.2 264 25.6 0.86 0.69 - 1.08 0.20

0.113 rs1683564 ( -475C>A) (DF-03) CC 290 37.3 386 37.5 1.00 AC 388 49.9 490 47.6 1.06 0.86 - 1.30 0.60 AA 99 12.7 153 14.9 0.89 0.66 - 1.21 0.46 0.695

rs10459228 ( -43825G>A) (VDR-110) GG 565 72.7 732 70.7 1.00 AG/AA 212 27.3 303 29.3 0.90 0.73 - 1.12 0.34

0.167 rs4516035 ( -26929G>A) (VDR-19) GG 235 30.2 322 31.1 1.00 AG 403 51.9 527 51.0 1.09 0.87 - 1.36 0.45 AA 139 17.9 185 17.9 1.09 0.82 - 1.45 0.57

125 0.507 rs10783219 (IVS1-1747A>T) (VDR-105) AA 278 35.8 357 34.6 1.00 AT 381 49.0 512 49.6 0.94 0.77 - 1.17 0.60 TT 118 15.2 163 15.8 0.88 0.66 - 1.18 0.40 0.386 rs11168292 (IVS2+15G>C) (VDR-112) GG 331 42.6 467 45.2 1.00 CG 363 46.7 450 43.6 1.17 0.96 - 1.43 0.13 CC 83 10.7 116 11.2 1.08 0.78 - 1.50 0.627 0.280 rs10875695 (IVS2+583G>T) (VDR-106) GG 472 61.1 613 59.4 1.00 GT 261 33.8 355 34.4 0.95 0.77 - 1.17 0.62 TT 39 5.1 64 6.2 0.79 0.52 - 1.21 0.28 0.307 rs11574026 (IVS2+5374C>T) (VDR-25) CC 645 83.0 859 83.0 1.00 CT/TT 132 17.0 176 17.0 0.99 0.77 - 1.28 0.94

0.899 rs11574027 (IVS2+6247G>T) (VDR-107) GG 752 96.8 1018 98.4 1.00 GT 25 3.2 17 1.6 1.99 1.05 - 3.77 0.04

0.036 rs11168287 (IVS2+8206C>T) (VDR-77) CC 184 23.7 268 26.0 1.00 CT 426 54.8 526 51.0 1.25 0.99 - 1.58 0.06 TT 167 21.5 237 23.0 1.09 0.82 - 1.45 0.54 0.503 rs4760648 (IVS2-4108G>A) (VDR-42) GG 289 37.2 326 31.6 1.00 AG 374 48.1 507 49.2 0.81 0.65 - 1.00 0.05 AA 114 14.7 198 19.2 0.64 0.48 - 0.85 0.002 0.002 rs2853564 (IVS2-1930C>T) (VDR-39) CC 286 36.9 434 42.0 1.00 CT 378 48.7 456 44.1 1.29 1.05 - 1.59 0.02 TT 112 14.4 144 13.9 1.29 0.96 - 1.74 0.09 0.026 rs2254210 (IVS3-816C>T) (VDR-36) CC 324 41.7 483 46.8 1.00 CT 355 45.7 443 42.9 1.21 0.99 - 1.48 0.06 TT 98 12.6 107 10.4 1.40 1.02 - 1.93 0.04 0.015 rs10735810 (Ex4+4T>C) (VDR-04) TT 286 35.3 339 32.9 1.00 CT 376 46.4 493 47.8 0.89 0.72 - 1.10 0.29 CC 149 18.4 199 19.3 0.90 0.69 - 1.18 0.46 0.372 rs2239186 (IVS4+3341T>C)

126 (VDR-35) TT 447 57.5 581 56.2 1.00 CT 295 38.0 388 37.6 1.00 0.82 - 1.22 0.98 CC 35 4.5 64 6.2 0.71 0.46 - 1.11 0.13 0.340 rs3782905 (IVS4+6584C>G) (VDR-40) CC 373 48.0 514 49.7 1.00 CG 352 45.3 425 41.1 1.15 0.94 - 1.41 0.16 GG 52 6.7 95 9.2 0.71 0.49 - 1.04 0.08 0.702 rs3819545 (IVS4-6046T>C) (VDR-101) TT 275 35.4 337 32.6 1.00 CT 394 50.7 524 50.7 0.95 0.77 - 1.18 0.66 CC 108 13.9 172 16.7 0.80 0.60 - 1.08 0.14 0.176 rs2189480 (IVS4-4868C>A) (VDR-32) CC 294 37.8 390 37.7 1.00 AC 385 49.5 498 48.1 1.02 0.83 - 1.25 0.85 AA 98 12.6 147 14.2 0.88 0.65 - 1.20 0.41 0.570 rs886441 (IVS4-4004C>T) (VDR-92) CC 464 59.7 672 65.1 1.00 CT/TT 313 40.3 361 34.9 1.24 1.02 - 1.51 0.03

0.024 rs12717991 (IVS4-166G>A) (VDR-97) GG 293 37.8 336 32.5 1.00 AG 355 45.7 501 48.5 0.84 0.68 - 1.04 0.10 AA 128 16.5 196 19.0 0.79 0.60 - 1.05 0.10 0.068 rs2239179 (IVS5+1064A>G) (VDR-34) AA 286 36.8 381 36.8 1.00 AG 384 49.4 507 49.0 0.99 0.81 - 1.22 0.94 GG 107 13.8 146 14.1 0.92 0.68 - 1.24 0.56 0.626 rs2239180 (IVS5+2784G>C) (VDR-108) GG 636 81.9 854 82.7 1.00 CG/CC 141 18.1 179 17.3 1.06 0.83 - 1.37 0.63

0.772 rs2107301 (IVS5+3260C>T) (VDR-81) CC 384 49.7 481 46.6 1.00 CT 319 41.3 447 43.3 0.92 0.75 - 1.12 0.41 TT 69 8.9 105 10.2 0.83 0.59 - 1.17 0.29 0.232 rs2239182 (IVS5+3419A>G) (VDR-84) AA 225 29.0 299 28.9 1.00 AG 391 50.3 526 50.9 0.96 0.77 - 1.20 0.71 GG 161 20.7 209 20.2 0.96 0.73 - 1.27 0.79 0.764 rs2248098 (IVS5-1885T>C) (VDR-109) TT 234 30.1 285 27.5 1.00 CT 371 47.7 529 51.1 0.85 0.68 - 1.06 0.14

127 CC 172 22.1 221 21.4 0.92 0.70 - 1.21 0.55 0.470 rs11574077 (IVS5-1456A>G) (VDR-94) AA 701 90.2 938 90.6 1.00 AG/GG 76 9.8 97 9.4 0.94 0.68 - 1.30 0.71

0.710 rs1544410 (IVS10+283G>A) (VDR-08) GG 315 40.6 412 39.8 1.00 AG 358 46.1 481 46.5 0.96 0.78 - 1.18 0.68 AA 103 13.3 142 13.7 0.89 0.65 - 1.20 0.43 0.437 rs3847987 (Ex11+721G>T) (VDR-64) GG 620 79.8 828 80.1 1.00 GT/TT 157 20.2 206 19.9 1.02 0.80 - 1.29 0.89

0.991 rs12721364 (*7098T>C) (VDR-98) TT 536 69.0 694 67.1 1.00 CT/CC 241 31.0 341 32.9 0.92 0.75 - 1.12 0.40

0.601 Adjusted for sex, center, age (continuous), and smoking status (ever, never)

128 CHAPTER 5: Effect of VDR pathway genes on vitamin D intake and exposure and

Renal Cancer Risk

Karami S 1, Brennan P 2, Navratilova M 6, Mates D 8, Zaridze D 3, Janout V 4, Kollarova H 4,

Bencko V 5, Matveev V 3, Szeszenia-Dabrowska N 7, Holcatova I 5, Yeager M 9, Chanock

S9, Menashe I 1, Rothman N 1, Chow W-H1, Boffetta P 2, Moore LE 1

1Division of Cancer Epidemiology and Genetics, National Cancer Institute, NIH, DHHS,

Bethesda, MD, USA

2International Agency for Research on Cancer, Lyon, France

3Institute of Carcinogenesis, Cancer Research Centre, Moscow, Russia

4Department of Preventive Medicine, Faculty of Medicine, Palacky University, Olomouc,

Czech Republic

5Institute of Hygiene and Epidemiology, Charles University, First Faculty of Medicine,

Prague, Czech Republic

6Department of Cancer Epidemiology and Genetics, Masaryk Memorial Cancer Institute,

Brno, Czech Republic

7Department of Epidemiology, Institute of Occupational Medicine, Lodz, Poland

8Institue of Public Health, Bucharest, Romania

9Core Genotyping Facility at the Advanced Technology Center of the National Cancer

Institute, NIH, Department of Health and Human Services

Running Title: Effect of VDR and RXRA on RCC Risk and vitamin D

Key words: Renal cancer, vitamin D, calcium, UVB, sunlight, VDR , RXR ,

129 ABSTRACT:

Found in food and produced in the body after exposure to ultraviolet (UV) rays, vitamin

D is metabolized to its active form within the kidney, the major organ for vitamin D metabolism and activity and calcium homeostasis. Mediated by binding to the high- affinity vitamin D receptor ( VDR ), vitamin D forms a heterodimer complex with the retinoid-X-receptor ( RXR ) gene. Variation in both genes has been shown to modify renal cell carcinoma (RCC) risk. Therefore, we investigated whether single nucleotide polymorphisms (SNPs) across the VDR and RXR-alpha genes modified associations between RCC risk and intake frequency of vitamin D and calcium rich foods and occupational UV exposure. Ten SNPs were examined among 987 RCC case and 1,298 controls from a hospital-based case-control study conducted in Central and Eastern

Europe. In-person interviews were performed to collect demographic, dietary, and occupational data. A positive association was observed between increasing dietary intake frequency of yogurt (p-trend= 0.04) and total calcium (p-trend= 0.18), while inverse associations were observed with intake frequency of egg (p-trend= 0.02) and total vitamin D (p-trend= 0.09). RXRA SNPs, located 3 ′ downstream of the coding sequence, were shown to modify associations between vitamin D rich foods and RCC risk. RXRA

SNPs, located in introns one and four, were shown to modify associations between calcium rich foods and RCC risk. No associations with occupational UV exposure, VDR or RXRA SNPS, and RCC risk were found. However, increased risk (OR= 1.70; 95% CI=

1.14-2.53) was seen with increasing occupational UV exposure among males with the

TTG rs4760648, rs2853564, and rs2254210 VDR haplotype compared to subjects with the referent CCA haplotype. Results suggest that VDR and RXRA genes in the vitamin D

130 pathway may modify associations between RCC risk and vitamin D intake or exposure.

Replication studies are warranted to confirm these findings.

INTRODUCTION:

Epidemiological studies have recently suggested that vitamin D may be involved in renal

cell carcinoma (RCC) etiology [1-4] . Found in food (vitamin D 2 and D 3) and produced in the body after exposure to ultraviolet (UV) rays from the sun (vitamin D 3), both vitamin

D2 and D 3 are hydroxylated in the liver and subsequently in the kidney to form

1,25(OH) 2D3, the biologically active form of the vitamin [5] . Previously, we reported a significant inverse association between occupational UV exposure and RCC risk [6] ; however, the association between dietary vitamin D intake and renal cancer risk was not assessed. Given that dietary vitamin D intake accounts for approximately 10% of vitamin

D levels [5] , dietary intake of vitamin D and vitamin D rich foods may play an important role in determining RCC risk. Furthermore, since within the kidney calcium has also been shown to influence vitamin D levels, dietary intake of calcium may also be important in

RCC etiology [5] .

Although the exact anti-carcinogenic mechanism of vitamin D is not fully understood, vitamin D and its metabolites are thought to impede carcinogenesis by stimulating cell differentiation, inhibiting cell proliferation, inducing apoptosis, and suppressing invasiveness, angiogenesis, and metastasis [7-10] . Vitamin D is mediated by binding to the vitamin D receptor ( VDR ), transcriptional factors that are part of the nuclear hormone receptor family. Forming a heterodimer complex with the retinoid-X-receptor ( RXR )

131 gene, VDR can regulate the transcription of other genes involved in cell regulation,

growth, and immunity [9, 11, 12] . Most epidemiological studies have generally focused

on the VDR gene; however recently, we evaluated 139 single nucleotide polymorphisims

(SNPs) across eight genes of the vitamin D pathway and found a significant association

between RCC risk and certain VDR and RXR-alpha (RXRA ) genetic variants [13] .

Since the kidney is the major organ for vitamin D metabolism and activity and calcium

homeostasis [5, 13] it is important to investigate the role of calcium and vitamin D in

RCC etiology, particularly since there are so few studies that have investigated this

association. To follow-up on this hypothesis we investigated whether common genetic

variation in VDR and RXRA , the two genes significantly associated with RCC risk, modified associations between RCC risk and dietary intake frequency of vitamin D and calcium rich foods and occupational UV exposure. Analyses were conducted among cases and controls residing in Central and Eastern Europe, an area with one of the highest rates of RCC worldwide and an area where foods are not fortified with vitamin D [14] .

Additionally, haplotype analyses were also preformed to determine if associations were modified between RCC risk and vitamin D.

MATERIALS & METHODS:

Study Population

Details regarding this study population were previously described [15] . Briefly, from

1999 through 2003, a hospital-based case-control study of RCC was conducted in seven centers in four countries of Central and Eastern Europe (Moscow, Russia; Bucharest,

132 Romania; Lodz, Poland; and Prague, Olomouc, Ceske-Budejovice, and Brno, Czech

Republic). Cases, 20 to 88 years of age, included patients newly (within 3 months)

diagnosed and histologically confirmed with RCC (IDC-O-2 codes C64) who had lived in

the study areas for at least one year. Controls, frequency-matched to cases on age (+/- 3

years), sex, and place of residence, were selected from patients admitted to participating

hospitals for diagnoses unrelated to smoking or urological disorders with the exception of

benign prostatic hyperplasia. No single disease made up more than 20% of the control

group. A number of controls were also recruited in parallel for studies of lung and head

and neck cancers [16, 17] . All participants were of Caucasian descent. A total of 1,097 incident, histologically confirmed RCC cases and 1,476 controls were included in this study. Suitable quantity and quality of genomic DNA was obtained from a subset of 987

(90.0%) RCC cases and 1,298 (87.9%) controls. The response rates across study centers for study participation ranged from 90.0% to 98.6% for cases and from 90.3% to 96.1 % for controls. All subjects provided written informed consent. This study was approved by the institutional review boards of all participating centers.

Dietary Intake:

Interviewers were trained in each center to perform face-to-face interviews with cases and controls during hospitalization using standard questionnaires. The questionnaire covered basic demographic characteristics, family history of cancer, history of tobacco consumption, and dietary habits. A food frequency questionnaire was comprised of 23 food items, which the study investigators selected by consensus during the planning stage of the study and further validated during the pilot stage by asking participants to name

133 common food items not already specified. Frequency of dietary intake for each food group was assessed for each item as never, less than once per month, less than once per week, one to two times per week, three to five times per week, and daily. A standardized questionnaire was used in each of the study centers that was translated from a common

English version and then back-translated into English to ensure the validity of the initial translation. The questionnaire was repeated for two different time periods: 1) the year prior to interview, and 2) the year prior to political and market changes in 1989 (1991 in

Russia). Data for the year prior to interview and the year prior to political change were then extrapolated to represent lifetime average dietary intake by multiplying the score for

each time period by the number of years the participant was alive during the time period,

then summing the time period scores and dividing by the total age of the individual.

In this study, intake frequencies of vitamin D rich foods, which include liver, egg, and

fish were combined for each participant to create a new dietary exposure variable for total

vitamin D intake frequency. Since calcium has been reported to influence vitamin D

levels and risk of certain cancers [5] , intake frequencies of cheese, milk, and yogurt were

also combined for each participant to creat a new total calcium intake frequency exposure

variable. In this study, specific details regarding the type of fish, cheese, milk, etc.

consumed was not available; therefore, each food category was assumed to have equal

weights when combined to estimate total vitamin D and calcium lifetime average dietary

intake frequencies. All dietary intake frequency variables were assessed in tertiles based

on frequency of dietary intake among controls.

134 Occupational UV Exposure:

Lifetime occupational information for jobs held for at least 12 months duration was also

ascertained during interviews through the use of a general occupational questionnaire.

Data collected for each job included title, detailed tasks, and type of employer, as well as

year of beginning and ending employment. These data were used to create a job-

exposure matrix (JEM) which classified each study subjects’ estimated occupational

exposure to sunlight; the JEM is described in more detail elsewhere [6] . To assess the relationship between occupational sunlight exposure, RCC risk and vitamin D pathway genes, cumulative exposure (low exposure-unit-years) across all jobs [calculated as duration (years) x frequency midpoint x intensity of exposure for each job and summed over jobs] was assessed. A categorical exposure matrix was used to evaluate the association between RCC risk and cumulative occupational UV exposure based on tertiles of exposure levels among all male controls. Since the association between occupational UV exposure and RCC risk in this population was previously observed only among males [6] , evaluation of genetic variants, occupational UV exposure, and RCC risk was only explored among male participants.

Laboratory Procedures.

Genomic DNA was extracted from whole blood buffy coat using a standard phenol chloroform method at the National Cancer Institute (NCI) laboratory. Genotyping was conducted with a GoldenGate ® Oligo Pool All (OPA) assay by Illumina®

(www.illumina.com ). DNA samples from RCC cases and controls were coded and

randomized on polymerase chain reaction (PCR) plates for genotyping analyses. A

135 random 5% duplicate sample was selected and genotyped for quality control.

Based on our previous results reported in this population [13] , ten SNPS across the VDR

(rs11574027, rs4760648, rs2853564, rs2254210 and rs886441) and RXRA (rs100791, rs3118523, rs748964, rs3118536, and rs10776909) gene, which were significantly associated with RCC risk, were further evaluated in this study. Tag SNPs were originally selected to provide 80% to 90% coverage across the genomic regions of interest, while some non-synonymous SNPs were selected for their putative functional significance. All

SNPs selected for analysis had a variant allele frequency of at least 5% as reported by the

NCI SNP500Cancer Database ( http://snp500cancer.nci.nih.gov ) [18] and a validated assay at the NCI Core Genotyping Facility (CGF) ( http://cgf.nci.nih.gov/home.cfm ).

The concordance rate between duplicate DNA samples ranged from 93% to 100% and completion rates ranged from 98% to 100%. The genotype frequencies among controls did not differ from the expected Hardy-Weinberg equilibrium proportions (p >0.05).

Statistical Analysis:

The distributions of select characteristics and known RCC risk factors (sex, age, smoking habits, self-reported hypertension status, body mass index (BMI), family history of cancer, country of residence, and pertinent dietary variables) were compared between cases and controls using the Chi-square test. Odds ratios (ORs) and 95% confidence intervals (95% CIs) using unconditional logistic regression were calculated to estimate the associations between individual SNPs and RCC risk.

136 SNP variants were assessed using a dominant model: Wald chi-square test for the

presence or absence of the variant allele (0, 1). All regression models were adjusted for

age, sex, study center, smoking status (ever, never), BMI, and self-reported hypertensive

status. To check for lack of fit for each model, Hosmer-Lemeshow goodness of fit test

statistics were calculated.

Haplotype blocks identified previously [13] were analyzed in relation to RCC risk using

Haplostats (R version 2.4.0; http://www.r-project.org ) [19] adjusted for age, sex, study center, BMI, self-reported hypertensive status, and smoking habit. Associations between common haplotypes (>5% frequency) and RCC risk were evaluated by computing ORs and 95% CIs using the most common haplotype as the referent category. Interactions were tested comparing regression models with and without interaction terms using a likelihood ratio test (LRT).

All analyses were conducted in STATA 9.0 unless otherwise specified (STATA

Corporation, College Station, TX).

RESULTS:

Table 1 provides a description of study participants and known RCC risk factors. All participants and genotyped participants showed a similar distribution of characteristics.

Cases and controls were comparable in age; however, cases were more likely to be female and to have excess BMI (>30 kg/m 2), self-reported hypertension, and a first- degree relative with cancer. After adjustment for age, BMI, hypertension, study center,

137 and sex the association with smoking was no longer observed [20] .

The associations between intake frequency of vitamin D and calcium rich foods and RCC

risk among genotyped participants are shown in Table 2. Reduced RCC risk was

observed with increasing intake frequency of egg and total vitamin D (based on liver,

egg, and fish); however, after adjusting for covariates the association with total vitamin D

intake frequency was no longer statistically significant (p-trend= 0.09). Increased RCC

risk was observed with increasing intake frequency of yogurt and total calcium (based on

cheese, milk and yogurt); though, the association with total calcium intake frequency was

no longer statistically significant after adjusting for covariates (p-trend= 0.18). Adjusted

logistic regression analyses between RCC risk and each of the food frequency groups

examined were supported by the Hosmer-Lemeshow goodness of fit test.

Associations between RCC risk, SNPs across the vitamin D pathway, and vitamin D rich

foods that were associated with RCC risk are shown in Table 3. Inverse associations

were observed for main effects between RCC risk and subjects with the wild type

rs1007971 (p-trend= 0.05), rs3118523 (p-trend= 0.01), rs748964 (p-trend= 0.01),

rs3118536 (p-trend= 0.02), and rs10776909 (p-trend= 0.05) RXRA alleles. Yet, reduced

RCC risk was observed with increasing intake frequency of egg among subjects with > 1

variant rs1007971 (p-trend <0.001), rs3118523 (p-trend <0.001), rs748964 (p-trend

<0.001), rs3118536 (p-trend <0.001), and rs10776909 (p-trend= 0.004) alleles across the

RXRA gene. Significant interactions were only seen for RXRA SNPs located downstream,

3 of the coding sequence: rs1007971 (p-interaction= 0.01), rs3118523 (p-interaction=

138 0.01) and rs748964 (p-interaction= 0.01). Regression analyses between RCC risk, the

five RXRA gene variants, and each of the dietary vitamin D rich food frequency groups examined were supported by the Hosmer-Lemeshow goodness of fit test. No significant

association was observed for frequency of dietary total vitamin D (based on liver, egg,

and fish). Furthermore, no association was observed between any of the VDR SNPs and

dietary intake frequency of vitamin D food groups (data not shown).

Associations between RCC risk, SNPs across the vitamin D pathway, and calcium rich

foods are shown in Table 4. Increased RCC risk was observed with increasing intake

frequency of yogurt among participants with the wild type AA rs3118538 (p-trend= 0.01)

and TT rs10776909 (p-trend= 0.01) alleles across the RXRA gene; rs10776909 located in

intron 1, was also shown to significantly modify associations between dietary intake

frequency of yogurt and RCC risk (p-interaction= 0.04). Similarly, increased RCC risk

was observed with increased total calcium (based on cheese, yogurt and milk) intake

frequency for subjects with the wild type AA rs3118538 (p-trend= 0.05) and TT

rs10776909 (p-trend= 0.02) alleles; significant interactions were observed for both SNPs,

located in introns 1 and 4, within the RXRA gene. Regression analyses between RCC risk,

the five RXRA gene variants, and each of the dietary calcium rich food frequency groups

examined were supported by the Hosmer-Lemeshow goodness of fit test. No association

was observed between frequency of dietary calcium rich foods and other SNPs ( n= 8)

evaluated in this study.

Previously we observed a significant reduction (38%) in RCC risk only among male

139 participants occupational exposed to UV [6] . In this study, neither VDR nor RXRA SNPs modified the observed associations (data not shown); however, VDR haplotypes were shown to modify associations between RCC risk and cumulative occupational UV exposure among male participants (Table 5). Haplotype analyses revealed males with the

VDR T-T-G (rs2254210, rs2853564, and rs4760648) haplotype were at a significantly increased (OR= 1.27; 95%CI= 1.03-1.57) risk of renal cancer compared to males with the most common referent haplotype C-C-A. When haplotypes were stratified by

occupational UV category, only males in the highest UV exposure category with the

same VDR haplotype, T-T-G, were shown to have a significantly stronger increase in

RCC risk (OR= 1.70; 95%CI= 1.14-2.53) compared to males with the referent haploype,

C-C-A. No associations between VDR haplotpes, RCC risk and cumulative occupational

UV exposure were observed among male participants in the lowest or middle UV

category. A borderline significant interaction was also observed between VDR

haplotypes, occupational UV exposure, and RCC risk (p-interaction =0.06).

DISCUSSION:

In this study, reduced RCC risk was observed with increasing intake frequency of egg,

while increased RCC risk was observed with increasing intake frequency of yogurt.

Analyses by SNPs across the VDR and RXRA gene showed that only the RXRA gene modified associations between RCC risk and frequency of dietary vitamin D and calcium rich foods. Specifically, significant interactions between RCC risk, intake frequency of eggs and RXRA SNPs located downstr eam, 3 of the coding sequence (rs1007971, rs3118523, and rs748964) were observed. Significant interactions were also observed

140 between RCC risk, intake frequency of yogurt and total calcium, and intronic RXRA

SNPs (rs3118538 and rs10776909). Moreover, no association between SNPs in the VDR or RXRA gene were associated with occupational UV exposure and RCC risk among male participants; however, with increasing occupational UV exposure, male carriers of the T-T-G VDR haplotype had a significant increase in RCC risk compared to male subjects with the referent C-C-A haplotype.

Based on numerous studies that have established a link between diet and cancer risk or progression, the World Health Organization recently acknowledged that after smoking, diet may be the second most important contributor to the global burden of cancer [21] .

While, dietary vitamin D intake constitutes a small proportion of vitamin D requirements

[5] , scientific studies suggest that dietary intake of vitamin D may play an important role in determining cancer risk [2, 21, 22] . This is particularly important in areas where ultraviolet levels of exposure may vary by season. Similar to the results we present in this study, an Italian case-control study of 767 RCC cases and 1,534 controls observed a significant inverse association between dietary vitamin D intake, based on a 78-food frequency questionnaire, and RCC risk [2] . However, an earlier Canadian RCC case- control study observed no association for intake of individual foods rich in vitamin D (i.e. fish and eggs) or calcium (i.e. milk, diary, and cheese). Calcium supplement intake on the other hand, was shown to significantly reduce RCC risk with increasing number years of consumption [22] . Most epidemiological studies investigating the association between

RCC risk and vitamin D or calcium rich foods have been inconsistent and generally null

[22-27] .

141

While no reports to date have investigated the association between RCC risk, dietary intake of vitamin D or calcium rich foods, and vitamin D pathway genes, inconsistent results have been reported for VDR SNPs and other cancers. For example, in a large

colorectal U.S. case-control study, increased dietary vitamin D and calcium intake was

associated with reduced colon cancer risk among subjects with the polyA SS and BsmI BB

genotypes compared to those with the wild type LL and bb VDR alleles [28] . Reduced

rectal cancer risk was also observed with increasing consumption of calcium intake

among carriers of the polyA SS and BsmI BB genotypes (p-interaction= 0.01) [28] . The

association between vitamin D and calcium intake was further evaluated in a small U.S.

case-control study of colorectal adenoma, which found increased cancer risk associated

with increased dietary vitamin D intake among participants with the BsmI BB allele compared to subjects with the wild type bb genotype (p-interaction= 0.09) [29] . In a third case-control study recently conducted in Britain, where milk is not fortified with vitamin D, reduced colorectal cancer risk was observed with increasing consumption of dairy products (p-interaction= 0.02) and to a lesser degree increasing consumption of milk (p-interaction= 0.18) among participants with the combined VDR variant ApaI aA/AA alleles compared to participants with the wild type aa allele [30] . Yet, no association between VDR polymorphisms, TaqI and BsmI , cancer risk and vitamin D intake (from diet or supplements) were observed in a U.S. population-based lymphoma case-control study [31] . Similarly, no associations between FokI , TaqI , BsmI , ApaI or polyA polymorphisms, cancer risk, and vitamin D intake were reported in a large U.S. nested breast cancer study [32] .

142

Inconsistent results have also been reported for epidemiological studies investigating

whether polymorphisms in vitamin D pathway genes modify the relationship between

cancer risk and UV exposure; no studies to date however, have investigated this

relationship with renal cancer risk. Similar to the results reported in our study, an

association between UV exposure and haplotypes comprised of the rs2254210 VDR SNP

were reported in a recent prostate case-control study conducted in Britain [33] . Four VDR

SNPs in two haplotype block regions (region 1: rs7139166, rs4516035; region 2:

rs2853559, rs2254210) were examined in relation to prostate cancer risk and UV

exposure; both haplotype regions were shown to significantly increase prostate cancer

risk among men with low UV exposure [33] . Conversely, inverse associations were

reported in this population when haplotypes comprising of five other VDR (rs11568820,

rs4516035, rs10735810, rs2107301, rs731236) SNPs were examined in relation to UV

exposure and prostate cancer risk [34] . Likewise, in a U.S. population-based case-control

study of prostate cancer, reduced cancer risk was reported among participants with the

Fok1 Ff/FF (OR=0.46; 95%CI=0.23-0.92) and Taq1 tt (OR=0.48; 95%CI=0.24-0.9581)

alleles in the highest UV exposure group [35] . A stronger inverse association was also

observed among subjects in the highest UV exposure group who were carriers of the FokI

and TaqI FFtt haplotype [35] .

Our study adds to the limited epidemiological evidence that vitamin D (from dietary intake or UV exposure) and vitamin D pathway way genes, in particular VDR and RXRA , play a role in RCC etiology. The biological activity of vitamin D is mediated by the high-

143 affinity receptor, VDR which acts as a ligand-activated transcription factor that forms a heterodimer with RXR [9, 11, 12, 36] ; this VDR-RXR heterodimer complex is directed to the vitamin D-responsive element in the promoter region of 1,25-regulated genes [9, 11,

12, 36] . Different polymorphisms in the VDR gene have been speculated to result in

variation of VDR expression and therefore to further result in changes to circulating

levels of active vitamin D [37] . Furthermore, RXRA may play a critical role in vitamin D

activity, particularly from diet, since this gene has been shown to regulate cholesterol

[38] , which is abundant in liver, eggs, cheese, milk, and yogurt, the food groups used to

categorize our vitamin D and calcium intake variables in this study.

Strengths of our study include the use of HapMap to tag genes of interest using high (80 -

90%) genomic coverage both 5 and 3 of the target genes. In this study we also observed

high participation rates, used newly diagnosed cases, included only histologically

confirmed cancers, and collected biologic materials from a high proportion of subjects.

The large sample size of this study provided sufficient statistical power to detect

relatively small associations; however, the power to detect gene-environment interactions

was limited and results need to be interpreted cautiously.

Inherent limitations of our study include: (1) possible misclassification of vitamin D and

calcium intake from the use of a 23-item lifetime food frequency questionnaire, which

assumed equal weights for all food categories (2) the inability to control for potential

confounders (e.g. dietary retinol and phosphorous), (3) lack of data regarding dietary

vitamin D or calcium supplements, which may have confounded results, and (4) limited

144 statistical power. Limitations stemming from the estimation of occupational UV exposure [6] , such as non-differential inaccurate or incomplete recall of all occupational histories and non-differential exposure misclassification, may also have biased results.

Additionally, while hospital-based case-control studies have potential limitations due to possible differences with population controls, these studies can improve response rates for the intense collection of biological specimens and therefore reduce the chances of bias in the assessment of gene-environment interactions [39] . Also, we did not employ a direct marker for current vitamin D status among these study subjects. While there is general agreement that the serum 25(OH) vitamin D level is the best indicator of vitamin

D status, the biochemical marker has a short half-life. Therefore, a single measurement of 25(OH) vitamin D may not reflect long-term vitamin D status that would be important in studies of cancer. Moreover, vitamin D status post diagnosis among cases may reflect disease rather than historical vitamin D levels of participants when they were “healthy.”

Lastly, while global p-values were assessed previously on these genes [13] , we cannot completely rule out the possibility of false positive findings in this study due to multiple comparisons.

In conclusion, among participants in the Central and Eastern European Renal Cell

Carcinoma Study VDR and RXRA genes of the vitamin D pathway were shown to modify associations between RCC risk and intake frequency of vitamin D and calcium rich foods, and occupational UV exposure. Additional studies with large sample sizes and similarly thorough genomic coverage of VDR pathway genes are needed to confirm results and to provide insight into the role of vitamin D in RCC etiology.

145

REFERENCES:

1. Ikuyama T, Hamasaki T, Inatomi H, Katoh T, Muratani T, Matsumoto T. Association of vitamin D receptor gene polymorphism with renal cell carcinoma in Japanese. Endocr

J 2002 Aug;49(4):433-8.

2. Bosetti C, Scotti L, Maso LD, et al. Micronutrients and the risk of renal cell cancer: a case-control study from Italy. Int J Cancer 2007 Feb 15;120(4):892-6.

3. Obara W, Suzuki Y, Kato K, Tanji S, Konda R, Fujioka T. Vitamin D receptor gene polymorphisms are associated with increased risk and progression of renal cell carcinoma in a Japanese population. Int J Urol 2007 Jun;14(6):483-7.

4. Karami S, Brennan P, Hung RJ, et al. Vitamin D receptor polymorphisms and renal cancer risk in Central and Eastern Europe. J Toxicol Environ Health A 2008;71(6):367-

72.

5. Holick MF. Vitamin D: Its role in cancer prevention and treatment. Progress in

Biophys Mol Biol 2006;92(1):49-59.

6. Karami S, Boffetta P, Stewart P, et al. Occupational Sunlight Exposure and Risk of

Renal Cell Carcinoma. (In Press) Cancer.

7. Trump DL, Hershberger PA, Bernardi RJ, Ahmed S, Muindi J, Fakih M, et al. Anti-

146 tumor activity of calcitriol: pre-clinical and clinical studies. J Steroid Biochem Mol Biol

2004;89-90(1-5):519-26.

8. Ordonez-Moran P, Larriba MJ, Pendas-Franco N, Aguilera O, Gonzalez-Sancho JM,

Munoz A. Vitamin D and cancer: an update of in vitro and in vivo data. Front Biosci

2005;10:2723-49.

9. Valdivielso JM, Fernandez E. Vitamin D receptor polymorphisms and diseases. Clin

Chim Acta 2006;371(1-2):1-12.

10. Hino O, Kobayashi T, Momose S, Kikuchi Y, Adachi H, Okimoto K. Renal carcinogenesis: genotype, phenotype and dramatype. Cancer Sci 2003;94(2):142-7.

11. Walters MR. Newly Identified actions of the vitamin D endocrine system. Endocr

Rev 1992;13:719-764.

12. Thibault F, Cancel-Tassin G, Cussenot O. Low penetrance genetic susceptibility to kidney cancer. Br J Urul Int 2006;98:735-738.

13. Karami S, Brennan P, Stewart P, et al. Analysis of SNPs and Haplotypes in VDR

Pathway Genes and Renal Cancer Risk. (In Press) PLoS One.

14. International Agency for Research on Cancer. GLOBACAN. Available at:

147 http://www-dep.iarc.fr/ Last Accessed on April 2, 2008.

15. Hung RJ, Moore L, Boffetta P, et al. Family history and the risk of kidney cancer: a multicenter case-control study in Central Europe. Cancer Epidemiol Biomarkers Prev

2007;16:1287-1290.

16. Norman AW. Sunlight, season, skin pigmentation, vitamin D, and 25-hydroxyvitamin

D: integral components of the vitamin D endocrine system. Am J Clin Nutr

1998;67:1108-10.

17. Matsuoka LY, Wortsman J, Haddad JG, Kolm P, Hollis BW. Racial pigmentation and the cutaneous synthesis of vitamin D. Arch Dermatol 1991;127:536-8.

18. Packer BR, Yeager M, Burdett L, et al. SNP500Cancer: a public resource for sequence validation, assay development, and frequency analysis for genetic variation in candidate genes. Nucleic Acids Res 2006;34(Database issue):D617-21.

19. R Project for Statistical Computing. Available at: http://www.r-project.org/ Last

Accessed on March 14, 2008.

20. Hunt JD, van der Hel OL, McMillan GP, Boffetta P, Brennan P. Renal cell carcinoma in relation to cigarette smoking: meta-analysis of 24 studies. Int J Cancer 2005;114:101-

8.

148

21. Thorne J, Campbell MJ. The vitamin D receptor in cancer. Proc Nutr Soc

2008;67(2):115-27.

22. Hu J, Mao Y, White K; Canadian Cancer Registries Epidemiology Research Group.

Diet and vitamin or mineral supplements and risk of renal cell carcinoma in Canada.

Cancer Causes Control 2003 Oct;14(8):705-14.

23. Rashidkhani B, Akesson A, Lindblad P, Wolk A. Major dietary patterns and risk of renal cell carcinoma in a prospective cohort of Swedish women. J Nutr

2005;135(7):1757-62.

24. Hu J, La Vecchia C, DesMeules M, Negri E, Mery L; Canadian Cancer Registries

Epidemiology Research Group. Meat and fish consumption and cancer in Canada. Nutr

Cancer 2008;60(3):313-24.

25. Bravi F, Bosetti C, Scotti L, et al. Food groups and renal cell carcinoma: a case- control study from Italy. Int J Cancer 2007;120(3):681-5.

26. Wolk A, Larsson SC, Johansson JE, Ekman P. Long-term fatty fish consumption and renal cell carcinoma incidence in women. JAMA 2006;296(11):1371-6.

27. Hsu CC, Chow WH, Boffetta P, et al. Dietary risk factors of renal cell carcinoma in

149 eastern and central Europe. Am J Epidemiol 2007;166(1):62-70.

28. Slattery ML, Neuhausen SL, Hoffman M, et al. Dietary calcium, vitamin D, VDR genotypes and colorectal cancer. Int J Cancer 2004, 111(5):750-6.

29. Kim HS, Newcomb PA, Ulrich CM, et al. Vitamin D receptor polymorphism and the risk of colorectal adenomas: evidence of interaction with dietary vitamin D and calcium.

Cancer Epidemiol Biomarkers Prev 2001;10(8):869-74.

30. Hubner RA, Muir KR, Liu JF, et al. Dairy products, polymorphisms in the vitamin D receptor gene and colorectal adenoma recurrence. Int J Cancer 2008 Aug 1;123(3):586-

93.

31. Purdue MP, Hartge P, Davis S, et al. Sun exposure, vitamin D receptor gene polymorphisms and risk of non-Hodgkin lymphoma. Cancer Causes Control 2007;18(9):

989-99.

32. McCullough ML, Stevens VL, Diver WR, et al. Vitamin D pathway gene polymorphisms, diet, and risk of postmenopausal breast cancer: a nested case-control study. Breast Cancer Res 2007;9(1):R9.

33. Rukin NJ, Luscombe C, Moon S, et al. Prostate cancer susceptibility is mediated by interactions between exposure to ultraviolet radiation and polymorphisms in the 5'

150 haplotype block of the vitamin D receptor gene. Cancer Lett 2007; 247(2): 328-35.

34. Moon S, Holley S, Bodiwala D, et al. Associations between G/A1229, A/G3944,

T/C30875, C/T48200 and C/T65013 genotypes and haplotypes in the vitamin D receptor gene, ultraviolet radiation and susceptibility to prostate cancer. Ann Hum Genet 2006

Mar;70(Pt 2):226-36.

35. John EM, Schwartz GG, Koo J, Van Den Berg D, Ingles SA. Sun exposure, vitamin

D receptor gene polymorphisms, and risk of advanced prostate cancer. Cancer Res 2005;

65(12): 5470-9.

36. Brown AJ, Dusso A, Slatopolsky E. Vitamin D. Am J Physiol Renal Physiol

1999;227(2 Pt2):F157-75.

37. van den Berg H. Bioavailability of vitamin D. Eur J Clin Nutr 1997;51 Suppl 1: S76-

9.

38. Hegele RA, Cao H. Single nucleotide polymorphisms of RXRA encoding retinoid X receptor alpha. J Hum Genet 2001;46(7):423-5.

39. Wacholder S, Chatterjee N, Hartge P. Joint effect of genes and environment distorted by selection biases: implications for hospital-based case-control studies. Cancer

Epidemiol Biomarkers Prev 2002;11:885-889.

151 Table 1. General characteristics of participants in the Central and Eastern European Renal Cell Carcinoma Study All Participants Genotyped Participants Variables Cases Controls Cases Controls N%N% ^ p-value N%N% ^ p-value Participants 1,097100.0 1,476100.0 987 100.0 1,298 100.0

Sex Males 64859.1 95264.5 58959.7 83864.6 Females 449 40.9 524 35.5 0.01 398 40.3 460 35.4 0.02

Age at Interview <45 867.8 1228.3 767.7 1088.3 45-54 27825.3 37925.7 25325.6 33325.7 55-64 33530.5 46031.2 30330.7 40531.2 65-74 35332.2 45230.6 31331.7 39730.6 75+ 45 4.1 63 4.3 0.61 424.3 55 4.2 0.58

Mean Age (std) 59.6 years (10.3) 59.3 years (10.3) 0.68 59.5 years (10.3) 59.3 years (10.3) 0.68

Center Romania-Bucharest 95 8.7 160 10.8 91 9.2 132 10.2 Poland-Lodz 99 8.7 19813.4 81 8.7 197 15.2 Russia-Moscow 317 28.9 463 31.4 288 29.2 368 28.4 *Czech Republic 586 53.4 655 44.4 <0.001 527 53.4 601 46.3 <0.001

BMI at Interview <25 32729.8 53236.2 28829.2 45835.3 25-29.9 476 43.4 62042.1 42943.5 556 42.8 30+ 293 26.7 319 21.7 <0.001 270 27.4 284 21.9 <0.001

Tobacco Status Never 510 46.6 599 40.7 454 46.1 528 40.7 Ever 584 53.4 874 59.3 0.003 53053.9 768 59.3 0.01

Hypertension No 60054.7 90661.4 53954.7 80061.7 Yes 496 45.3 569 38.6 0.001 44745.3 497 38.3 0.001

Familial History of Cancer No 1st degree relative with cancer 733 66.8 1074 72.8 654 66.3 932 71.8 1st degree relative with cancer 364 33.2 402 27.2 0.001 333 33.7 366 28.2 0.004

* Brno, Olomouc, Prague, Ceske-Budejovice ^ Crude p-values from Chi-square test

152 Table 2. RCC risk by frequency of dietary intake of vitamin D and calcium rich foods among genotyped participants

Cases Controls Unadjusted Adjusted Intake Frequency Variable N % N % OR UCI - LCI p-trend OR UCI - LCI p-trend

Liver Low (<33%) 318 41.0 411 39.7 1.00 1.00 Medium (33-66%) 315 40.6 380 36.7 1.07 0.87 - 1.32 1.11 0.89 - 1.38 High (>66%) 143 18.4 244 23.6 0.76 0.59 - 0.98 0.82 0.63 - 1.07 0.08 0.28 Egg Low (<33%) 274 35.3 345 33.3 1.00 1.00 Medium (33-66%) 392 50.5 454 43.9 1.09 0.88 - 1.34 1.11 0.90 - 1.38 High (>66%) 110 14.2 236 22.8 0.59 0.45 - 0.77 0.64 0.48 - 0.86 0.002 0.02 Fresh Fish Low (<33%) 260 33.5 345 33.3 1.00 1.00 Medium (33-66%) 274 35.3 344 33.2 1.06 0.84 - 1.32 1.06 0.84 - 1.34 High (>66%) 242 31.2 346 33.4 0.93 0.74 - 1.17 0.94 0.74 - 1.19 0.53 0.60 Salt Fish Low (<33%) 259 33.4 346 33.4 1.00 1.00 Medium (33-66%) 334 43.0 419 40.5 1.06 0.86 - 1.32 1.02 0.82 - 1.27 High (>66%) 183 23.6 270 26.1 0.91 0.71 - 1.16 0.91 0.71 - 1.18 0.50 0.53 Total Fish Low (<33%) 290 37.4 380 36.7 1.00 1.00 Medium (33-66%) 236 30.4 309 29.9 1.00 0.80 - 1.26 0.98 0.77 - 1.23 High (>66%) 250 32.2 346 33.4 0.95 0.76 - 1.18 0.95 0.75 - 1.19 0.64 0.65 ^Total Vitamin D Low (<33%) 284 36.6 345 33.3 1.00 1.00 Medium (33-66%) 279 36.0 353 34.1 0.96 0.77 - 1.20 0.95 0.76 - 1.20 High (>66%) 213 27.4 337 32.6 0.77 0.61 - 0.97 0.81 0.63 - 1.03 0.03 0.09 Cheese Low (<33%) 193 24.9 286 27.6 1.00 1.00 Medium (33-66%) 302 38.9 337 32.6 1.33 1.04 - 1.69 1.31 1.03 - 1.68 High (>66%) 281 36.2 412 39.8 1.01 0.80 - 1.28 0.99 0.77 - 1.26 0.83 0.71 Yogurt Low (<33%) 202 26.0 347 33.5 1.00 1.00 Medium (33-66%) 278 35.8 313 30.2 1.53 1.20 - 1.93 1.48 1.16 - 1.88 High (>66%) 296 38.1 375 36.2 1.36 1.08 - 1.71 1.31 1.03 - 1.67 0.02 0.04 Milk Low (<33%) 179 23.1 264 25.5 1.00 1.00 Medium (33-66%) 292 37.6 389 37.6 1.11 0.87 - 1.41 1.10 0.86 - 1.42 High (>66%) 305 39.3 382 36.9 1.18 0.92 - 1.50 1.14 0.89 - 1.46 0.19 0.31 *Total Calcium Low (<33%) 226 29.1 345 33.3 1.00 1.00 Medium (33-66%) 274 35.3 355 34.3 1.18 0.94 - 1.48 1.16 0.91 - 1.48 High (>66%) 276 35.6 335 32.4 1.26 1.00 - 1.59 1.20 0.92 - 1.58 0.05 0.18 Main effects adjusted for sex, age, study center, self-reported hypertensive status, BMI, and smoking status (never, ever) Dietary variable categorizing in tertiles based on controls ^Total Vitamin D (L, M, H) variable based on liver, egg, and salt and fresh water fish consumption ^ Vitamin D intake risk adjusted for sex, age, center, self-reported hypertensive status, BMI, smoking status, and occupational UV exposure *Total calcium variable based on cheese, yogurt, and milk consumption

153 0.59 0.11 0.72 0.82 0.99 11) 17) .08) .25) .22) WildType 83 (0.52-1.35) 193/263 0.97 (0.74-1.27) 1 Variant 1 > 7 1.13 (0.75-1.68) 165/246 0.87 (0.65-1.16) 4 1.04 (0.70-1.55) 164/239 0.90 (0.67-1.20) 8 0.84 (0.58-1.21) 155/205 1.00 (0.74-1.36) 8 0.96 (0.64-1.44) 168/235 0.91 (0.68-1.21) Dietary Intake Frequency of Vitamin D Foods Rich 1 77/68 1.00 207/277 1.00 Cases/Controls Cases/Controls 89/97 1.00 195/248 1.00 %) 110/112 1.00 174/233 1.00 %) 113/111 1.00 171/234 1.00 %) 110/119 1.00 174/226 1.00 p-trend 0.58 0.14 DietaryIntake Frequency of Total Vitamin D p-trend 0.07 0.53 DietaryIntake Frequency of Total Vitamin D p-trend 0.21 0.21 DietaryIntake Frequency of Total Vitamin D p-trend 0.06 0.45 DietaryIntake Frequency of Total Vitamin D p-trend 0.10 0.33 DietaryIntake Frequency of Total Vitamin D T TotalVitamin D N/N OR (LCI-UCI) N/N OR (LCI-UCI) *LRT 0.01 0.01 0.01 0.37 0.21 )(>66%) High 71/102 0.86 (0.54-1.38) 142/235 0.78 (0.57-1 )(>66%) High 65/113 0.63 (0.39-1.00) 148/224 0.90 (0.66-1 )(>66%) High 51/66 0.69 (0.39-1.24) 162/271 0.82 (0.61-1. )(>66%) High 68/105 0.65 (0.41-1.02) 145/232 0.88 (0.63-1 )(>66%) High 59/95 0.65 (0.40-1.06) 154/242 0.86 (0.63-1. 0.36 0.82 0.73 0.37 0.49 WildType foods of renal and cell risk carcinoma 0.004 <0.001 <0.001 <0.001 <0.001 6-1.46) 291/367 1.26 (0.97-1.64)(33-66%) Medium 86/90 0. 64-1.40) 253/331 1.34 (1.01-1.80)(33-66%) Medium 114/10 71-1.54) 251/328 1.27 (0.95-1.69)(33-66%) Medium 115/11 69-1.46) 242/291 1.24 (0.93-1.67)(33-66%) Medium 124/14 59-1.37) 263/320 1.25 (0.95-1.65)(33-66%) Medium 111/11 DietaryIntake Frequency of Egg s and occupationalexposure UV 1 Variant 1 > Cases/Controls Cases/Controls High (>66%) High 31/90 0.29 (0.17-0.49) 79/146 0.98 (0.68-1.42 DietaryIntake Frequency of Egg High (>66%) High 31/98 0.32 (0.19-0.53) 79/138 0.99 (0.69-1.43 High (>66%) High 19/62 0.21 (0.11-0.42) 91/174 0.86 (0.62-1.20 DietaryIntake Frequency of Egg High (>66%) High 37/91 0.44 (0.26-0.72) 73/145 0.78 (0.54-1.13 DietaryIntake Frequency of Egg High (>66%) High 31/81 0.35 (0.20-0.60) 79/155 0.81 (0.57-1.16 DietaryIntake Frequency of Egg self-reported hypertesive status, smoking status, status, smoking status, self-reported hypertesive into tertiles based on dietary intake among control into tertiles basedon intake dietaryamong f-reported hypertesive status, and smoking status and smoking f-reportedstatus, hypertesive nd smoking status smoking nd SNPs on dietary rich on D frequency SNPs of intake vitamin Main Effect Main 482/714 0.72 (0.58-0.88)(33-66%) Medium 141/130 0.95 (0. 484/697 0.79 (0.64-0.96)(33-66%) Medium 143/133 1.05 (0. 214/224 1.00(<33%) Low 92/68 1.00 180/270 1.00(<33%) Low 563/811 0.70 (0.56-0.88)(33-66%) Medium 103/94 0.91 (0.5 475/663 0.86 (0.70-1.04)(33-66%) Medium 152/170 1.00 (0. 295/321 1.00(<33%) Low 123/101 1.00 149/237 1.00(<33 Low 259/310 1.00(<33%) Low 97/88 1.00 175/250 1.00(<33%) Low 518/725 0.82 (0.67-1.00)(33-66%) Medium 131/141 0.90 (0. 293/338 1.00(<33%) Low 119/107 1.00 153/231 1.00(<33 Low 302/372 1.00(<33%) Low 113/111 1.00 159/227 1.00(<33 Low RXRA Cases/Controls AA ^p-trend 0.01 p-trend GG CG/GG ^p-trend 0.05 p-trend CC ^p-trend 0.01 p-trend TT ^p-trend 0.05 p-trend AG/GG AC/CC AA ^p-trend 0.02 p-trend CG/CC CT/CC -51) -28) -31) -12) -15) RXRA RXRA RXRA RXRA RXRA Dietary effect adjustedDietary BMI, foreffect center, age, study Table Effect 3.of rs748964( (*5628C>G) rs3118523( (*7060A>G) DietaryIntake Frequency of Egg All values adjusted values center, study All BMI,for age, sel rs3118536( (IVS4-542A>C) SNPrs1007971( (*7453G>C) N/N OR (LCI-UCI) Egg N/N OR (LCI-UCI) N/N OR (LCI-UCI) *LR rs10776909( (IVS1-4732T>C) All dietary intake frequency variables frequency categorized intake dietary All Main effects adjusted for center, Main effects age, study sex, a ^ ^ p-trend based on model additive Likelihood Ratio Test * 1

154 0.02 0.04 0.01 0.02 (UCI-LCI) *LRT trols Wild Type -15 (rs10776909) -15 0.58 0.16 RXRA 1 1 Variant > oods and risk ofrenaloods andcellrisk carcinoma (0.86(0.70-1.04)) (0.82(0.67-1.00)) 0.03 0.09 CT/CC AC/CC g status 42) 114/157 0.94 (0.58-1.54)93) 182/218 1.71 (1.18-2.49) 99/146 0.73 (0.47-1.14) 177/189 1.49 (1.06-2.11) 0.01 0.05 -2.64) 101/105 1.29 (0.81-2.07)-1.59) 169/198 1.84 (1.29-2.63) 112/117 1.10 (0.73-1.64) 162/238 1.21 (0.89-1.66) WildType (0.62(0.38-1.00)), p-trend=0.05; (0.58(0.34-0.99)), p-trend=0.02; CC CC -31 (rs3118538) -31 intake tertiles based on dietary intake among controls tertiles based onamong intake dietary RXRA nd smoking status nd smoking BMI, and smokin status, self-reported hypertensive (0.69(0.42-1.13)), (0.68(0.39-1.18)), CT AC 1 1 Variant SNPs frequency on dietary ofcalciumrichSNPs f intake > Cases/Controls Cases/Controls Cases/Controls Cases/Con N/N OR (UCI-LCI) N/N OR (UCI-LCI) *LRT N/N OR (UCI-LCI) N/N OR RXRA Table 4. Effect ofTable Effect 4. Dietary Intake Frequency ofYogurt Intake Frequency Dietary (<33%) Low Medium(33-66%) (>66%)High 85/91p-trend 77/91 1.23 (0.73-2.08) 1.00 ofTotal Calcium Intake Frequency Dietary 97/128 185/212(<33%) 0.87 Low (0.51-1.50) 1.88Medium(33-66%) (1.34 (>66%)High 199/247 94/94p-trend 80/97 1.69 (1.18-2. 1.22 (0.78-1.89) 133/266 1.00 85/119 1.00 180/261 0.76 (0.47-1.22) 1.18 (0.88 191/216 0.41 1.39 (1.00-1. 146/248 1.00 87/110 1.00 0.25 91/109 123/247 1.00 1.00 135/236 1.00 Main effects adjusted center, effects Main for age,a study sex, Ratio Test *Likelihood rs10776909 (IVS1-4732T>C) effects: main Caclium rich foods and milk yogurt, include cheese, Caclium rs3118538 effects: (IVS4-542A>C) main Dietary effect adjusted forcenter, effect Dietary study age, sex, variables categorized frequency into intake Dietary

155 0.06 LRT 0.13 0.01 8) 0.52 >16.10 (low-exposure-unityears) Cases Controls osure groups osure .25 33.5 25.6 1.70 (1.14-2.53) 0.21 High Exposure Cumulative I) p-value % % OR (LCI-UCI) p-value >6.00-16.10 (low-exposure-unityears) .5 7.7 0.76 (0.42-1.39) 0.38 6.6 5.3 1.83 (0.88-3.81) 0.11 Cases Controls .49 22.2 17.2 1.46 (0.99-2.16) 0.06 19.2 20.3 1.15 (0.75-1.7 status, and dietary vitamin D frequency intake D frequency status, vitamin and dietary 0.82 6.00 < er stratificationer cumulativeby occupational exp UV (low-exposure-unityears) Low Cumulative Exposure LowCumulative Exposure Cumulative Medium 29.7 31.0 0.93 (0.63-1.37) 0.72 31.7 28.7 1.23 (0.87-1.74) 0 Cases Controls (0.64(0.48-0.85)), 0.002 p-trend= (0.77(0.58-1.04)),p-trend= 0.03 (0.71(0.52-0.98)), p-trend= 0.02 AA CC CC center, BMI, self-reported hypertensive status, smoking BMI,center, status, smoking self-reported hypertensive 0.31 0.02 ported hypertensive status, and smoking status status, and portedsmoking hypertensive (0.81(0.65-1.00)), (1.00(0.75-1.33)), (0.86(0,63-1.18)), AG CT *MainEffect CT haplotypes and RCC risk among malesbeforeamong haplotypesaft and risk and RCC % % OR (LCI-UCI) p-value % % OR (LCI-UCI) p-value % % OR (LCI-UC 5.7 6.6 0.97 (0.67-1.42) 0.88 5.3 7.1 0.74 (0.36-1.50) 0.40 5 19.5 18.5 1.19 (0.94-1.51) 0.16 16.9 18.0 0.84 (0.52-1.37) 0 39.9 43.831.7 28.3 1.00 1.27 (1.03-1.57) 44.2 41.5 1.00 38.5 42.7 1.00 37.4 46.2 1.00 Cases Controls VDR Table5. Haplotypes Global P Global C-C-G VDR rs2254210-5' rs2853564, 3'-rs4760648, C-C-A T-T-G C-T-G rs2853564(IVS2-1930C>T) effects: main rs2245210(IVS3-816C>T) effects: main *Main effects adjusted BMI,effects *Main center, for age, study self-re occupational UV adjustedeffect Cumulative forage, study rs4760648 (IVS2-4108G>A) effects: main

156 Chapter 6: Past, Present and Future of Vitamin D and Cancer Risk Studies

6.1 Dissertation Study Results:

This study sought to investigate whether increased vitamin D exposure, via occupational sunlight exposure or dietary intake, was associated with reduced renal cell carcinoma

(RCC) risk and to explore if genetic variations within the vitamin D pathway modified risk. Specifically, a statistically significant (38%) reduction in RCC risk was observed with occupational ultraviolet (UV) exposure among male participants. No association between UV exposure and RCC risk was observed among female participants. When analyses were stratified by latitude, as another surrogate for intensity of UV exposure, a stronger (73%) reduction in risk was observed between UV exposure and RCC risk among Russian males residing at the highest study center latitude site.

Comprehensive evaluation of genetic variation across eight vitamin D pathway genes

(VDR , RXRA , RXRB , CYP24A1 , GC , THRAP4 , THRAP5 , STAT1 ) revealed, both the

vitamin D receptor ( VDR ) and retinoid-X-receptor-alpha (RXRA ) genes were

significantly associated with renal cancer risk. Globally significant at the gene level,

both genes ( VDR and RXRA ) contained chromosomal regions with particularly strong

associations. Across the VDR gene, three haplotypes within two regions (intron 2 and

intron 4) were significantly associated with increased cancer risk. Across the RXRA

gene, one haplotype located downstream, 3´ of the coding sequence, increased RCC risk

approximately 35% among individuals with the variant haplotype compared to subjects

with the most common referent haplotype.

157

Further analysis of the VDR and RXRA genes were evaluated to see if select single nucleotide polymorphisms (SNPs) across these genes modified associations between

RCC risk and dietary intake frequency of vitamin D or calcium rich foods, and occupational UV exposure. First, unadjusted dietary analyses showed reduced RCC risk with increasing intake frequency of vitamin D rich foods, specifically egg and total vitamin D (based on liver, egg, and fish); increased RCC risk was observed with increasing intake frequency of calcium rich food, specifically yogurt and total calcium

(based on yogurt, milk, and cheese). After adjustment of covariates, total vitamin D and total calcium intake frequency were no longer significant although the direction in risk remained consistent. SNP analyses across VDR and RXRA genes showed that only the

RXRA gene modified associations between RCC risk and frequency of dietary vitamin D and calcium rich foods. RXRA SNPs, located downstream, 3 ′ of the coding sequence, were shown to modify associations between vitamin D rich foods (specifically egg) and

RCC risk. Subsequently, RXRA SNPs, located in introns one and four, were shown to modify associations between calcium rich foods (specifically yogurt and total calcium) and RCC risk. No association with occupational UV exposure, VDR or RXRA SNPS, and

RCC risk were found; however, among males in the highest occupational UV exposure category, carriers with the TTG (rs4760648, rs2853564 and rs2254210) VDR haplotype had observed an increase in risk compared to those with the referent CCA haplotype.

6.2 Vitamin D:

An essential nutrient for the prevention of rickets, vitamin D has long been recognized

158 for its role in bone health. Regulating and maintaining serum calcium and phosphorus concentrations, vitamin D promotes calcium absorption from the gut, thus enabling bone mineralization [67, 68] . Only recently, has vitamin D been acknowledged for its importance in maintaining and promoting proper function of the nervous and organ system [67-75] . Modulating both neuromuscular and immune functions and reducing inflammation, vitamin D plays a complex and vital role in human health [67, 68] .

Clinical and epidemiological studies conducted within the past decade suggest that vitamin D may be linked to a number of health conditions. Vitamin D insufficiency is thought to be linked to increased risk of heart disease, hypertension, a variety of different auto-immune diseases, as well increased risk of cancer [81, 137, 138] . While the exact anti-carcinogenic mechanism of vitamin D is fully not understood, vitamin D and its metabolites are thought to impede carcinogenesis by stimulating cell differentiation, inhibiting cell proliferation, inducing apoptosis, and suppressing invasiveness, angiogenesis, and metastasis [18, 73-75] .

Vitamin D is a fat soluble steroid hormone found in both foods and produced in the body after exposure to the sun. There are two forms of vitamin D, ergocacliferol (vitamin D 2) derived from plants and cholecalciferol (vitamin D 3) generated within the skin after UVB exposure to sunlight [78, 84, 86] . Neither vitamin D 2 nor D 3 have significant biological activity and must be hydroxylated in the body, first in the liver to 25-hydroxylase

(25(OH)) and then in the kidney to calcitriol (1,25 dihydroxycholecalciferol

(1,25(OH) 2D)), the hormonally active form of the vitamin [78, 84, 86] . Both increased

159 exposure to sunlight and increased dietary intake of vitamin D will increase serum

concentrations of 25(OH) vitamin D making this metabolite a useful indicator of vitamin

D status [88, 89] .

Considerable progress in understanding the relationship between vitamin D and certain

health conditions, particularly cancer, has been made over the past decade. Yet,

epidemiological studies of the vitamin are quite limited for many sites, such as that of the

kidney, the major organ involved in vitamin D metabolism and activity, and far from

consistent. There are many questions still to be answered regarding the complex

relationship between vitamin D and the risks of developing or surviving cancer. While

we have vastly improved our understanding of vitamin D in recent years, the knowledge required for making recommendations for cancer prevention, such as the most effective level/dose of dietary vitamin D intake or sunlight exposure, is still unknown.

Furthermore, our knowledge regarding how vitamin D pathway genes may modify the relationship between cancer risk and vitamin D is still limited.

6.3 Vitamin D Sunlight Exposure:

Cholecalciferol is the preferred form of vitamin D because it is the natural form of the vitamin that is made in the skin and is approximately 50% to 80% more effective than ergocacliferol in maintaining 25(OH) vitamin D levels [273] . In general, most individuals can obtain 100% of vitamin D requirements through sunlight exposure [77,

273] . Certain factors, such as skin color, age, health, latitude, cloud cover, and pollution, affect the amount of vitamin D obtained from solar UVB exposure. However, adequate

160 levels of vitamin D among healthy Caucasians can be made in the skin with as little as ten to fifteen minutes of natural unprotected sun exposure at least two times per week to the face, arms, hands, or back [76, 78] .

Most studies of vitamin D UV exposure have been in relation to breast, prostate, colon or lymphatic cancers. Only a limited number of studies, the majority being ecological, have investigated the association between vitamin D UVB exposure and renal cancer risk.

Kidney cancer mortality risk across the United States (U.S.) was shown to be inversely associated with UVB radiation in an ecological study investigating cancer deaths from

1970-1994 [94] . Similarly, an inverse correlation between kidney cancer mortality risk and UV exposure was also observed across European countries [95] , where potential confounding factors such as those related to diet and socioeconomic status were also considered, but were not shown to affect results.

The association between kidney cancer incidence and UVB irradiance across 175

European countries was investigated in a recent ecological study, where a protective effect was observed for both males and females with increasing estimated sunlight exposure [97] . Residents of countries situated at the highest latitudes had the highest incidence rates of renal cancer [97] . Further protective evidence between kidney cancer risk and UVB exposure was described by Boscoe and colleagues, who through the use of daily satellite-measured solar UVB levels, investigated over three million incidental cancer cases between 1998 and 2002 and cancer deaths from 1993 to 2002 in the U.S.

[96] .

161

With the exception of our results, to date, no case-control study has examined the relationship between occupational sunlight exposure and renal cancer risk. Furthermore, only one occupational cohort study has explored the relationship between sunlight exposure and kidney cancer risk, where a statistically significant 30% reduction in cancer risk was observed among male construction workers in Sweden [98]. Additional supporting evidence for kidney cancer risk and UV exposure was shown in a cohort study investigating the risk of second primary tumors and UV exposure among 10,886 melanoma and 35,620 non-melanoma skin cancer cases [99] . A protective effect for several solid cancers, including that of the kidney, was more prominent for participants residing in sunny countries compared to participants residing in less sunny countries, suggesting that although sun exposure is a risk factor for melanoma, it may also be a protective factor for internal cancers through the production of vitamin D.

Epidemiological studies, the majority of which are ecological, have generally observed a protective effect between sunlight exposure and kidney cancer mortality or incidence risk. Unfortunately, the intrinsic limitations resulting from the use of aggregate data on exposure and disease make it difficult to establish a causal relationship between UV exposure and cancer risk from these ecological studies. Furthermore, the lack of consistent evidence between other cancer sites [101-134] and UV exposure, particularly occupational UV exposure, may reflect the degree of misclassification due to exposure assessment, usually based on work tasks and job titles, which limits the weight of evidence. Most studies also fail to account for recreational sunlight exposures or

162 exposure during vacations [125, 126] . To our knowledge, no studies to date have examined the relative levels of 25(OH) vitamin D among outdoor workers and those with greater recreational time outdoors. This leads to the related issue that factors correlated with sunlight (such as outdoor exercise) may be confounding the relationship between vitamin D UV exposure and cancer risk.

While the causal relationship between sunlight exposure and cancer risk can probably best be established by examining 25(OH) vitamin D serum levels, the majority of non- ecological studies investigating the relationship between sunlight exposure and cancer risk have been case-controls studies, where 25(OH) vitamin D levels may not accurately reflect vitamin D status among participants. First of all, a single measurement of 25(OH) vitamin D is not reflective of long-term vitamin D status, particularly among cases recently diagnosed with cancer. Secondly, 25(OH) vitamin D levels are related to dietary intake and sun exposure. Therefore, assessment of dietary sources of vitamin D must also be accounted for in order to more accurately assess the relationship between UV exposure and cancer risk when analyzing 25(OH) vitamin D levels. Unfortunately, the majority of studies conducted to date have not considered dietary sources of vitamin D when examining the association between 25(OH) vitamin D levels and cancer risk among

UV exposed participants.

6.4 Dietary Vitamin D:

Based on numerous studies that have established a link between diet and cancer risk or progression, the World Health Organization recently acknowledged that after smoking,

163 diet may be the second most preventable cause of cancer [296]. While, dietary vitamin D intake constitutes a small proportion (less than 10%) of vitamin D requirements [135] , scientific studies suggest that dietary intake of vitamin D may play an important role in determining cancer risk, particularly in areas where UVB intensity levels vary by season.

Dietary sources of vitamin D are limited to only a few natural food sources, such as fatty fish like salmon, herring or mackerel. Fortified foods such as milk and cereals constitute a larger proportion of dietary vitamin D intake, particularly in the U.S., where food sources are commonly fortified with vitamin D. Fortification of foods with vitamin D however, is not a common practice throughout most parts of the world. Another source of vitamin D intake includes supplements, which have rarely been examined.

No studies to date have investigated the association between vitamin D supplement intake and renal cancer risk. Furthermore, studies examining the relationship between dietary vitamin D intake and kidney cancer are limited. An Italian case-control study of 767 cases and 1,534 controls observed a statistically significant inverse association between dietary vitamin D intake, based on a 78-item food frequency questionnaire, and cancer risk [141] . The association was more pronounced among females than males when analyses were stratified by sex. Inconsistent results have also been reported for studies exploring the association between specific vitamin D food sources, such as fatty fish consumption, and renal cancer risk. A Swedish cohort study of 61, 433 middle aged (40-

76 years) women reported an inverse association between fatty fish consumption and renal cancer risk (p-trend= 0.02). Among the 150 incidental renal cases observed, no association was found for consumption of lean fish [34] . A cohort study of male smokers

164 in Finland however, observed increased risk of renal cancer with increasing consumption

of Baltic herring (p-trend <0.001) [297] . A high correlation between fish consumption and contaminated blood levels, high in mercury, lead, cadmium, dioxin, and polychlorinated biphenyls was also observed, suggesting that heavy metals contaminants in this region of the Baltic Sea may have confounded results.

While our study did find an association between intake frequency of dietary vitamin D rich foods (mostly egg) and RCC risk, poor assessment of dietary vitamin D may explain the discrepancy between epidemiological studies that have investigated dietary vitamin D intake and cancer risk [142-191, 297, 298] . For example, fatty fish and not lean fish constitute the main source of vitamin D obtained from fish consumption. Therefore, studies that report an association with total fish consumption may be too simplistic and may not represent the larger picture that exists between dietary vitamin D intake and cancer risk. Although our study did not find an association between RCC risk and dietary intake frequency of fish consumption, null results for both intake frequency of saltwater, freshwater fish, and total fish consumption were observed. Furthermore, failure to account for potential confounders, such as dietary vitamin D supplements and other micronutrients that affect vitamin D status, such as folate, calcium, retinols, and phosphorus, may bias results. Similarly, possible confounding due to healthy dietary or lifestyle patterns may also bias results.

6.5 Common Vitamin D Receptor (VDR) Polymorphisms:

Within the kidney, vitamin D mediates its biological effect by binding to VDR [75, 192] .

165 Localized to the human chromosome 12q13-14, VDR contains 14 exons and spans approximately 75 kilobases in length of genomic deoxyribonucleic acid (DNA) [298] .

The expression of VDR variants are associated with a number of conditions, including autoimmune diseases, resistance to vitamin D therapy, susceptibility to infections and cancer [194] . Epidemiological studies investigating the association between variations in the VDR gene and cancer risk have primarily focused on five common polymorphisms

(FokI, BsmI, TaqI, ApaI , and poly(A) ) [194, 195, 284] . With the exception of FokI , the functional implications of these common VDR variants are not well understood and are suspected to be in linkage disequilibrium (LD) with one or more functional polymorphisms, which may affect messenger ribonucleic acid (mRNA) stability [200] .

The FokI variant on the other hand is not linked to other VDR polymorphisms and results in a three amino acid shorter molecule with high biological activity [196, 197] .

With the exception of our study results, which comprehensively evaluated 29 SNPs across the VDR gene, only three studies to date have investigated the association between common VDR variants and kidney cancer risk. One-hundred thirty-five RCC cases and

150 controls were genotyped in a recently published Japanese case-control study investigating the association between common VDR variants and cancer risk.

Participants with the AA ApaI allele were shown to have significant increased risk of renal cancer compared to participants with the aa allele [202] . In a second Japanese case- control study, Ikuyama and collogues reported significant increased risk of rapid growth renal cancer among participants with the TT TaqI allele compared to participants with the referent tt allele [203] . However, no association between FokI , TaqI , or BsmI SNPs and

166 RCC risk was observed in a large hospital-based case-control study conducted in Europe

[201] .

The association between common VDR SNPs (rs1544410 and rs10735810) and RCC risk was not observed in our study. However, most epidemiological studies investigating the association between overall cancer risk and common VDR polymorphisms have been inconsistent [77, 118, 198, 204-240, 298] . Furthermore, while the allelic variants of the

VDR gene occur naturally in the human population, allelic frequencies for these common polymorphisms appear to vary considerably by race, and ethnic group. Therefore, the need for larger studies to explore the relationship between common VDR variants and cancer risk within specific populations is critical to validate results. Additionally, most epidemiological studies fail to investigate gene-environment interactions, which may partially explain some of the variations observed across different studies. Cancer risk associated with common VDR variants may be modified by age, diet, or other lifestyle/environmental factors and studies that fail to account for such potential interactions may mask associations.

6.6 Vitamin D Pathway Genes:

The transport, metabolism, binding, activation, and mechanistic action of vitamin D within the body are complex. Although VDR is the only nuclear protein that binds the biologically most active vitamin D metabolite with high affinity, the concert of genes that interact with vitamin D and its pathway genes is vast [299] . If vitamin D is an important factor in the development of cancer, then common functional sequence polymorphism in

167 the genes that influence vitamin D metabolism, transport, binding, function and/or

expression may predispose individuals to disease.

Vitamin D, dietary or synthesized, is released into the bloodstream where the vitamin

binds with high affinity to the vitamin D binding protein [242] . In plasma, the group

specific component (GC) vitamin D binding protein serves as the major carrier of vitamin

D and its metabolites in plasma to target tissue [242, 243] . The vitamin D binding

protein has a stronger affinity for vitamin D 3 than D 2 [195] . In plasma, greater than 99% of circulating vitamin D is protein bound, which is why vitamin D binding proteins are approximately 20 times higher in concentration in plasma than vitamin D concentrations

[195] .

Inside the cell, the vitamin D binding protein is degraded and vitamin D is hydroxylated in the liver into 25(OH) vitamin D where the vitamin is then released and metabolized to

1,25(OH) 2 vitamin D by 1, α-hydroxylasse or 24-hydroxylase ( CYP24A1 ) [195, 249-251] .

The biological activity of 1,25(OH) 2 vitamin D is mostly mediated by a high-affinity

receptor, the vitamin D receptor where research has shown that cells, including those of

cancer cells, express specific ( VDR ) receptors for calcitriol [273, 300] . VDR is found in a

number of organs and cells (Table 7) and is known to be involved in cell proliferation and

differentiation, which affect cancer risk [273, 300] .

168 Table 7: Tissues that express VDR [273, 300]

Adipose Lung Adrenal Cortex Lymphocytes (B & T) Adrenal Mendulla Macrophages Bone Monocytes Bone marrow Muscle Brain Neurons Breast tissue Osteoblast Cancer cells Ovary Cartilage Pancreatic islets Colon epithelial Parathyroid Decidua Parotide Dendritic cells Placenta Endothelial cells Pituitary Epididymis Prostate Glia Retina Hair follicle Skeletal muscle Heart Skin Intestine Stomach Keratinocytes Testis Kidney Thyroid Liver Uterus

Inside the nucleus, active vitamin D binds to VDR and forms a heterodimer complex with the retinoid-X-receptor ( RXR ) gene [254] . This complex is then directed to the vitamin D response element ( VDRE ) in the promoter region of a broad spectrum of 1,25 regulated genes to initiate transcription, thus, modulating via up-regulating or down-regulating gene expression [253, 254] . Greater than 60 genes in various tissue are known to be regulated by 1,25(OH) 2 vitamin D [195, 273, 300] .

Although VDR , and to some degree RXR , are generally the two genes believed crucial in the mediation of calcitriol to regulate the transcription of various genes, the interaction of other genes with VDR and/or RXR may impede this process. For example, both the thyroid hormone receptor associated protein ( THRAP ) genes and the STAT1 gene have

169 been shown to inhibit the VDR-RXR complex to VDRE , which hinders transcription [260-

262, 264-266] .

Comprehensive analysis of vitamin D pathway genes has only been examined in relation to breast and prostate cancer risk. Three vitamin D pathway genes ( VDR, GC, and

CYP24A1 ) were analyzed in relation to breast cancer risk in a nested case-control study of 500 breast cancer cases and 500 matched controls [236] . No association was observed between breast cancer risk and the five common VDR (FokI, TaqI, BsmI, ApaI , and poly(A) ) variants, the two GC variants, or the single CYP24A1 variant that were examined. Additionally, dietary vitamin D intake was not shown to modify the association between these SNPs and breast cancer risk [236] . Three vitamin D pathway genes were also assessed in relation to prostate cancer risk in a population-based case- control study of middle-aged (40 to 64 years old) men [301] . Common variants across

VDR (N= 22), CYP24A1 (N= 14), and CYP27B1 (N= 2) SNPs were analyzed and only two VDR loci (rs2107301 and rs2238135), located in introns 2 and 4, respectively, were associated with a 2 fold to 2.5 fold higher risk of prostate cancer. Haplotypes for VDR and CYP24A1 were not associated with prostate cancer risk; furthermore, other gene- environment interactions were not considered in the study [301] .

As described above, the vitamin D pathway is complex and the genes that interact with vitamin D and vitamin D pathway genes are extensive. To date, only two vitamin D pathway papers in relation to cancer risk have been published [236, 301] . While the results of one of these studies [301] were similar to the results reported in our study,

170 additional research that comprehensively examines common variants across vitamin D

pathway genes are needed to validate results. Furthermore, the complexities associated

with gene-environment interactions, such as diet, occupation, and other risk factors, have

yet to be understood in relation to these genes and should also be examined in future

studies.

6.7 Gene-environment interactions- Vitamin D pathway genes, dietary vitamin D intake,

and UV exposure:

To date no epidemiological study has comprehensively investigated whether vitamin D

pathway genes modify the association between RCC risk and dietary vitamin D or

calcium intake. However, VDR SNPs have been analyzed in relation to vitamin D intake with regards to other cancers. For example, in a small U.S. colorectal case-control study, a nearly significant interaction between vitamin D intake, the BsmI VDR SNP and

colorectal cancer risk was reported (p-interaction= 0.09) [302] . A statistically significant

interaction was also reported between colorectal cancer risk, consumption of dairy

products (p-interaction= 0.02) and the VDR ApaI SNP in a recent case-control study conducted in Britain, an area where milk is not fortified with vitamin D [303] .

Additionally, a significant interaction was observed between the VDR combined polyA/BsmI SNPs, calcium intake and rectal cancer risk in a large (952 cases and 1,205 controls) U.S. rectal cancer case-control study (p-interaction= 0.01) [105] .

Excluding the results of our study, the association between VDR polymorphisms in relation to RCC risk and UV exposure has yet to be evaluated; however, a handful of

171 studies have looked at this association in relation to other cancers. For instance, a strong interaction (p-interaction= 0.004) between sunlight exposure, VDR polymorphisms ( TaqI and BsmI ) and lymphoma risk was newly reported in a U.S. population-based case- control study [129] . Yet, no gene-environment interactions were observed between prostate cancer risk, UV exposure, and the TaqI or FokI VDR SNPs in a recent case- control study conducted among Caucasians in Northern Europe [304, 305] .

As mentioned before, the complexities associated with gene-environment interactions, especially those that involve diet or occupational exposures, have yet to be understood in relation to vitamin D pathway genes. However, additional studies particularly those that investigate the risk of renal cancer and vitamin D (via diet, supplement intake and/or sunlight) exposure and vitamin D pathway genes are needed to help elucidate the etiology of RCC.

6.8 Strengths & Limitations of the Dissertation Study:

The Central and Eastern European Renal Cell Carcinoma (CEERCC) Study is extraordinary in that it is the largest kidney cancer case-control study conducted to date with biological samples. Furthermore, approximately half of the cases recruited for this study were residents of the Czech Republic, an area with among the highest reported incidence of RCC in the world [3, 12]. Although there were various strengths and limitations associated with each of the three manuscripts, which are described in more detail below, all three papers are noteworthy in that they included a large samples size, had high participation rates among both cases and controls, included only early-onset

172 cases (within three months of diagnosis), and were comprised of only histologically

confirmed cancers.

Manuscript One:

The only case-control study to date to evaluate the association between renal cancer risk and occupational UV exposure, this research is extraordinary in that the collection and assessment of detailed occupational exposure histories were conducted by trained industrial hygienists. Both detailed job and industry titles were obtained from the

International Standard Classification of Occupation (ISCO) and the Statistical

Classification of Economic Activities of the European Community (NACE) coding systems to assign individual-specific exposure information into a job exposure matrix for occupational UV exposure. Consistent exposure assessment for frequency, intensity, and confidence of occupational sunlight exposure were also confirmed by inter-rater agreement scores.

Several limitations were inherent in this paper. First, sunlight exposure outside of the work environment was not assessed in this study. Vitamin D synthesis from the sun is most likely to occur when UV index levels reach their peak between 10 am and 2 pm (11 am and 3 pm daylight savings time), the times when individuals are generally at work.

However, sunlight exposure during non-work hours may have affected results. Non- differential misclassification of sunlight exposure may have biased results towards the null; all sources of sunlight exposure, such as recreational exposure to UV light, history of sunburns, and sunbathing activities on a regular basis or during vacations, which have

173 been associated with other cancers, were not considered or measured in this study.

Secondly, while participants were primarily of Central European descent, data regarding hair color, eye color, tanning ability, use of sunscreen or protective clothing, such as hats, gloves, long pants, etc. were not ascertained for this study. Moreover, while we were able to control for known RCC risk factors, such as hypertension, smoking, and body mass index (BMI), other potential risk modifiers such as genetics, were not considered in this paper; however, these potential risk modifiers were considered in the third manuscript and are also described in more detail in the Appendix. Additional limitations of this paper included non-differential inaccurate or incomplete recall of all occupational histories, non-differential exposure misclassification, and the use of hospital-based controls. While this manuscript had sufficient statistical power to detect relatively small associations, the number of cases and controls with only high intensity jobs were small, thus limiting the precision of our association within this subgroup.

Lastly, in my dissertation proposal, I had planned to assess how well sun exposure categories were assigned, by calculating correlations between exposure rates and questionnaire data regarding the type of environment where participants regularly worked. For participants in which occupational sunlight exposure assessments were highly correlated (r´ >0.80) with questionnaire data regarding participants’ regular work environment, stratified analyses for cumulative exposure, frequency-adjusted duration of exposure, frequency-adjusted duration of exposure for low intensity jobs, and frequency- adjusted duration of exposure for any high intensity jobs were to be reassessed.

Unfortunately, questionnaire data regarding the type of environment where participants

174 regularly worked were not available for analysis due to some study centers’

unwillingness to back translate participants’ responses into English.

Manuscript Two:

This paper is unique in that it is the first genetic susceptibility study to comprehensively examine the association between VDR and other pathway genes in relation to RCC risk.

An additional strength of this study is the use of HapMap to tag genes of interest using high (80% to 90%) genomic coverage, both upstream and downstream of the target genes. The collection of biological material from a high proportion of subjects provided sufficient statistical power to detect relatively small associations between genotypes and risk. Analysis of genes with significant global p-values reduced the risk of Type I errors in this study, while the use of two different Haplotype-based methods (HaploWalk and

HaploStats) reduced the likelihood of finding significant results based on chance.

However, inherent limitations stemming from uncontrolled confounding, such as diet or occupational exposures may have biased results. Additionally, a hospital-based case- control study design may have potential limitations due to the lack of population controls.

Lastly, while there is general agreement that serum 25(OH) vitamin D levels are the best indicator of current vitamin D status, for both dietary intake and UV exposure, the biochemical marker has a half-life of only three weeks [88, 89] and would not have been beneficial for this case-control study since a single measurement of 25(OH) vitamin D would unlikely reflect long-term vitamin D status, particularly among cases that were recently diagnosed.

175 Manuscript Three:

This manuscript is exceptional in that both vitamin D exposure and intake were examined in relation to RCC risk and vitamin D pathway genes. Furthermore, this paper had sufficient statistical power to detect relatively small genotype associations overall and within particular subgroups, although the power to detect gene-evironment interactions was limited. The major sources of dietary vitamin D in this study were frequency of dietary liver, egg, and fish, which only constitute 15 International Units (IU), 20IU, and

250-360IU (based on type of fish) of vitamin D per serving, respectively [76] . Though dietary intake in general only constitutes a small source of vitamin D requirements, misclassification of vitamin D and/or calcium intake from the use of a 23-item lifetime food frequency questionnaire likely resulted in incomplete control of confounding. Other potential confounders, such as dietary retinol and phosphorous intake, were also not considered in this study. Limitations described above regarding the assessment of occupational UV exposure also pertain to this manuscript, since the relationship between occupational UV exposure, RCC risk, and vitamin D pathway genes were assessed.

Additionally, while hospital-based case-control studies have potential limitations due to the lack of population controls, these studies can improve response rates for the intense collection of biological specimens and therefore reduce the chance of bias in the assessment of gene-environment interactions [306] . Finally, the use of a case-control study also limits us from using serum 25(OH) levels as a marker for long term vitamin D status, since this marker is the best indicator of current vitamin D for both dietary intake and UV exposure [88, 89, 306] .

176 6.9 Direction for Future vitamin D Studies:

Epidemiologic studies of cancer risk typically fall in one of four categories: ecologic, case-control, cohort or controlled trials. The first three are characterized as observational studies because they observe the relationship between risk or protective factors and outcomes in non-experimental populations, people whose behavior or exposures were not determined by researchers. Considered the least informative type of epidemiologic study, ecological studies assess the relationship between exposures and outcome in groups of people, usually geographic groups. Since ecologic studies do not take into account individual exposures or outcomes, they may find false relationships or miss true ones.

Case-control studies are retrospective, beginning with a group of individuals with the disease or outcome of interest and another group of individuals that do not share the outcome of interest. These types of studies look backward in time to identify factors that were more common in one group compared to the other and try to explain why the groups have experienced different outcomes. To the extent that the studies rely on recall to determine exposures or the fact that participating cases or controls do not reflect the same exposures as the base population, biases may result.

Cohort studies are prospective in nature and identify a population which has been or will be exposed to some factors. Cohort studies follow populations forward and outcomes are assessed against risk factors identified before the outcomes have occurred, reducing biases resulting from both recall and selection. Because of this, cohort studies are generally considered more instructive. Nested case-control studies recruit cases and

177 controls from a cohort study population. A number of studies of circulating vitamin D are nested case-control designs, in which blood stored at the baseline of the cohort study is analyzed in a select group of cases and control for cost reasons. Therefore, these studies also reduce recall or selection biases. Trials have often been considered the most informative study design because biases can be controlled through subject randomization and blinding of both the investigators and subjects to the exposure (e.g., with placebos).

In view of the above rationale, if I were to construct my own study design to investigate the association between vitamin D (via sunlight exposure and dietary intake) and cancer risk, I would advocate using a nested case-control study design. The collection and assessment of blood samples, both prior to and after diagnosis (for cases), would provide a more accurate description of long-term vitamin D status. In the first manuscript, gender differences related to UV exposure and cancer risk were speculated to relate to gender related hormonal differences which may have played a role in responses to acute UVB exposure as well as UV-induced tumor development. The use of serum 25(OH) vitamin

D levels from a nested case-control study would facilitate assessing whether or not sex differences exist with regards to UV exposure and/or dietary vitamin D intake.

Detailed information could also be collected from a nested case-control study regarding dietary intake (including supplements), recreational UV exposures outside of the work environment, and personal characteristics such as skin sensitivity. This would also decrease the likelihood of exposure misclassification while allowing for the adjustment of potential confounders. In addition, this type of study design would also facilitate the

178 investigation of serum 25(OH) vitamin D in relation to season of detection and cancer survival, a new prospect in the investigation of vitamin D and cancer.

6.10 Seasonality of Diagnosis:

One area of investigation within the larger question of whether sunlight exposure can provide cancer benefits is the association between season of detection and cancer survival. Two decades ago, some studies in New Zealand [307, 308] and Finland [309] , noted a more favorable prognosis among breast cancer patients diagnosed in summer months compared to patients diagnosed during other seasons. The findings were hypothesized to be due to seasonal changes in hormone levels. Studies on seasonality of diagnosis and survival currently focus on higher levels of 25(OH) vitamin D levels in summer/fall months to explain this relationship.

In Norway, Robsahm and colleagues found that breast, colon, and prostate cancer patients diagnosed in fall months demonstrated 30% lower fatality risks than those diagnosed in the winter [310] . There was, however, no survival benefit to occupational sun exposure. Four other Norwegian studies found marked improvements in survival for patients with breast [311] , lung (in males) [312] , and colon cancers [313] , including

Hodgkin lymphoma [314] diagnosed in autumn or summer months. Investigators hypothesized that vitamin D may confer survival benefits at some sites by enhancing the therapeutic effects of treatment [310, 313, 315] .

The majority of these studies have relied on season as a surrogate for vitamin D exposure

179 and have not used individual measurements of vitamin D 25(OH) levels. Some studies have recognized the possibility that seasonal variations in other factors, such as fruit and vegetable consumption or physical activity [310, 313] , might provide an alternative explanation. Other investigators suspect the seasonal associations with prognosis may be related to seasonal relationships to detection, health care delivery, or winter infections.

However, the question of greater importance may be to determine whether the relative importance of year-long levels of 25(OH) vitamin D or seasonal/less regular intense exposures is more beneficial.

6.11 Benefits of Vitamin D as a Chemo-preventive Agent:

Clinical trials of vitamin D or its analogs as chemo-preventive agents in humans may assist greatly in elucidating the role of vitamin D on cancer risk and survival. A recent meta-analysis of randomized controlled trials showed supportive evidence that vitamin D supplementation decreased the risk of all-cause mortality [316] . Consistent, though not statistically significant, colorectal mortality risk was shown to decrease (RR= 0.88; 95%

CI= 0.74-1.06) with vitamin D supplementation in a five year randomized double-blind controlled trial [317] . Significant reduction in overall cancer risk (RR= 0.23; 95% CI=

0.09–0.60) however, was observed in a large randomized trial, followed for four years, for moderate (1,000 IU) vitamin D and calcium supplementation [318] . Further beneficial effects of vitamin were observed in a clinical trial for patients with inoperable advanced hepatocellular carcinoma (HCC), where a daily dose of 10 mg of Seocalcitol for up to one year resulted in reduced tumor dimension among HCC patients [319] . Similar effects were also observed among 37 patients with metastatic androgen-independent prostate

180 cancer (AIPC) who were treated with oral calcitriol/docetaxel. High dose treatment of

calcitriol/docetaxel showed prostate-specific antigen (PSA) reductions of at least 50%

[320] . However, when 19-nor-1alpha-25-dihydroxyvitamin D 2 (paricalcitol), a vitamin D analogue, was assessed in a Phase I/II trial of patients with AIPC, no substantial drop of

50% or greater in PSA was observed among the participants (N= 18) who received intravenous paricalcitol three times per week [321] . Null results were also reported by

Wactawski-Wende and colleagues, in a seven-year clinical trial of vitamin D

supplementation and invasive colorectal cancer among females obtaining 500mg calcium

with 200IU vitamin D twice a day [322] . The null trial findings may, however, have

reflected low trial doses, insufficient follow-up, or the fact that participants may have

already benefited from baseline levels of vitamin D and calcium intake since at

enrollment, participants had a mean total calcium and vitamin D level that was twice that

of the national average.

6.12 Replication of Results:

There are currently two renal carcinoma studies underway at the National Cancer

Institute (NCI) that will allow us the opportunity to replicate results presented in the

dissertation, as well as expand upon analyses already conducted. The first, the United

States Renal Cell Carcinoma (USRCC) Study, is a case-control study of approximately

1,200 cases and 1,200 matched controls from the Detroit and Chicago areas. Data

regarding demographic characteristics, education, exposure to tobacco smoke and

alcohol, family history, medical history, and occupational history were collected for this

study. Additionally, genomic DNA was also ascertained from a high proportion of

181 participants.

The association between occupational sunlight exposure and renal cancer can be assessed

using data from the USRCC Study. Compared to the CEERCC Study, misclassification

of occupational sunlight exposure in this study is expected to be reduced. In addition to collecting lifetime occupational data regarding the title, the detailed tasks, the company, and the year of beginning and ending employment for jobs held for at least 12 months duration, data regarding hours per week individuals spent walking or biking to work as well as the number of hours participants spent per week at each job was also ascertained in this study. This information will allow us to more accurately assess occupational UV exposure assessment, as well as the ability to adjust for recreation UV exposures.

The relationship between dietary vitamin D intake and renal cancer risk for the USRCC

Study will be assessed based on an 84-item food frequency questionnaire. Dietary vitamin D intake will also be more accurately assessed in the USRCC Study compared to the CEERCC Study since there are more detailed questions regarding specific vitamin D food sources. For example, data regarding the type of fish consumed (herring, sardines, anchovies, tuna, and salmon) were collected. Additionally, since the U.S. regularly fortifies certain foods with vitamin D this will provide more variability across exposure categories, especially since data regarding the consumption of soy, milk, and cereal products were gathered. Fortunately, unlike the CEERCC Study, data regarding potential confounding dietary factors, like retinol and folate intake, were collected and can be assessed.

182

We hope to replicate results from the vitamin D pathway analysis presented from the

CEERCC Study with DNA analysis of genes in the USRCC. Since the USRCC study includes a high proportion of African American participants, the opportunity to look at associations stratified by race will be available. This will be the largest RCC study with biological material to include African Americans.

The second RCC study sponsored by the NCI and the International Agency for Research on Cancer (IARC) is the Follow-up CERRCC Study. Of the 1,097 RCC cases from the

CEERCC Study, data on treatment, death, recurrence/progression of second and/or third primary tumors, and tobacco and alcohol exposures will be ascertained. Information from this study can be used to assess whether genetic variants in vitamin D pathway genes are associated with survival. Additionally, the association between occupational

UV exposure and cancer survival may also be reviewed, although as described above the limitations of the study should be considered carefully when interpreting results.

6.13 In Summary:

According to the Surveillance Epidemiology and End Results (SEER) Program statistics, an estimated 67,720 Americans are expected to be diagnosed with skin cancer (excluding

Basal and Squamous) in 2008, while an estimated 11,200 will die from the disease [323] .

Based on these facts and those reported in Table 8, mass media approaches have been widely used in public health programs to address behavioral risk factors, such as sunscreen use or sun exposure avoidance in efforts to prevent skin cancer. While it is

183 well accepted that reduced sun exposure can decrease the risk of cataracts and skin

cancers [267] , most public health messages related to sun avoidance have failed to acknowledge the recent beneficial discoveries relating vitamin D (which is generally

obtained by solar UV exposure) to a variety of different health conditions, such as auto-

immune diseases, depression, hypertension, and cancer [274-276] . Furthermore, UV prevention efforts have also been partly to blame for the epidemic rates of vitamin D deficiency reported over the past 25 years [81, 268-271] . Although it is not known whether the increased incidence of certain cancers over the past two decades, such as that of the kidney, is associated with the reduced prevalence of vitamin D, the health benefits of UV exposure should not be ignored, particularly since epidemiological UV studies are quite limited for many sites.

Table 8: SEER Skin Cancer (excluding Basal and Squamous) Rates

from 2001-2005 [323]

Among Men Among Women Incidence Rates All Races 27.1 per 100,000 men 17.0 per 100,000 women Whites 31.0 per 100,000 men 19.9 per 100,000 women

Mortality Rates All Races 5.3 per 100,000 men 2.2 per 100,000 women Whites 5.9 per 100,000 men 2.4 per 100,000 women

5-Year Survival Rates White 88.4% 93.2%

Given the potential cancer prevention promise of vitamin D, which includes but is not

limited to inhibition of clonal tumor cell proliferation, hematopoieses, induction of

immune cell differentiation, reduction of inflammation, and apoptosis, it is important to

184 investigate the relationship between vitamin D and cancer risk [18, 68, 73-75]. This is particularly relevant for renal cell carcinoma since the kidney is a major organ for vitamin D metabolism and activity, and calcium homeostasis [6, 81, 277] .

Epidemiological studies are needed to determine exactly how much sun exposure and dietary vitamin D intake are required for adequate vitamin D levels among individuals.

Additionally, the results of these studies can help determine whether a policy driven intervention that fortifies foods is the best approach for increasing vitamin D levels. This type of intervention would not only increase vitamin D levels among consumers, but would also avoid the risks associated with excessive sun exposure.

The results of the present dissertation study have contributed to the limited data currently available on vitamin D sunlight exposure and dietary vitamin D intake in relation to renal cell cancer risk. Although there is still much to be learned and researched regarding vitamin D and renal cancer, until more well-designed epidemiological studies for specific cancers are conducted, questions regarding the most effective/beneficial dose of vitamin

D intake or sunlight exposure and risk of surviving or developing cancer still exist.

Furthermore, the complexities of how vitamin D pathway genes may modify cancer risk are still not well understood. However, with increased understanding of the vitamin, eventually public health messages or other interventions will help individuals to predict their own risk assessment for vitamin D (via sunlight exposure or dietary intake) in relation to cancer risk

185 References:

[1] Motzer RJ, Russo P, Nanus DM, Berg WJ. Renal cell carcinoma. Curr Probl Cancer

1997;21(4):185-32.

[2] Lipworth L, Tarone RE, McLaughlin JK. The epidemiology of renal cell carcinoma. J

Urol 2006;176(6 Pt 1):2353-8.

[3] International Agency for Research on Cancer. GLOBACAN. Available at: http://www-dep.iarc.fr/ Last Accessed on April 2, 2008.

[4] Moore LE, Wilson RT, Campleman SL. Lifestyle factors, exposures, genetic susceptibility, and renal cell cancer risk: a review. Cancer Invest 2005;23(3):240-55.

[5] SEER Cancer Statistics Review, 1975-2001 (NCI 2004). Available at http://seer.cancer.gov Last Accessed on April 2, 2008.

[6] Giovannucci E. The epidemiology of vitamin D and cancer incidence and mortality: a review (United States). Cancer Causes Control 2005;16(2):83-95.

[7] Garland CF, Garland FC, Gorham ED, Lipkin M, Newmark H, Mohr SB, et al. The role of vitamin D in cancer prevention. Am J Public Health 2006;96(2):252-61.

186 [8] Chow WH, Devesa SS, Moore LE. Chapter 43: Epidemiology of renal cell carcinoma.

Comprehensive Textbook of Genitourinary Oncology, Third edition. Lippincott: William

& Wilkins, 2006. (p669-79).

[9] Cancer incidence in five continents (Vol 1X): Curado MP, Edwards B, Shin HR,

Storm H, Ferlay J, Heanue M, Boyle P, eds (2007) Cancer Incidence in Five Continents,

Vol. IX IARC Scientific Publications No. 160, Lyon, IARC.

[10] Murai M, Oya M. Renal cell carcinoma: etiology, incidence and epidemiology.

Curr Opin Urol 2004;14(4):229-33.

[11] Godley P, Kim SW. Renal cell carcinoma. Curr Opin Oncol 2002;14(3):280-5.

[12] Hung RJ, Moore L, Boffetta P, Feng BJ, Toro JR, Rothman N, et al. Family history and the risk of kidney cancer: a multicenter case-control study in Central Europe. Cancer

Epidemiol Biomarkers Prev 2007;16(6):1287-90.

[13] Whang YE, Godley PA. Renal cell carcinoma. Curr Opin Oncol 2003;15(3):213-6.

[14] Drucker BJ. Renal cell carcinoma: current status and future prospects. Cancer Treat

Rev 2005;31(7):536-45.

[15] Surveillance Epidemiology and End Results Program: Kidney and Renal Pelvis.

187 Available at: http://seer.cancer.gov/statfacts/html/kidrp.html?statfacts_page=kidrp.

html&x=16&y=20 Last Accessed on April 2, 2008.

[16] Wronkowski Z, Romejko M, Zwierko M. Survival of cancer patients in Poland.

Cancer Detect Prev 1993;17(4-5):469-74.

[17] Damhuis RA, Kirkels WJ. Improvement in survival of patients with cancer of the kidney in Europe. Eur J Cancer 1998;34(14):2232-5.

[18] Hino O, Kobayashi T, Momose S, Kikuchi Y, Adachi H, Okimoto K. Renal carcinogenesis: genotype, phenotype and dramatype. Cancer Sci 2003;94(2):142-7.

[19] Hsu CC, Chow WH, Boffetta P, Moore L, Zaridze D, Moukeria A, et al. Dietary risk factors of renal cell carcinoma in eastern and central Europe. Am J Epidemiol

2007;166(1):62-70.

[20] Mellemgaard A, McLaughlin JK, Overvad K, Olsen JH. Dietary risk factors for renal cell carcinoma in Denmark. Eur J Cancer 1996;32A(4):673-82.

[21] Hu J, Mao Y, White K. Diet and vitamin or mineral supplements and risk of renal cell carcinoma in Canada. Cancer Causes Control 2003;14(8):705-14.

[22] Rashidkhani B, Lindblad P, Wolk A. Fruits, vegetables and risk of renal cell

188 carcinoma: a prospective study of Swedish women. Int J Cancer 2005;113(3):451-5.

[23] Nicodemus KK, Sweeney C, Folsom AR. Evaluation of dietary, medical and lifestyle risk factors for incident kidney cancer in postmenopausal women. Int J Cancer

2004;108(1):115-21.

[24] Parker AS, Cerhan JR, Lynch CF, Ershow AG, Cantor KP. Gender, alcohol consumption, and renal cell carcinoma. Am J Epidemiol 2002;155(5):455-62.

[25] Curti BD. Renal cell carcinoma. JAMA 2004;292(1):97-100.

[26] Ursin G, Bjelke E, Heuch I, Vollset SE. Milk consumption and cancer incidence: a

Norwegian prospective study. Br J Cancer 1990;61(3):456-9.

[27] Prineas RJ, Folsom AR, Zhang ZM, Sellers TA, Potter J. Nutrition and other risk factors for renal cell carcinoma in postmenopausal women. Epidemiology 1997;8(1):31-

6.

[28] Wolk A, Lindblad P, Adami HO. Nutrition and renal cell cancer. Cancer Causes

Control 1996;7(1):5-18.

[29] Lee JE, Hunter DJ, Spiegelman D, Adami HO, Bernstein L, van den Brandt PA, et al. Intakes of coffee, tea, milk, soda and juice and renal cell cancer in a pooled analysis of

189 13 prospective studies. Int J Cancer 2007;121(10):2246-53.

[30] Galeone C, Pelucchi C, Talamini R, Negri E, Montella M, Ramazzotti V, et al. Fibre

intake and renal cell carcinoma: a case-control study from Italy. Int J Cancer

2007;121(8):1869-72.

[31] Bosetti C, Rossi M, McLaughlin JK, Negri E, Talamini R, Lagiou P, et al.

Flavonoids and the risk of renal cell carcinoma. Cancer Epidemiol Biomarkers Prev

2007;16(1):98-101.

[32] Lee JE, Giovannucci E, Smith-Warner SA, Spiegelman D, Willett WC, Curhan GC.

Intakes of fruits, vegetables, vitamins A, C, and E, and carotenoids and risk of renal cell cancer. Cancer Epidemiol Biomarkers Prev 2006;15(12):2445-52.

[33] Bravi F, Bosetti C, Scotti L, Talamini R, Montella M, Ramazzotti V, et al. Food groups and renal cell carcinoma: a case-control study from Italy. Int J Cancer

2007;120(3):681-5.

[34] Wolk A, Larsson SC, Johansson JE, Ekman P. Long-term fatty fish consumption and renal cell carcinoma incidence in women. JAMA 2006;296(11):1371-6.

[35] Rashidkhani B, Akesson A, Lindblad P, Wolk A. Major dietary patterns and risk of renal cell carcinoma in a prospective cohort of Swedish women. J Nutr

190 2005;135(7):1757-62.

[36] van Dijk BA, Schouten LJ, Kiemeney LA, Goldbohm RA, van den Brandt PA.

Vegetable and fruit consumption and risk of renal cell carcinoma: results from the

Netherlands cohort study. Int J Cancer 2005;117(4):648-54.

[37] Hu J, Ugnat AM; Canadian Cancer Registries Epidemiology Research Group.

Active and passive smoking and risk of renal cell carcinoma in Canada. Eur J Cancer

2005;41(5):770-8.

[38] Mucci LA, Lindblad P, Steineck G, Adami HO. Dietary acrylamide and risk of renal

cell cancer. Int J Cancer 2004;109(5):774-6.

[39] Handa K, Kreiger N. Diet patterns and the risk of renal cell carcinoma. Public Health

Nutr 2002;5(6):757-67.

[40] Bianchi GD, Cerhan JR, Parker AS, Putnam SD, See WA, Lynch CF, et al. Tea

consumption and risk of bladder and kidney cancers in a population-based case-control

study. Am J Epidemiol 2000;151(4):377-83.

[41] Augustsson K, Skog K, Jägerstad M, Dickman PW, Steineck G. Dietary heterocyclic amines and cancer of the colon, rectum, bladder, and kidney: a population-based study.

Lancet 1999;353(9154):703-7.

191

[42] De Stefani E, Fierro L, Mendilaharsu M, Ronco A, Larrinaga MT, Balbi JC, et al.

Meat intake, 'mate' drinking and renal cell cancer in Uruguay: a case-control study. Br J

Cancer 1998;78(9):1239-43.

[43] Yuan JM, Gago-Dominguez M, Castelao JE, Hankin JH, Ross RK, Yu MC.

Cruciferous vegetables in relation to renal cell carcinoma. Int J Cancer 1998;77(2):211-6.

[44] Hu J, La Vecchia C, DesMeules M, Negri E, Mery L; Canadian Cancer Registries

Epidemiology Research Group. Meat and fish consumption and cancer in Canada. Nutr

Cancer 2008;60(3):313-24.

[45] Hu J, Chen Y, Mao Y, Desmeules M, Mery L; Canadian Cancer Registries

Epidemiology Research Group. Alcohol drinking and renal cell carcinoma in Canadian men and women. Cancer Detect Prev 2008;32(1):7-14.

[46] Ward MH, Rusiecki JA, Lynch CF, Cantor KP. Nitrate in public water supplies and the risk of renal cell carcinoma. Cancer Causes Control 2007;18(10):1141-51.

[47] Hogervorst JG, Schouten LJ, Konings EJ, Goldbohm RA, van den Brandt PA.

Dietary acrylamide intake and the risk of renal cell, bladder, and prostate cancer. Am J

Clin Nutr 2008;87(5):1428-38.

192 [48] Benichou J, Chow WH, McLaughlin JK, Mandel JS, Fraumeni JF Jr. Population

attributable risk of renal cell cancer in Minnesota. Am J Epidemiol 1998;148(5):424-30.

[49] Sweeney C, Farrow DC, Schwartz SM, Eaton DL, Checkoway H, Vaughan TL.

Glutathione S-transferase M1, T1, and P1 polymorphisms as risk factors for renal cell

carcinoma: a case-control study. Cancer Epidemiol Biomarkers Prev 2000;9(4):449-54.

[50] Wszolek MF, Wotkowicz C, Libertino JA. Surgical management of large renal

tumors. Nat Clin Pract Urol 2008;5(1):35-46.

[51] McLaughlin JK, Lindblad P, Mellemgaard A, McCredie M, Mandel JS, Schlehofer

B, et al. International renal-cell cancer study. I. Tobacco use. Int J Cancer

1995;60(2):194-8.

[52] Hunt JD, van der Hel OL, McMillan GP, Boffetta P, Brennan P. Renal cell carcinoma in relation to cigarette smoking: meta-analysis of 24 studies. Int J Cancer

2005;114(1):101-8.

[53] Wolk A, Gridley G, Niwa S, Lindblad P, McCredie M, Mellemgaard A , et al.

International renal-cell cancer study. VII. Role of diet. Int J Cancer 1996;65(1):67-73.

[54] Bergstrom A, Hsieh CC, Lindblad P, Lu CM, Cook NR, Wolk A. Obesity and renal cell cancer--a quantitative review. Br J Cancer 2001;85(7):984-90.

193

[55] Bjorge T, Tretli S, Engeland A. Relation of height and body mass index to renal cell carcinoma in two million Norwegian men and women. Am J Epidemiol

2004;160(12):1168-76.

[56] Gago-Dominguez M, Castelao JE, Yuan JM, Ross RK, Yu MC. Lipid peroxidation: a novel and unifying concept of the etiology of renal cell carcinoma (United States).

Cancer Causes Control 2002;13(3):287-93.

[57] Calle EE, Kaaks R. Overweight, obesity and cancer: epidemiological evidence and proposed mechanisms. Nat Rev 2004;4(8):579-91.

[58] Dhote R, Thiounn N, Debre B, Vidal-Trecan G. Risk factors for adult renal cell carcinoma. Urol Clin North Am 2004;31(2):237-47.

[59] Mellemgaard A, Niwa S, Mehl ES, Engholm G, McLaughlin JK, Olsen JH. Risk factors for renal cell carcinoma in Denmark: Role of medication and medical history. Int

J Epidemiol 1994;23(5):923-30.

[60] Schlehofer B, Pommer W, Mellemgaard A, Stewart JH, McCredie M, Niwa S, et al.

International renal cell cancer study. VI. the role of medical and family history. Int J

Cancer 1996;66(6):723-6.

194 [61] Chow WH, McLaughlin JK, Mandel JS, Wacholder S, Niwa S, Fraumeni JF Jr. Risk

of renal cell cancer in relation to diuretics, antihypertensive drugs and hypertension.

Cancer Epidemiol Biomarkers Prev 1995;4(4):327-31.

[62] McLaughlin JK, Chow WH, Mandel JS, Mellemgaard A, McCredie M, Lindblad P,

et al. International renal-cell cancer study. VII. Role of diuretics, other anti-hypertensive

medications and hypertension. Int J Cancer 1995;63(2):216-21.

[63] Shapiro JA, Williams MA, Weiss NS, Stergachis A, LaCroix AZ, Barlow WE.

Hypertension, antihypertensive medication use, and risk of renal cell carcinoma. Am J

Epidemiol 1999:149(6):521-30.

[64] Fraser GE, Phillips RL, Beeson WL. Hypertension, antihypertensive medication and

risk of renal carcinoma in California Seven-Day Adventists. Int J Epidemiol

1990;19(4):832-8.

[65] Heath CW, Lally CA, Calle EE, McLaughlin JK, Thun MJ. Hypertension, diuretics, and antihypertensive medication as posible risk factors for renal cell cancer. Am J

Epidemiol 1997;145(7):607-13.

[66] Coughlin SS, Neaton JD, Randall B, Sengupta A. Predictors of mortality from kidney cancer in 332,547 men screened for the Multiple Risk Factor Intervention Trial.

Cancer 1997;79(11):2171-7.

195

[67] van den Berg H. Bioavailability of vitamin D. Eur J Clin Nutr 1997;51 Suppl 1: S76-

9.

[68] Cranney A, Horsley T, O'Donnell S, Weiler H, Puil L, Ooi D, et al. Effectiveness

and safety of vitamin D in relation to bone health. Evid Rep Technol Assess 2007;(158):

1-235.

[69] John EM, Koo J, Schwartz GG. Sun exposure and prostate cancer risk: evidence for a protective effect of early-life exposure. Cancer Epidemiol Biomarkers Prev

2007;16(6):1283-6.

[70] Grant WB, Garland CF. Evidence supporting the role of vitamin D in reducing the risk of cancer. J Intern Med 2002;252(2):178-9.

[71] Zittermann A. Vitamin D in preventive medicine: are we ignoring the evidence? Br J

Nutr 2003;89(5):552-72.

[72] Wei MY, Garland CF, Gorham ED, Mohr SB, Giovannucci E. Vitamin D and prevention of colorectal adenoma: a meta-analysis. Cancer Epidemioll Biomarkers Prev

2008;17(11)2958-69.

[73] Trump DL, Hershberger PA, Bernardi RJ, Ahmed S, Muindi J, Fakih M, et al. Anti-

196 tumor activity of calcitriol: pre-clinical and clinical studies. J Steroid Biochem Mol Biol

2004;89-90(1-5):519-26.

[74] Ordonez-Moran P, Larriba MJ, Pendas-Franco N, Aguilera O, Gonzalez-Sancho JM,

Munoz A. Vitamin D and cancer: an update of in vitro and in vivo data. Front Biosci

2005;10:2723-49.

[75] Valdivielso JM, Fernandez E. Vitamin D receptor polymorphisms and diseases. Clin

Chim Acta 2006;371(1-2):1-12.

[76] Dietary Supplement Fact Sheet: Vitamin D. National Institutes of Health. Available

at: http://ods.od.nih.gov/factsheets/vitamind.asp Last Accessed on December 19, 2008.

[77] John EM, Schwartz GG, Koo J, Wang W, Ingles SA. Sun exposure, vitamin D

receptor gene polymorphisms, and breast cancer risk in a multiethnic population. Am J

Epidemiol 2007;166(12):1409-19.

[78] Norman AW. Sunlight, season, skin pigmentation, vitamin D, and 25-

hydroxyvitamin D: integral components of the vitamin D endocrine system. Am J Clin

Nutr 1998;67(6):1108-10.

[79] Holick MF, Matsuoka LY, Wortsman J. Age, vitamin D, and solar ultraviolet

radiation. Lancet 1989:2(8671):1104-5.

197

[80] Lips P. Vitamin D deficiency and secondary hyperparathyroidism in the elderly:

consequences for bone loss and fractures and therapeutic implications. Endocrine Rev

2001:22(4);477-501.

[81] Holick MF. Sunlight and vitamin D for bone health and prevention of autoimmune diseases, cancers, and cardiovascular disease. Am J Clin Nutr 2004;80(6 Suppl):1678S-

88S.

[82] Palm MD, O'Donoghue MN. Update on photoprotection. Dermatol Ther

2007;20(5):360-76.

[83] Sayre RM, Dowdy JC. Darkness at Noon: Sunscreens and Vitamin D 3. Photochem

Photobiol 2007;83(2):459-63.

[84] Matsuoka LY, Wortsman J, Haddad JG, Kolm P, Hollis BW. Racial pigmentation

and the cutaneous synthesis of vitamin D. Arch Dermatol 1991;127(4):536-8.

[85] Bendich A, Langseth L. Safety of vitamin A. J Clin Nutr 1989;49(2):358-71.

[86] Deeb KK, Trump DL, Johnson CS. Vitamin D signaling pathways in cancer:

potential for anticancer therapeutics. Nat Rev Cancer 2007;7(9):684-700.

198 [87] Zehnder D, Bland R, Williams MC, McNinch RW, Howie AJ, Stewart PM, et al.

External expression of 25-hydroxyvitamin d(3)-1 alpha-hydroxylase. J Clin Endocrinol

Metab 2001;86(2):888-94.

[88] Holick MF. Vitamin D deficiency. N Engl J Med 2007;357(3):266-81.

[89] Holick MF. Vitamin D Status: Measurement, Interpretation, and Clinical

Application. Ann Epidemiol 2008; Mar 8. [Epub ahead of print]

[90] Apperly FL. The relation of solar radiation to cancer mortality in North America.

Cancer Res 1941;1:191-5.

[91] Kricker A, Armstrong B. Does sunlight have a beneficial influence on certain cancers? Prog Biophys Mol Biol 2006;92(1):132-9.

[92] Garland CF, Garland FC. Do sunlight and vitamin D reduce the likelihood of colon cancer. Int J Epidemiol 1980:9(3);227-31.

[93] Grant WB. An ecologic study of dietary and solar ultraviolet-B links to breast carcinoma mortality rates. Cancer 2002;94(1):272-81.

[94] Grant WB. An estimate of premature cancer mortality in the United States due to inadequate doses of solar ultraviolet-B radiation. Cancer 2002;94(6):1867-75.

199

[95] Grant WB. Ecologic studies of solar UV-B radiation and cancer mortality rates.

Recent Results Cancer Res 2003;164:371-7.

[96] Boscoe FP, Schymura MJ . Solar ultraviolet-B exposure and cancer incidence and mortality in the United States, 1993-2002. BMC Cancer 2006;6:264.

[97] Mohr SB, Gorham ED, Garland CF, Grant WB, Garland FC . Are low ultraviolet B and high animal protein intake associated with risk of renal cancer? Int J Cancer

2006;119(11):2705-9.

[98] Håkansson N, Floderus B, Gustavsson P, Feychting M, Hallin N. Occupational sunlight exposure and cancer incidence among Swedish construction workers.

Epidemiology 2001;12(5):552-7.

[99] Tuohimaa P, Pukkala E, Scélo G, Olsen JH, Brewster DH, Hemminki K, et al. Does solar exposure, as indicated by the non-melanoma skin cancers, protect from solid cancers: vitamin D as a possible explanation. Eur J Cancer 2007;43(11):1701-12.

[100] Garland CF, Garland FC. Do sunlight and vitamin D reduce the likelihood of colon cancer? Int J Epidemiol 2006;35(2):217-20.

[101] Grant WB, Garland CF. The association of solar ultraviolet B (UVB) with reducing

200 risk of cancer: multifactorial ecologic analysis of geographic variation in age-adjusted cancer mortality rates. Anticancer Res 2006;26(4a):2687-99.

[102] Emerson JC, Weiss NS. Colorectal cancer and solar radiation. Cancer Causes

Control 1992;3(1):95-9.

[103] Freedman DM, Dosemeci M, McGlynn K. Sunlight and mortality from breast, ovarian, colon, prostate, and non-melanoma skin cancer: a composite death certificate based case-control study. Occup Environ Med 2002;59(4):257-62.

[104] Kampman E, Slattery ML, Caan B, Potter JD. Calcium, vitamin D, sunshine exposure, dairy products and colon cancer risk (United States). Cancer Causes Control

2000;11(5):459-66.

[105] Slattery ML, Neuhausen SL, Hoffman M, Caan B, Curtin K, Ma KN, et al. Dietary calcium, vitamin D, VDR genotypes and colorectal cancer. Int J Cancer 2004;111(5):750-

6.

[106] Garland FC, Garland CF, Gorham ED, Young JF. Geographic variation in breast cancer mortality in the United States: a hypothesis involving exposure to solar radiation.

Prev Med 1990;19(6):614-22.

[107] Gorham ED, Garland FC, Garland CF. Sunlight and breast cancer incidence in the

201 USSR. Int J Epidemiol 1990;19(4):820-4.

[108] Laden F, Spiegelman D, Neas LM, Colditz GA, Hankinson SE, Manson JE, et al.

Geographic variation in breast cancer incidence rates in a cohort of U.S. women J Natl

Cancer Inst 1997;89(18):1373-8.

[109] John EM, Schwartz GG, Dreon DM, Koo J. Vitamin D and breast cancer risk: the

NHANES I Epidemiologic follow-up study, 1971-1975 to 1992. National Health and

Nutrition Examination Survey. Cancer Epidemiol Biomarkers Prev 1999;8(5):399-406.

[110] Schwartz GG, Hulka BS. Is vitamin D deficiency a risk factor for prostate cancer?

(Hypothesis). Anticancer Res 1990;10(5A):1307-11.

[111] Hanchette CL, Schwartz GG. Geographic patterns of prostate cancer mortality.

Evidence for a protective effect of ultraviolet radiation. Cancer 1992;70(12):2861-9.

[112] Schwartz GG, Hanchette CL. UV, latitude, and spatial trends in prostate cancer mortality: all sunlight is not the same (United States). Cancer Causes Control

2006;17(8):1091-101.

[113] John EM, Dreon DM, Koo J, Schwartz GG. Residential sunlight exposure is associated with a decreased risk of prostate cancer. J Steroid Biochem Mol Biol 2004;89-

90(1-5):549-52.

202

[114] Tseng M, Breslow RA, Graubard BI, Ziegler RG. Dairy, calcium, and vitamin D intakes and prostate cancer risk in the National Health and Nutrition Examination

Epidemiologic Follow-up Study cohort. Am J Clin Nutr 2005;81(5):1147-54.

[115] Luscombe CJ, Fryer AA, French ME, Liu S, Saxby MF, Jones PW, et al. Exposure to ultraviolet radiation: association with susceptibility and age at presentation with prostate cancer. Lancet 2001;358(9282):641-2.

[116] Bodiwala D, Luscombe CJ, French ME, Liu S, Saxby MF, Jones PW, et al.

Associations between prostate cancer susceptibility and parameters of exposure to ultraviolet radiation. Cancer Lett 2003;200(2):141-8.

[117] Bodiwala D, Luscombe CJ, French ME, Liu S, Saxby MF, Jones PW, et al.

Susceptibility to prostate cancer: studies on interactions between UVR exposure and skin type. Carcinogenesis 2003;24(4):711-7.

[118] John EM, Schwartz GG, Koo J, Van Den Berg D, Ingles SA. Sun exposure, vitamin D receptor gene polymorphisms, and risk of advanced prostate cancer. Cancer

Res 2005;65(12):5470-9.

[119] Bentham G. Association between incidence of non-Hodgkin's lymphoma and solar ultraviolet radiation in England and Wales. BMJ 1996;312(7039):1128-31.

203

[120] McMichael AJ, Giles GG. Have increases in solar ultraviolet exposure contributed

to the rise in incidence of non-Hodgkin's lymphoma? Br J Cancer 1996;73(7):945-50.

[121] Langford IH, Bentham G, McDonald AL. Mortality from non-Hodgkin lymphoma

and UV exposure in the European Community. Health Place 1998;4(4):355-64.

[122] Freedman DM, Zahm SH, Dosemeci M. Residential and occupational exposure to

sunlight and mortality from non-Hodgkin's lymphoma: composite (threefold) case-

control study. BMJ 1997;314(7092):1451-5.

[123] Hu S, Ma F, Collado-Mesa F, Kirsner RS. Ultraviolet radiation and incidence of

non-Hodgkin's lymphoma among Hispanics in the United States. Cancer Epidemiol

Biomarkers Prev 2004;13(1):59-64.

[124] Hartge P, Devesa SS, Grauman D, Fears TR, Fraumeni JF Jr. Non-Hodgkin's

lymphoma and sunlight. J Natl Cancer Inst 1996;88(5):298-300.

[125] Hughes AM, Armstrong BK, Vajdic CM, Turner J, Grulich AE, Fritschi L, et al.

Sun Exposure may protect against Non-Hodgkin Lymphoma: a case-control study. Int J

Cancer 2004;112(5):865-71.

[126] Smedby KE, Hjalgrim H, Melbye M, Torrang A, Rostgaard K, Munksgaard L, et

204 al. Ultraviolet radiation exposure and risk of malignant lymphomas. J Natl Cancer Inst

2005;97(3):199-209.

[127] Hartge P, Lim U, Freedman DM, Colt JS, Cerhan JR, Cozen W, et al. Ultraviolet

radiation, dietary vitamin D, and risk of non-Hodgkin lymphoma (United States). Cancer

Causes Control 2006;17(8):1045-52.

[128] Hughes AM, Armstrong BK, Vajdic CM, Turner J, Grulich A, Fritschi L, et al.

Pigmentary characteristics, sun sensitivity and non-Hodgkin lymphoma. Int J Cancer

2004;110(3):429-34.

[129] Purdue MP, Hartge P, Davis S, Cerhan JR, Colt JS, Cozen W, et al. Sun exposure, vitamin D receptor gene polymorphisms and risk of non-Hodgkin lymphoma. Cancer

Causes Control 2007;18(9):989-99.

[130] van Wijngaarden E, Savitz DA. Occupational sunlight exposure and mortality from non-Hodgkin lymphoma among electric utility workers. J Occup Environ Med

2001;43(6):548-53.

[131] Adami J, Gridley G, Nyren O, Dosemeci M, Linet M, Glimelius B, et al. Sunlight and non-Hodgkin's lymphoma: a population-based cohort study in Sweden. Int J Cancer

1999;80(5):641-5.

205 [132] Tavani A, Bosetti C, Franceschi S, Talamini R, Negri E, La Vecchia C.

Occupational exposure to ultraviolet radiation and risk of non-Hodgkin lymphoma Eur J

Cancer Prev 2006;15(5):453-7.

[133] Newton R, Roman E, Fear N, Carpenter L. Non-Hodgkin's lymphoma and solar ultraviolet radiation. Data are inconsistent. BMJ 1996;313(7052):298.

[134] Miligi L, Costantini AS, Veraldi A, Benvenuti A, Vineis P. Cancer and pesticides: an overview and some results of the Italian multicenter case-control study on hematolymphopoietic malignancies. Ann NY Acad Sci 2006;1076:366-77.

[135] Holick MF. Vitamin D: Its role in cancer prevention and treatment. Progress in

Biophys Mol Biol 2006;92(1):49-59.

[136] Institute of Medicine, Food and Nutrition Board. Dietary Reference Intakes:

Calcium, Phosphorus, Magnesium, Vitamin D, and Fluoride. Washington, DC: National

Academy Press, 1997. Available at: http://www.iom.edu/?id=54338 Last Accessed on

December 19, 2008.

[137] Melamed ML, Muntner P, Michos ED, Uribarri J, Weber C, Sharma J, et al. Serum

25-Hydroxyvitamin D Levels and the Prevalence of Peripheral Arterial Disease. Results from NHANES 2001 to 2004. Arterioscler Thromb Vasc Biol 2008;28(6):1179-85.

206 [138] Travera-Mendoza LE, White JH. Cell Defenses and the Sunshine Vitamin. Sci Am

2007:297(5):62-5, 68-70, 72.

[139] Wagner CL, Greer FR, and the Section on Breastfeeding and Committee on

Nutrition. Prevention of rickets and vitamin D deficiency in infants, children, and adolescents. Pediatrics 2008;122(5):1142-52.

[140] Canadian mothers and babies don't get enough vitamin D 2007 Canadian Paediatric

Society Recommendation. Available at: http://www.cps.ca/english/Media/News

Releases/2007/VitaminD.htm Last Accessed on December 19, 2008.

[141] Bosetti C, Scotti L, Maso LD, Talamini R, Montella M, Negri E, et al.

Micronutrients and the risk of renal cell cancer: a case-control study from Italy. Int J

Cancer 2007;120(4):892-6.

[142] Pritchard RS, Baron JA, Gerhardsson de Verdier M. Dietary calcium, vitamin D, and the risk of colorectal cancer in Stockholm, Sweden. Cancer Epidemiol Biomarkers

Prev 1996;5(11):897-900.

[143] La Vecchia C, Braga C, Negri E, Franceschi S, Russo A, Conti E, et al. Intake of selected micronutrients and risk of colorectal cancer. Int J Cancer 1997;73(4):525-30.

[144] Ferraroni M, La Vecchia C, D'Avanzo B, Negri E, Franceschi S, Decarli A.

207 Selected micronutrient intake and the risk of colorectal cancer. Br J Cancer

1994;70(6):1150-5.

[145] Boutron MC, Faivre J, Marteau P, Couillault C, Senesse P, Quipourt V. Calcium, phosphorus, vitamin D, dairy products and colorectal carcinogenesis: a French case- control study Br J Cancer 1996;74(1):145-51.

[146] Marcus PM, Newcomb PA. The association of calcium and vitamin D, and colon and rectal cancer in Wisconsin women. Int J Epidemiol 1998;27(5):788-93.

[147] Peters RK, Pike MC, Garabrant D, Mack TM. Diet and colon cancer in Los

Angeles County, California. Cancer Causes Control 1992;3(5):457-73.

[148] Lin J, Zhang SM, Cook NR, Manson JE, Lee IM, Buring JE. Intakes of calcium and vitamin D and risk of colorectal cancer in women. Am J Epidemiol 2005;161(8):755-

64.

[149] Kesse E, Boutron-Ruault MC, Norat T, Riboli E, Clavel-Chapelon F. Dietary calcium, phosphorus, vitamin D, dairy products and the risk of colorectal adenoma and cancer among French women of the E3N-EPIC prospective study. Int J Cancer

2005;117(1):137-44.

[150] McCullough ML, Robertson AS, Rodriguez C, Jacobs EJ, Chao A, Carolyn J, et al.

208 Calcium, vitamin D, dairy products, and risk of colorectal cancer in the Cancer

Prevention Study II Nutrition Cohort (United States). Cancer Causes Control

2003;14(1):1-12.

[151] Terry P, Baron JA, Bergkvist L, Holmberg L, Wolk A. Dietary calcium and vitamin D intake and risk of colorectal cancer: a prospective cohort study in women. Nutr

Cancer 2002;43(1):39-46.

[152] Pietinen P, Malila N, Virtanen M, Hartman TJ, Tangrea JA, Albanes D, et al. Diet and risk of colorectal cancer in a cohort of Finnish men. Cancer Causes Control

1999;10(5):387-96.

[153] Järvinen R, Knekt P, Hakulinen T, Aromaa A. Prospective study on milk products, calcium and cancers of the colon and rectum. Eur J Clin Nutr 2001;55(11):1000-7.

[154] Kearney J, Giovannucci E, Rimm EB, Ascherio A, Stampfer MJ, Colditz GA, et al.

Calcium, vitamin D, and dairy foods and the occurrence of colon cancer in men. Am J

Epidemiol 1996;143(9):907-17.

[155] Bostick RM, Potter JD, Sellers TA, McKenzie DR, Kushi LH, Folsom AR.

Relation of calcium, vitamin D, and dairy food intake to incidence of colon cancer among older women. The Iowa Women's Health Study. Am J Epidemiol 1993;137(12):1302-17.

209 [156] Heilbrun LK, Nomura A, Hankin JH, Stemmermann GN. Dietary vitamin D and calcium and risk of colorectal cancer. Lancet 1985;1(8434):925.

[157] Ishihara J, Inoue M, Iwasaki M, Sasazuki S, Tsugane S. Dietary calcium, vitamin

D, and the risk of colorectal cancer. Am J Clin Nutr 2008;88(6):1576-83.

[158] Zheng W, Anderson KE, Kushi LH, Sellers TA, Greenstein J, Hong CP, et al. A prospective cohort study of intake of calcium, vitamin D, and other micronutrients in relation to incidence of rectal cancer among postmenopausal women. Cancer Epidemiol

Biomarkers Prev 1998;7(3):221-5.

[159] Grant WB, Garland CF, Gorham ED. An estimate of cancer mortality rate reductions in Europe and the US with 1,000 IU of oral vitamin D per day. Recent Results

Cancer Res 2007;174:225-34.

[160] Hartman TJ, Albert PS, Snyder K, Slattery ML, Caan B, Paskett E, et al. The association of calcium and vitamin D with risk of colorectal adenomas. J.Nutr

2005;135(2):252-9.

[161] Shin MH, Holmes MD, Hankinson SE, Wu K, Colditz GA, Willett WC. Intake of dairy products, calcium, and vitamin d and risk of breast cancer. J Natl Cancer Inst

2002;94(17):1301-11.

210 [162] McCullough ML, Rodriguez C, Diver WR, Feigelson HS, Stevens VL, Thun MJ, et al. Dairy, calcium, and vitamin D intake and postmenopausal breast cancer risk in the

Cancer Prevention Study II Nutrition Cohort. Cancer Epidemiol Biomarkers Prev

2005;14(12):2898-904.

[163] Gissel T, Rejnmark L, Mosekilde L, Vestergaard P. Intake of vitamin D and risk of breast cancer--a meta-analysis. J Steroid Biochem Mol Biol 2008;111(3-5):195-9.

[164] Simard A, Vobecky J, Vobecky JS. Vitamin D deficiency and cancer of the breast: an unprovocative ecological hypothesis. Can J Public Health 1991;82(5):300-3.

[165] Witte JS, Ursin G, Siemiatycki J, Thompson WD, Paganini-Hill A, Haile RW. Diet and premenopausal bilateral breast cancer: a case-control study. Breast Cancer Res Treat

1997;42(3):243-51.

[166] Levi F, Pasche C, Lucchini F, La Vecchia C. Dietary intake of selected micronutrients and breast-cancer risk. Int J Cancer 2001;91(2):260-3.

[167] Frazier AL, Li L, Cho E, Willett WC, Colditz GA. Adolescent diet and risk of breast cancer. Cancer Causes Control 2004;15(1):73-82.

[168] Frazier AL, Ryan CT, Rockett H, Willett WC, Colditz GA. Adolescent diet and risk of breast cancer. Breast Cancer Res 2003;5(3):R59-64.

211

[169] Chan JM, Giovannucci E, Andersson SO, Yuen J, Adami HO, Wolk A. Dairy

products, calcium, phosphorous, vitamin D, and risk of prostate cancer (Sweden). Cancer

Causes Control 1998;9(6):559-66.

[170] Chan JM, Pietinen P, Virtanen M, Malila N, Tangrea J, Albanes D, et al. Diet and

prostate cancer risk in a cohort of smokers, with a specific focus on calcium and

phosphorus (Finland). Cancer Causes Control 2000;11(9):859-67.

[171] Kristal AR, Cohen JH, Qu P, Stanford JL. Associations of energy, fat, calcium, and vitamin D with prostate cancer risk. Cancer Epidemiol Biomarkers Prev 2002;11(8):719-

25.

[172] Tavani A, Bertuccio P, Bosetti C, Talamini R, Negri E, Franceschi S, et al. Dietary intake of calcium, vitamin D, phosphorus and the risk of prostate cancer. Eur Urol

2005;48(1):27-33.

[173] Deneo-Pellegrini H, De Stefani E, Ronco A, Mendilaharsu M. Foods, nutrients and prostate cancer: a case-control study in Uruguay. Br J Cancer 1999;80(3-4):591-7.

[174] Vlajinac HD, Marinkovic JM, Ilic MD, Kocev NI. Diet and prostate cancer: a case- control study. Eur J Cancer 1997; 33(1):101-7.

212 [175] Berndt SI, Carter HB, Landis PK, Tucker KL, Hsieh LJ, Metter EJ, et al. Calcium

intake and prostate cancer risk in a long-term aging study: the Baltimore Longitudinal

Study of Aging. Urology 2002;60(6):1118-23.

[176] Giovannucci E, Rimm EB, Wolk A, Ascherio A, Stampfer MJ, Colditz GA, et al.

Calcium and fructose intake in relation to risk of prostate cancer. Cancer Res

1998;58(3):442-7.

[177] Rodriguez C, McCullough ML, Mondul AM, Jacobs EJ, Fakhrabadi-Shokoohi D,

Giovannucci EL, et al. Calcium, dairy products, and risk of prostate cancer in a

prospective cohort of United States men. Cancer Epidemiol Biomarkers Prev

2003;12(3):597-603.

[178] Genkinger JM, Hunter DJ, Spiegelman D, Anderson KE, Arslan A, Beeson WL, et

al. Dairy products and ovarian cancer: a pooled analysis of 12 cohort studies. Cancer

Epidemiol Biomarkers Prev 2006;15(2):364-72.

[179] Cramer DW, Kuper H, Harlow BL, Titus-Ernstoff L. Carotenoids, antioxidants and ovarian cancer risk in pre- and postmenopausal women. Int J Cancer 2001;94(1):128-34.

[180] Goodman MT, Wu AH, Tung KH, McDuffie K, Kolonel LN, Nomura AM, et al.

Association of dairy products, lactose, and calcium with the risk of ovarian cancer. Am J

Epidemiol 2002;156(2):148-57.

213

[181] Bidoli E, La Vecchia C, Talamini R, Negri E, Parpinel M, Conti E, et al.

Micronutrients and ovarian cancer: a case-control study in Italy. Ann Oncol

2001;12(11):1589-93.

[182] Huncharek M, Muscat J, Kupelnick B. Dairy products, dietary calcium and vitamin

D intake as risk factors for prostate cancer: a meta-analysis of 26,769 cases from 45 observational studies. Nutr Cancer 2008;60(4):421-41.

[183] Beer T, Myrthue A. Calcitriol in the treatment of prostate cancer. Anticancer Res

2006;26(4A):2647-51.

[184] Salazar-Martinez E, Lazcano-Ponce EC, Gonzalez Lira-Lira G, Escudero-De los

RP, Hernandez-Avila M. Nutritional determinants of epithelial ovarian cancer risk: a case-control study in Mexico. Oncology 2002;63(2):151-7.

[185] La Vecchia C, Decarli A, Negri E, Parazzini F, Gentile A, Cecchetti G, et al.

Dietary factors and the risk of epithelial ovarian cancer. J Natl Cancer Inst

1987;79(4):663-9.

[186] Bosetti C, Negri E, Franceschi S, Pelucchi C, Talamini R, Montella M, et al. Diet and ovarian cancer risk: a case-control study in Italy. Int J Cancer 2001;93(6):911-5.

214 [187] Fernandez E, Chatenoud L, La Vecchia C, Negri E, Franceschi S. Fish

consumption and cancer risk. Am J Clin Nutr 1999;70(1):85-90.

[188] Zheng T, Holford TR, Leaderer B, Zhang Y, Zahm SH, Flynn S, et al. Diet and

nutrient intakes and risk of non-Hodgkin's lymphoma in Connecticut women. Am J

Epidemiol 2004;159(5):454-66.

[189] Polesel J, Talamini R, Montella M, Parpinel M, Dal Maso L, Crispo A, et al.

Linoleic acid, vitamin D and other nutrient intakes in the risk of non-Hodgkin lymphoma:

an Italian case-control study. Ann Oncol 2006;17(4):713-8.

[190] Chang ET, Balter KM, Torrang A, Smedby KE, Melbye M, Sundstrom C, et al.

Nutrient intake and risk of non-Hodgkin's lymphoma. Am J Epidemiol

2006;164(12):1222-32.

[191] Fritschi L, Ambrosini GL, Kliewer EV, Johnson KC. Dietary fish intake and risk of leukemia, multiple myeloma, and non-Hodgkin lymphoma. Cancer Epidemiol

Biomarkers Prev 2004;13(4):532-7.

[192] Walters MR. Newly Identified actions of the vitamin D endocrine system. Endocr

Rev 1992;13(4):719-64.

[193] Thibault F, Cancel-Tassin G, Cussenot O. Low penetrance genetic susceptibility to

215 kidney cancer. Br J Urul Int 2006;98(4):735-8.

[194] Slattery ML. Vitamin D receptor gene (VDR) associations with cancer. Nutr Rev

2007;65(8 Pt 2):S102-4

[195] Dusso AS, Brown AJ, Slatopolsky E. Vitamin D. Am J Physiol Renal Physiol

2005;289(1):F8-28.

[196] Arai H, Miyamoto K, Taketani Y, Yamamoto H, Iemori Y, Morita K, et al. A vitamin D receptor gene polymorphism in the translation initiation codon: effect on protein activity and relation to bone mineral density in Japanese women. J Bone Miner

Res 1997;12(6):915-21.

[197] Jurutka PW, Remus LS, Whitfield GK, Thompson PD, Hsieh JC, Zitzer H, et al.

The polymorphic N terminus in human vitamin D receptor isoforms influences transcriptional activity by modulating interaction with transcription factor IIB. Mol

Endocrinol 2000;14(3):401-20.

[198] Huang SP, Huang CY, Wu WJ, Pu YS, Chen J, Chen YY, et al. Association of vitamin D receptor FokI polymorphism with prostate cancer risk, clinicopathological features and recurrence of prostate specific antigen after radical prostatectomy. Int J

Cancer 2006;119(8):1902-7.

216 [199] Liu Z, Calderon JI, Zhang Z, Sturgis E, Spitz MR, Wei Q. Polymorphisms of vitamin D receptor gene protect against the risk of head and neck cancer. Pharmacogenet

Genomics 2005;15(3):159-65.

[200] Uitterlinden AG, Fang Y, van Meurs JB, Pols HA, Van Leeuwen JP. Genetics and biology of vitamin D receptor polymorphisms. Gene 2004;338(2):143-56.

[201] Karami S, Brennan P, Hung RJ, Boffetta P, Toro J, Wilson RT, et al. Vitamin D receptor polymorphisms and renal cancer risk in Central and Eastern Europe. J Toxicol

Environ Health A. 2008;71(6):367-72.

[202] Obara W, Suzuki Y, Kato K, Tanji S, Konda R, Fujioka T. Vitamin D receptor gene polymorphisms are associated with increased risk and progression of renal cell carcinoma in a Japanese population. Int J Urol 2007;14(6):483-7.

[203] Ikuyama T, Hamasaki T, Inatomi H, Katoh T, Muratani T, Matsumoto T.

Association of vitamin D receptor gene polymorphism with renal cell carcinoma in

Japanese. Endocr J 2002;49(4):433-8.

[204] Medeiros R, Morais A, Vasconcelos A, Costa S, Pinto D, Oliveira J, et al. The role of vitamin D receptor gene polymorphisms in the susceptibility to prostate cancer of a southern European population. J Hum Genet 2002;47(8):413-8.

217 [205] Hamasaki T, Inatomi H, Katoh T, Ikuyama T, Matsumoto T. Significance of vitamin D receptor gene polymorphism for risk and disease severity of prostate cancer and benign prostatic hyperplasia in Japanese. Urol Int 2002;68(4):226-31.

[206] Gsur A, Madersbacher S, Haidinger G, Schatzl G, Marberger M, Vutuc C, et al.

Vitamin D receptor gene polymorphism and prostate cancer risk. Prostate 2002;51(1):30-

4.

[207] Hamasaki T, Inatomi H, Katoh T, Ikuyama T, Matsumoto T. Clinical and pathological significance of vitamin D receptor gene polymorphism for prostate cancer which is associated with a higher mortality in Japanese. Endocr J 2001;48(5):543-9.

[208] Mishra DK, Bid HK, Srivastava DS, Mandhani A, Mittal RD. Association of vitamin D receptor gene polymorphism and risk of prostate cancer in India. Urol Int

2005;74(4):315-8.

[209] Yang Y, Wang S, Ye Z, Yang W. Association of single nucleotide polymorphism of vitamin D receptor gene start codon and the susceptibility to prostate cancer in the Han nationality in Hubei area. Zhonghua Nan Ke Xue 2004;10(6):411-4.

[210] Tayeb MT, Clark C, Haites NE, Sharp L, Murray GI, McLeod HL. Vitamin D receptor, HER-2 polymorphisms and risk of prostate cancer in men with benign prostate hyperplasia. Saudi Med J 2004;25(4):447-51

218

[211] Cheteri MB, Stanford JL, Friedrichsen DM, Peters MA, Iwasaki L, Langlois MC, et al. Vitamin D receptor gene polymorphisms and prostate cancer risk. Prostate

2004;59(4):409-18.

[212] Bodiwala D, Luscombe CJ, French ME, Liu S, Saxby MF, Jones PW, et al.

Polymorphisms in the vitamin D receptor gene, ultraviolet radiation, and susceptibility to prostate cancer. Environ Mol Mutagen 2004;43(2):121-7.

[213] Chokkalingam AP, McGlynn KA, Gao YT, Pollak M, Deng J, Sesterhenn IA, et al.

Vitamin D receptor gene polymorphisms, insulin-like growth factors, and prostate cancer risk: a population-based case-control study in China. Cancer Res 2001;61(11):4333-6.

[214] Tayeb MT, Clark C, Haites NE, Sharp L, Murray GI, McLeod HL. CYP3A4 and

VDR gene polymorphisms and the risk of prostate cancer in men with benign prostate hyperplasia. Br J Cancer 2003;88(6):928-32.

[215] Bousema JT, Bussemakers MJ, van Houwelingen KP, Debruyne FM, Verbeek AL, de La Rosette JJ, et al. Polymorphisms in the vitamin D receptor gene and the gene and the risk of benign prostatic hyperplasia. Eur Urol 2000;37(2):234-8.

[216] Andersson P, Varenhorst E, Söderkvist P. Androgen receptor and vitamin D receptor gene polymorphisms and prostate cancer risk. Eur J Cancer 2006;42(16):2833-7.

219

[217] Oakley-Girvan I, Feldman D, Eccleshall TR, Gallagher RP, Wu AH, Kolonel LN, et al. Risk of early-onset prostate cancer in relation to germ line polymorphisms of the vitamin D receptor. Cancer Epidemiol Biomarkers Prev 2004;13(8):1325-30.

[218] Mikhak B, Hunter DJ, Spiegelman D, Platz EA, Hollis BW, Giovannucci E.

Vitamin D receptor (VDR) gene polymorphisms and haplotypes, interactions with plasma

25-hydroxyvitamin D and 1,25-dihydroxyvitamin D, and prostate cancer risk. Prostate

2007;67(9):911-23.

[219] Cicek MS, Liu X, Schumacher FR, Casey G, Witte JS. Vitamin D receptor genotypes/haplotypes and prostate cancer risk. Cancer Epidemiol Biomarkers Prev

2006;15(12):2549-52.

[220] Li H, Stampfer MJ, Hollis JB, Mucci LA, Gaziano JM, Hunter D, et al. A prospective study of plasma vitamin D metabolites, vitamin D receptor polymorphisms, and prostate cancer. PLoS Med 2007;4(3):e103.

[221] Watanabe M, Fukutome K, Murata M, Uemura H, Kubota Y, Kawamura J, et al.

Significance of vitamin D receptor gene polymorphism for prostate cancer risk in

Japanese. Anticancer Res 1999;19(5C):4511-4.

[222] Blazer DG 3rd, Umbach DM, Bostick RM, Taylor JA. Vitamin D receptor

220 polymorphisms and prostate cancer. Mol Carcinog 2000;27(1):18-23.

[223] Chaimuangraj S, Thammachoti R, Ongphiphadhanakul B, Thammavit W. Lack of association of VDR polymorphisms with Thai prostate cancer as compared with benign prostate hyperplasia and controls. Asian Pac J Cancer Prev 2006;7(1):136-9.

[224] Maistro S, Snitcovsky I, Sarkis AS, da Silva IA, Brentani MM. Vitamin D receptor polymorphisms and prostate cancer risk in Brazilian men. Int J Biol Markers

2004;19(3):245-9.

[225] Huang SP, Chou YH, Wayne Chang WS, Wu MT, Chen YY, Yu CC, et al.

Association between vitamin D receptor polymorphisms and prostate cancer risk in a

Taiwanese population. Cancer Lett 2004;207(1):69-77

[226] Suzuki K, Matsui H, Ohtake N, Nakata S, Takei T, Koike H, et al. Vitamin D receptor gene polymorphism in familial prostate cancer in a Japanese population. Int J

Urol 2003;10(5):261-6.

[227] Habuchi T, Suzuki T, Sasaki R, Wang L, Sato K, Satoh S, et al. Association of vitamin D receptor gene polymorphism with prostate cancer and benign prostatic hyperplasia in a Japanese population. Cancer Res 2000;60(2):305-8.

[228] Sillanpää P, Hirvonen A, Kataja V, Eskelinen M, Kosma VM, Uusitupa M, et al.

221 Vitamin D receptor gene polymorphism as an important modifier of positive family history related breast cancer risk. Pharmacogenetics 2004;14(4):239-45

[229] Hou MF, Tien YC, Lin GT, Chen CJ, Liu CS, Lin SY, et al. Association of vitamin

D receptor gene polymorphism with sporadic breast cancer in Taiwanese patients. Breast

Cancer Res Treat 2002;74(1):1-7.

[230] Trabert B, Malone KE, Daling JR, Doody DR, Bernstein L, Ursin G, et al. Vitamin

D receptor polymorphisms and breast cancer risk in a large population-based case-control study of Caucasian and African-American women. Breast Cancer Res 2007;9(6):R84.

[231] Lowe LC, Guy M, Mansi JL, Peckitt C, Bliss J, Wilson RG, et al. Plasma 25- hydroxy vitamin D concentrations, vitamin D receptor genotype and breast cancer risk in a UK Caucasian population. Eur J Cancer 2005;41(8):1164-9.

[232] Bretherton-Watt D, Given-Wilson R, Mansi JL, Thomas V, Carter N, Colston KW.

Vitamin D receptor gene polymorphisms are associated with breast cancer risk in a UK

Caucasian population. Br J Cancer 2001;85(2):171-5.

[233] Ruggiero M, Pacini S, Aterini S, Fallai C, Ruggiero C, Pacini P. Vitamin D receptor gene polymorphism is associated with metastatic breast cancer. Oncol Res

1998;10(1):43-6.

222 [234] Guy M, Lowe LC, Bretherton-Watt D, Mansi JL, Peckitt C, Bliss J, et al. Vitamin

D receptor gene polymorphisms and breast cancer risk. Clin Cancer Res

2004;10(16):5472-81.

[235] Ingles SA, Garcia DG, Wang W, Nieters A, Henderson BE, Kolonel LN, et al.

Vitamin D receptor genotype and breast cancer in Latinas (United States). Cancer Causes

Control 2000;11(1):25-30.

[236] McCullough ML, Stevens VL, Diver WR, Feigelson HS, Rodriguez C, Bostick

RM, et al. Vitamin D pathway gene polymorphisms, diet, and risk of postmenopausal breast cancer: a nested case-control study. Breast Cancer Res 2007;9(1):R9.

[237] Curran JE, Vaughan T, Lea RA, Weinstein SR, Morrison NA, Griffiths LR.

Association of A vitamin D receptor polymorphism with sporadic breast cancer development. Int J Cancer 1999;83(6):723-6.

[238] Abbas S, Nieters A, Linseisen J, Slanger T, Kropp S, Mutschelknauss EJ, et al.

Vitamin D receptor gene polymorphisms and haplotypes and postmenopausal breast cancer risk. Breast Cancer Res 2008;10(2):R31.

[239] Lundin AC, Söderkvist P, Eriksson B, Bergman-Jungeström M, Wingren S.

Association of breast cancer progression with a vitamin D receptor gene polymorphism.

South-East Sweden Breast Cancer Group. Cancer Res 1999;59(10):2332-4.

223

[240] Chen WY, Bertone-Johnson ER, Hunter DJ, Willett WC, Hankinson SE.

Associations between polymorphisms in the vitamin D receptor and breast cancer risk.

Cancer Epidemiol Biomarkers Prev 2005;14(10):2335-9.

[241] Agoston ES, Hatcher MA, Kensler TW, Posner GH. Vitamin D analogs as anti- carcinogenic agents. Anticancer Agents Med Chem 2006;6(1):53-71.

[242] Yang F, Bergeron JM, Linehan LA, Lalley PA, Sakaguchi AY, Bowman BH.

Mapping and conservation of the group-specific component gene in mouse. Genomics

1990;7(4):509-16.

[243] Verboven C, Rabijns A, De Maeyer M, Van Baelen H, Bouillon R, De Ranter C. A structural basis for the unique binding features of the human vitamin D-binding protein.

Nat Struct Biol 2002;9(2):131-6.

[244] Negri AL. Proximal tubule endocytic apparatus as the specific renal uptake mechanism for vitamin D-binding protein/25-(OH)D3 complex. Nephrology (Carlton)

2006;11(6):510-5.

[245] Ray R. Molecular recognition in vitamin D-binding protein. Proc Soc Exp Biol

Med 1996;212(4):305-12.

224 [246] Willnow TE, Nykjaer A. Pathways for kidney-specific uptake of the steroid hormone 25-hydroxyvitamin D3. Curr Opin Lipidol 2002;13(3):255-60.

[247] Fröhlander N, Ljungberg B. Serum protein groups in renal cell carcinoma. Hum

Hered 1986;36(2):119-22.

[248] Germenis A, Dimopoulos MA, Fertakis A, Dimopoulos C. Genetic markers in renal adenocarcinoma. J Urol 1984;132(1):173-4.

[249] Sawada N, Kusudo T, Sakaki T, Hatakeyama S, Hanada M, Abe D, et al. Novel metabolism of 1 alpha,25-dihydroxyvitamin D3 with C24-C25 bond cleavage catalyzed by human CYP24A1. Biochemistry 2004;43(15):4530-7.

[250] Väisänen S, Dunlop TW, Sinkkonen L, Frank C, Carlberg C. Spatio-temporal activation of chromatin on the human CYP24 gene promoter in the presence of

1alpha,25-Dihydroxyvitamin D3. J Mol Biol 2005;350(1):65-77.

[251] Masuda S, Strugnell SA, Knutson JC, St-Arnaud R, Jones G. Evidence for the activation of 1alpha-hydroxyvitamin D2 by 25-hydroxyvitamin D-24-hydroxylase: delineation of pathways involving 1alpha,24-dihydroxyvitamin D2 and 1alpha,25- dihydroxyvitamin D2. Biochim Biophys Acta 2006;1761(2):221-34.

[252] Tissandié E, Guéguen Y, Lobaccaro JM, Aigueperse J, Souidi M. Vitamin D:

225 metabolism, regulation and associated diseases. Med Sci (Paris) 2006;22(12):1095-100.

[253] Lin R, White JH. The pleiotropic actions of vitamin D. BioEssays 2003;26(1):21-8.

[254] Brown AJ, Dusso A, Slatopolsky E. Vitamin D. Am J Physiol Renal Physiol

1999;227(2 Pt2):F157-75.

[255] Madej A, Puzianowska-Kuznicka M, Tanski Z, Nauman J, Nauman A. Vitamin D

receptor binding to DNA is altered without the change in its expression in human renal

clear cell cancer. Nephron Exp Nephrol 2003;93(4):150-7.

[256] Nagakura K, Hayakawa M, Hata M, Nakamura H. 1,25-Dihydroxyvitamin D3

receptors and their relationship to histological features in renal cell carcinoma. J Urol

1987;137(6):1300-3.

[257] Buentig N, Stoerkel S, Richter E, Dallmann I, Reitz M, Atzpodien J. Predictive

impact of retinoid X receptor-alpha-expression in renal-cell carcinoma. Cancer Biother

Radiopharm 2004;19(3):331-42.

[258] Skoromny ĭ NA, Beketov AI, Vasil'ev KK, Kolbasin PN. Morphofunctional changes in the vascular system of the rabbit brain in rocking during anti-orthostatic position and administration of dimephosphon. Kosm Biol Aviakosm Med 1991;25(3):31-

5.

226

[259] Obara W, Konda R, Akasaka S, Nakamura S, Sugawara A, Fujioka T. Prognostic significance of vitamin D receptor and retinoid X receptor expression in renal cell carcinoma. J Urol 2007;178(4 Pt 1):1497-503.

[260] Rachez C, Lemon BD, Suldan Z, Bromleigh V, Gamble M, Näär AM, et al.

Ligand-dependent transcription activation by nuclear receptors requires the DRIP complex. Nature 1999;398(6730):824-8.

[261] Rachez C, Suldan Z, Ward J, Chang CP, Burakov D, Erdjument-Bromage H, et al.

A novel protein complex that interacts with the vitamin D3 receptor in a ligand- dependent manner and enhances VDR transactivation in a cell-free system. Genes Dev

1998;12(12):1787-800.

[262] Puzianowska-Kuznicka M, Nauman A, Madej A, Tanski Z, Cheng S, Nauman J.

Expression of thyroid hormone receptors is disturbed in human renal clear cell carcinoma. Cancer Lett 2000;155(2):145-52.

[263] Kamiya Y, Puzianowska-Kuznicka M, McPhie P, Nauman J, Cheng SY, Nauman

A. Expression of mutant thyroid hormone nuclear receptors is associated with human renal clear cell carcinoma. Carcinogenesis 2002;23(1):25-33.

[264] Vidal M, Ramana CV, Dusso AS. Stat1-vitamin D receptor interactions antagonize

227 1,25-dihydroxyvitamin D transcriptional activity and enhance stat1-mediated transcription. Mol Cell Biol 2002;22(8):2777-87.

[265] Matsuzaki J, Tsuji T, Zhang Y, Wakita D, Imazeki I, Sakai T, et al. 1alpha,25-

Dihydroxyvitamin D3 downmodulates the functional differentiation of Th1 cytokine- conditioned bone marrow-derived dendritic cells beneficial for cytotoxic T lymphocyte generation. Cancer Sci 2006;97(2):139-47.

[266] Muthian G, Raikwar HP, Rajasingh J, Bright JJ. 1,25 Dihydroxyvitamin-D3 modulates JAK-STAT pathway in IL-12/IFNgamma axis leading to Th1 response in experimental allergic encephalomyelitis. J Neurosci Res 2006;83(7):1299-309.

[267] Afaq F, Adhami VM, Mukhtar H. Photochemoprevention of ultraviolet B signaling and photocarcinogenesis. Mutat Res 2005;571(1-2):153-73.

[268] Reichrath J. The challenge resulting from positive and negative effects of sunlight: how much solar UV exposure is appropriate to balance between risks of vitamin D deficiency and skin cancer? Prog Biophys Mol Biol 2006;92(1):9-16.

[269] Kreiter SR, Schwartz RP, Kirkman HN Jr, Charlton PA, Calikoglu AS, Davenport

ML. Nutritional rickets in African American breast-fed infants. J Pediatr

2000;137(2):153-7.

228 [270] Pugliese MT, Blumberg DL, Hludzinski J, Kay S. Nutritional rickets in suburbia. J

Am Coll Nutr 1998;17(6):637-41.

[271] Sills IN, Skuza KA, Horlick MN, Schwartz MS, Rapaport R. Vitamin D deficiency rickets. Reports of its demise are exaggerated. Clin Pediatr (Phila)1994;33(8):491-3.

[272] Calvo MS, Whiting SJ, Barton CN. Vitamin D intake: a global perspective of current status. J Nutr 2005;135(2):310-6.

[273] Kimball S, Fuleihan Gel-H, Vieth R. Vitamin D: a growing perspective. Crit Rev

Clin Lab Sci 2008;45(4):339-414.

[274] Berk M, Sanders KM, Pasco JA, Jacka FN, Williams LJ, Hayles AL, et al. Vitamin

D deficiency may play a role in depression. Med Hypotheses 2007;69(6):1316-9.

[275] Schwalfenberg G. Not enough vitamin D: health consequences for Canadians. Can

Fam Physician 2007;53(5):841-54.

[276] Souberbielle JC, Friedlander G, Kahan A, Cormier C. Evaluating vitamin D status.

Implications for preventing and managing osteoporosis and other chronic diseases. Joint

Bone Spine 2006;73(3):249-53.

[277] Klassen CD, Watkins III JB. Casarett & Doull's Essentials of Toxicology. NY,

229 McGraw Hill, 6th ed., 2001. (p208-19).

[278] Scélo G, Constantinescu V, Csiki I, Zaridze D, Szeszenia-Dabrowska N, Rudnai P, et al. Occupational exposure to vinyl chloride, acrylonitrile and styrene and lung cancer risk. Cancer Causes Control 2004;15(5):445-52.

[279] Hashibe M, Boffetta P, Zaridze D, Shangina O, Szeszenia-Dabrowska N, Mates D, et al. Contribution of tobacco and alcohol to the high rates of squamous cell carcinoma of the supraglottis and glottis in Central Europe. Am J Epidemiol 2007;165(7):814-20.

[280] ILO. International standard classification of occupations (ISCO). Geneva, 1968 (Rev.

Ed.)

[281] Durusoy R, Boffetta P, Mannetje A, Zaridze D, Szeszenia-Dabrowska N, Rudnai P, et al. Lung cancer risk and occupational exposure to meat and live animals. Int J Cancer

2006;118(10):2543-7.

[282] Eurostat. NACE Rev, 1: statistical classification of economic activities in the

European community. Luxembourg, 1996.

[283] Mannetje A, Kromhout H. The use of occupation and industry classifications in general population studies. International Journal of Epidemiology 2003;32(2):419-28.

230 [284] Packer BR, Yeager M, Burdett L, Welch R, Beerman M, Qi L, et al.

SNP500Cancer: a public resource for sequence validation, assay development, and frequency analysis for genetic variation in candidate genes. Nucleic Acids Res

2006;34(Database issue):D617-21.

[285] Versage JL, Severin D, Chu MC, Petersen JM. Development of a Multitarget Real-

Time TaqMan PCR Assay for Enhanced Detection of Francisella tularensis in Complex

Specimens. J Clin Microbiol 2003;41(12):5492-9.

[286] Illumina ® Available at: http://www.illumina.com Last Accessed on March 14,

2008.

[287] Carlson CS, Eberle MA, Kruglyak L, Nickerson DA. Mapping complex disease loci in whole-genome association studies. Nature 2004;429(6990):446-52.

[288] Core Genotyping Facility. Available at: http://cgf.nci.nih.gov/home.cfm Last accessed on March 14, 2008.

[289] Brennan P, van der Hel O, Brennan P, Zaridze D, Matveev V, Holcatova I, et al.

Tobacco smoking, body mass index, hypertension, and kidney cancer in Central and

Eastern Europe. Br J Cancer 2008;99(11):1912-5.

[290] Haploview. Available at: http://www.broad.mit.edu/mpg/haploview/index.php Last

231 Accessed on March 14, 2008.

[291] R Project for Statistical Computing. Available at: http://www.r-project.org/ Last

Accessed on March 14, 2008.

[292] Chen BE, Sakoda LC, Hsing AW, Rosenberg PS. Resampling-based multiple hypothesis testing procedures for genetic case-control association studies. Genet

Epidemiol 2006;30(6):495-507.

[293] Rosenberg PS, Che A, Chen BE. Multiple hypothesis testing strategies for genetic case-control association studies. Stat Med 2006;25(18):3134-49.

[294] Pawitan Y, Murthy KR, Michiels S, Ploner A. Bias in the estimation of false discovery rate in microarray studies. Bioinformatics 2005;21(20):3865-72.

[295] Power Program V3.o National Cancer Institute, Division of Cancer Epidemiology

& Genetics. Available at: http://dceg.cancer.gov/tools/design/power/readme. Last

Accessed on April 2, 2008.

[296] Thorne J, Campbell MJ. The vitamin D receptor in cancer. Proc Nutr Soc

2008;67(2):115-27.

[297] Robin T Wilson. Fish intake and renal cell cancer risk in Finland. Presentation at

232 the 8 th National Tribal Conference on Environmental Management. June 24, 2008.

Available at: http://www.ntcem8.org/Robin%20Taylor%20Wilson_%20Fish%20and%20

Renal%20Cell%20Cancer_06_24_2008.ppt Last accessed on December 07, 2008.

[298] Zmuda JM, Cauley JA, Ferrell RE. Molecular epidemiology of vitamin D receptor

gene variants. Epidemiol Rev 2000;22(2):203-17.

[299] Ondková S, Macejová D, Brtko J. Role of dihydroxyvitamin D(3) and its nuclear

receptor in novel directed therapies for cancer. Gen Physiol Biophys 2006;25(4):339-53.

[300] Norman AW. From vitamin D to hormone D: fundamentals of the vitamin D

endocrine system essential for good health. Am J Clin Nutr 2008;88(2):491S-499S

[301] Holick CN, Stanford JL, Kwon EM, Ostrander EA, Nejentsev S, Peters U.

Comprehensive association analysis of the vitamin D pathway genes, VDR, CYP27B1,

and CYP24A1, in prostate cancer. Cancer Epidemiol Biomarkers Prev 2007;16(10):1990-

9.

[302] Kim HS, Newcomb PA, Ulrich CM, Keener CL, Bigler J, Farin FM , et al. Vitamin

D receptor polymorphism and the risk of colorectal adenomas: evidence of interaction

with dietary vitamin D and calcium. Cancer Epidemiol Biomarkers Prev. 2001;10(8):869-

74.

[303] Hubner RA, Muir KR, Liu JF, Logan RF, Grainge MJ, Houlston RS , et al. Dairy

233 products, polymorphisms in the vitamin D receptor gene and colorectal adenoma recurrence. Int J Cancer 2008 Aug 1;123(3):586-93.

[304] Moon S, Holley S, Bodiwala D, Luscombe CJ, French ME, Liu S, Saxby MF,

Jones PW, Fryer AA, Strange RC. Associations between G/A1229, A/G3944, T/C30875,

C/T48200 and C/T65013 genotypes and haplotypes in the vitamin D receptor gene, ultraviolet radiation and susceptibility to prostate cancer. Ann Hum Genet 2006; 70(Pt 2):

226-36.

[305] Rukin NJ, Luscombe C, Moon S, Bodiwala D, Liu S, Saxby MF, Fryer AA,

Alldersea J, Hoban PR, Strange RC. Prostate cancer susceptibility is mediated by interactions between exposure to ultraviolet radiation and polymorphisms in the 5' haplotype block of the vitamin D receptor gene. Cancer Lett 2007; 247(2): 328-35.

[306] Wacholder S, Chatterjee N, Hartge P. Joint effect of genes and environment distorted by selection biases: implications for hospital-based case-control studies. Cancer

Epidemiol Biomarkers Prev 2002;11(9):885-9.

[307] Mason BH, Holdaway IM, Skinner SJ, Stewart AW, Kay RG, Neave LM, et al.

Association between season of first detection of breast cancer and disease progression.

Breast Cancer Res Treat 1987;9(3):227-32.

234 [308] Mason BH, Holdaway IM, Stewart AW, Neave LM, Kay RG. Season of tumour detection influences factors predicting survival of patients with breast cancer. Breast

Cancer Res Treat 1990;15(1):27-37.

[309] Sankila R, Joensuu H, Pukkala E, Toikkanen S. Does the month of diagnosis affect survival of cancer patients? Br J Cancer 1993;67(4):838-41.

[310] Robsahm TE, Tretli S, Dahlback A, Moan J. Vitamin D3 from sunlight may improve the prognosis of breast-, colon- and prostate cancer (Norway). Cancer Causes

Control 2004;15(2):149-58.

[311] Porojnicu AC, Lagunova Z, Robsahm TE, Berg JP, Dahlback A, Moan J. Changes in risk of death from breast cancer with season and latitude: sun exposure and breast cancer survival in Norway. Breast Cancer Res Treat 2007;102(3)323-8.

[312] Porojnicu AC, Robsahm TE, Dahlback A, Berg JP, Christiani D, Bruland OS, et al.

Seasonal and geographical variations in lung cancer prognosis in Norway Does Vitamin

D from the sun play a role? Lung Cancer 2007;55(3):263-70.

[313] Moan J, Porojnicu AC, Robsahm TE, Dahlback A, Juzeniene A, Tretli S, et al.

Solar radiation, vitamin D and survival rate of colon cancer in Norway J Photochem

Photobiol B 2005;78(3):189-93.

235 [314] Porojnicu AC, Robsahm TE, Ree AH, Moan J. Season of diagnosis is a prognostic

factor in Hodgkin's lymphoma: a possible role of sun-induced vitamin D. Br J Cancer

2005; 93(5):571-4.

[315] Porojnicu AC, Robsahm TE, Berg JP, Moan J. Season of diagnosis is a predictor of cancer survival. Sun-induced vitamin D may be involved: a possible role of sun-induced

Vitamin D. J Steroid Biochem Mol Biol 2007;103(3-5):675-8.

[316] Autier P, Gandini S. Vitamin D supplementation and total mortality: a meta- analysis of randomized controlled trials. Arch Intern Med 2007;167(16):1730-7.

[317] Trivedi DP, Doll R, Khaw KT. Effect of four monthly oral vitamin D3

(cholecalciferol) supplementation on fractures and mortality in men and women living in the community: randomised double blind controlled trial. BMJ 2003;326(7387):469.

[318] Lappe JM, Travers-Gustafson D, Davies KM, Recker RR, Heaney RP. Vitamin D and calcium supplementation reduces cancer risk: results of a randomized trial. Am J Clin

Nutr 2007;85(6):1586-91.

[319] Dalhoff K, Dancey J, Astrup L, Skovsgaard T, Hamberg KJ, Lofts FJ, et al. A phase II study of the vitamin D analogue Seocalcitol in patients with inoperable hepatocellular carcinoma. Br J Cancer 2003;89(2):252-7.

236 [320] Beer TM, Eilers KM, Garzotto M, Egorin MJ, Lowe BA, Henner WD. Weekly

high-dose calcitriol and docetaxel in metastatic androgen-independent prostate cancer. J

Clin Oncol 2003;21(1):123-8.

[321] Schwartz GG, Hall MC, Stindt D, Patton S, Lovato J, Torti FM. Phase I/II study of

19-nor-1alpha-25-dihydroxyvitamin D2 (paricalcitol) in advanced, androgen-insensitive prostate cancer. Clin Cancer Res 2005;11(24 Pt 1):8680-5.

[322] Wactawski-Wende J, Kotchen JM, Anderson GL, Assaf AR, Brunner RL,

O’Sullivan MJ, et al. Calcium plus vitamin D supplementation and the risk of colorectal

cancer. N Engl J Med 2006;354(7):684-96.

[323] SEER Cancer Statistics Review, 1975-2005 (NCI 2008). Available at

http://seer.cancer.gov Last Accessed on December 7, 2008.

[324] Karami S, Boffetta P, Rothman N, Hung RJ, Stewart T, Zaridze D, et al. Renal cell

carcinoma, occupational pesticide exposure and modification by glutathione S-transferase

polymorphisms. Clin Cancer Res 2008;14(15):4726-34.

237 Appendix:

The results of this study indicated that occupational ultraviolet (UV) exposure was

associated with renal cell carcinoma (RCC) risk. Therefore we attempted to explore

whether risk factors associated with renal cancer risk, such as excess body mass index

(BMI), hypertension, or smoking, modified the association between occupational UV

exposure and renal cancer risk; these analyses were also stratified by sex, given that

increased occupation UV exposure was associated with decreased renal cancer risk only

among male participants in our study. Table 9 shows that among all participants BMI did

not modify the association between occupational UV exposure and RCC risk. However,

when analyses were stratified by sex (Table 10 & 11) BMI modified the association

between occupation UV exposure and renal cancer risk only among males (Table 10) for cumulative exposure (p-interaction= 0.04) and frequency-adjusted duration of exposure

(p-interaction= 0.04). With increasing UV exposure, low risk of RCC was observed among male participants in the lowest BMI category (>25 kg/m 2). These results are

biologically plausible since obese individuals have been reported to have lower levels of

circulating 25(OH) vitamin D due to reduced bioavailability [81-83] . No association

between occupational UV exposure, BMI, and RCC risk was observed among females

(Table 11).

238 1.87 0.45 0.092 1.81 1.93 2.14 3.73 0.70 0.447 1.90 0.40 0.081 1.71 0.80 0.246 ------00 00 00 % OR LCI - UCI P-trend *lrt Among Participants with 30+ BMI Participants Among 1.41 0.94 96 33.7 97 31.5 1.19 0.76 1.39 941.49 33.0 100 32.5 1.19 0.78 991.37 34.7 100 32.5 1.26 0.83 89 40.1 89 34.2 1.37 0.88 3.71 0.34 51 81.0 34 70.8 1.24 0.42 1.40 0.98 102 35.8 99 32.1 1.21 0.78 1.25 0.38 45 20.3 63 24.2 1.01 0.59 ------Among Participants with 25-29.9 BMI Participants Among s withOnlyLow Intensity Jobs (yrs) withAny High Intensity Jobs (yrs) to occupationalto sunlightby index body mass 1.33 0.55 147 32.1 208 34.4 1.00 0.70 1.56 1611.60 35.2 207 34.2 1.00 0.72 1671.58 36.5 204 33.7 1.07 0.77 154 41.2 191 38.1 0.97 0.69 5.22 0.78 70 83.3 87 83.7 1.54 0.64 1.33 0.56 153 33.4 209 34.5 1.00 0.71 1.14 0.22 77 20.6 121 24.2 0.83 0.55 ------ols tatus, and self-reported hypertension self-reported and tatus, Among Participants with <25 BMI Participants Among 6 12.5 11 14.5 1.00 14 16.7 17 16.3 1.00 12 19.0 14 29.2 1.00 118 37.7 176 34.1 1.00119 38.0 178 34.5 1.00118 44.5 175 39.8 144 1.00 31.4 189 31.2 1.00 144 31.4 193 31.9 1.00 143 38.2 189 37.7 89 1.00 31.2 109 35.4 1. 90 31.6 111 36.0 1. 88 39.6 108 41.5 1. Cases Controls Cases Controls Cases Controls 5.25 5.25 5.25 9.30 Frequency-Adjusted Duration Exposure of for Subject Frequency-Adjusted < >9.55 87 27.8>9.30 172 33.3 0.89 0.59 85 27.2>9.30 170 32.9 0.89 0.59 43 16.2 105 23.9 0.71 0.44 All values adjusted for age, sex, center, smoking s smoking sex, center, age, for adjusted All values contr levelson among exposure based values Tertile p-value test ratio *Likelihood Frequency-Adjusted Duration Exposure of(yrs) Frequency-Adjusted < Duration Exposure of for Subject Frequency-Adjusted < Sunlight(low Exposure Cumulative exposure-unit-yrs) N< % N % OR LCI - UCI P-trend N % N % OR LCI - UCI P-trend N % N >5.25-9.55 108 34.5 168 32.6>5.25-9.30 1.07 0.73 109 34.8 168 32.6>5.25-9.30 1.09 0.75 104 39.2 160 36.4 1.07 0.73 >9.30 42 87.5 65 85.5 1.23 0.29 Table 9: Risk of renal cell carcinoma and exposure Risk Table9:of carcinoma cell renal

239 *lrt 0.038 0.040 P-trend UCI 2.23 0.57 2.34 0.43 2.34 0.50 0.694 1.98 2.87 3.06 1.94 0.998 0.186 ------LCI OR % N % N Among Male Participants with BMI 30+ BMIwith Participants Male Among P-trend UCI 1.20 0.28 59 36.91.12 55 0.16 32.7 1.20 58 0.64 36.3 59 35.1 1.29 0.71 2.54 0.91 34 70.8 18 66.7 0.64 0.17 1.30 601.43 37.5 63 37.5 1.10 561.33 0.61 35.0 48 28.6 1.59 43 0.88 38.4 42 29.8 1.60 0.83 1.04 0.08 24 21.4 41 29.1 0.95 0.47 ------male participants male LCI OR % N % Among Male Participants with 25-29.9 BMI Participants Male Among N 87 31.0 136 33.5 0.8089 0.53 31.7 149 36.7 0.75 0.51 47 20.9 84 26.1 0.65 0.41 0.01 0.01 0.005 P-trend UCI 0.84 0.87 1.87 0.372 42 75.0 65 77.4 1.05 0.44 0.82 1.02 108 38.40.99 154 37.9 0.88 0.59 960.91 34.2 125 30.8 0.96 82 0.65 36.4 110 34.2 0.88 0.58 s withs Only Low Jobs Intensity (yrs) with HighAny Jobs Intensity (yrs) ------to occupational sunlight to occupational byamong index mass body controls LCI , and self-reported hypertension , and self-reported OR % N % 9 26.5 1625.0 1.00 1425.0 1922.6 1.00 1429.2 9 33.3 1.00 N 69 37.7 10229.1 1.0074 40.4 10931.1 1.0074 49.7 10737.3 86 1.00 30.6 11628.6 1.00 96 34.2 13232.5 1.00 96 42.7 12839.8 41 1.00 25.6 50 29.8 1.00 46 28.8 61 36.3 1.00 45 40.2 58 41.1 1.00 Among Male Participants with BMI <25 BMIwith Participants Male Among Cases Controls Cases Controls Cases Controls 6.00 6.30 6.30 13.50 *Likelihood ratio test ratio p-value *Likelihood All values adjusted for age, center, smoking status smoking center, age, All adjusted for values male exposure among levels on Tertilebased values Table 10: Risk of renal cell carcinoma and exposureTable 10: Risk cell of carcinoma renal Sunlight Cumulative Exposure (low exposure-unit-yrs) < of Exposure (yrs) Duration Frequency-Adjusted < of Exposure for Duration Frequency-Adjusted Subject < >16.10 47 25.7>13.50 115 32.8 0.51 50 0.31 27.3>13.50 118 33.6 0.54 25 0.34 16.8 70 24.4 0.46 0.26 Frequency-Adjusted Duration of Exposure for Duration Frequency-Adjusted Subject < >13.50 25 73.5 48 75.0 0.59 0.19 >6.00-16.10 67 36.6 134 38.2>6.30-13.50 0.64 59 0.40 32.2 124 35.3>6.30-13.50 0.63 50 0.39 33.6 110 38.3 0.56 0.35

240 *lrt P-trend 1.60 0.63 0.122 1.77 1.67 1.86 UCI 1.63 0.66 0.181 1.79 0.77 0.279 16.37 0.84 0.869 ------LCI OR % Participants with BMI30+ Participants N % N P-trend female participants female 1.88 0.66 54 43.2 61 43.6 0.85 0.45 1.441.55 37 29.6 42 30.0 0.900.46 1.55 35 28.0 40 28.6 0.840.43 35 31.8 38 31.9 0.920.46 UCI 1.93 0.58 53 42.4 59 42.1 0.86 0.46 1.80 0.99 41 37.3 43 36.1 0.90 0.45 7.88 0.89 13 86.7 18 85.7 1.30 0.10 ------LCI OR % N % ng Female Participants with BMI25-29.9 Participants Female ng Female Among N 70 39.5 69 34.7 1.12 0.66 0.03 P-trend s s withLowOnly (yrs) Jobs Intensity with Any IntensityHigh (yrs) Jobs to occupationaltosunlight by mass bodyindex among 3.87 3.973.89 50 28.2 61 30.7 0.830.48 3.90 51 28.8 58 29.1 0.900.52 49 32.9 57 31.8 0.890.51 UCI 3.56 0.06 71 40.1 67 33.7 1.14 0.68 3.66 0.07 45 30.2 51 28.5 1.01 0.57 e controls e ------, and self-reported hypertension ,self-reported and LCI OR % N % 1 7.1 216.7 1.00 310.7 210.0 1.00 213.3 314.3 1.00 N 4333.1 7646.1 1.004333.1 7746.7 1.004337.1 7549.0 5631.6 1.00 7135.7 1.00 5631.6 7236.2 1.00 5536.9 7139.7 3528.0 1.00 3927.9 1.00 3628.8 3927.9 1.00 3430.9 3831.9 1.00 Among Female Participants with BMI <25 Participants Female Among Amo Cases Controls Cases Controls Cases Controls 4.50 4.50 4.50 5.70 < Frequency-Adjusted Duration Duration Exposureof for Subject Frequency-Adjusted >5.80 39 30.0>5.70 43 26.1 1.85 43 0.96 33.1>5.70 45 27.3 2.03 30 1.06 25.9 35 22.9 1.81 0.90 >4.50-5.80 48 36.9 46 27.9>4.50-5.70 2.14 44 1.15 33.8 43 26.1>4.50-5.70 2.08 43 1.11 37.1 43 28.1 2.06 1.09 Table 11: Risk of renal cell carcinoma and exposure Table 11: ofRisk carcinoma cell renal Sunlight (low Exposure Cumulative exposure-unit-yrs) < < Frequency-Adjusted Duration Duration Exposureof (yrs) Frequency-Adjusted < Duration Exposureof for Subject Frequency-Adjusted >5.70 13 92.9 10 83.3 NA 25 89.3 18 90.0 0.860.09 All values adjusted for age, center, smoking smoking status center, age, for adjusted valuesAll femal levels based exposureon among values Tertile test p-value ratio *Likelihood

241 The association between occupational UV exposure and RCC risk was not modified when smoking status (results not shown) or hypertensive status (Table 12-14) was considered. However, with increasing cumulative (p-trend= 0.03) and frequency- adjusted duration (p-trend= 0.03) of occupational UV exposure, non-hypertensive males had a stronger reduction in RCC risk compared to hypertensive males (Table 13).

Additionally, since the skin loses its ability to convert 7-dehydrocholesterol (7DHC) to pre-vitamin from the sun as the human body ages [81-83] , occupational UV exposure and

RCC risk was also analyzed by median age. However, in this study age did not modify associations between UV exposure and RCC risk, nor did age modify associations between occupational UV exposure and cancer risk when stratified by sex (results not shown).

242 *lrt P-trend 1.75 0.26 0.322 1.64 1.66 1.71 UCI 1.78 0.21 0.259 1.36 0.773.49 0.735 0.40 0.802 ------LCI OR % ypertensive Participants ypertensive Participants N % N Cases Controls P-trend 1.16 0.31 148 31.0 164 30.0 1.23 0.86 1.291.40 178 37.21.33 205 185 37.5 38.7 1.18 207 0.86 169 37.9 44.7 1.20 192 0.87 40.3 1.22 0.88 UCI 1.16 0.32 156 32.61.13 167 30.6 0.232.69 1.26 0.89 0.65 66 17.5 82 111 82.0 23.3 53 0.89 0.58 76.8 1.46 0.61 ------LCI OR s s withOnly Low Jobs Intensity (yrs) with Any HighIntensity (yrs) Jobs to occupational to sunlight hypertensive bystatus ols tatus, and body mass index mass body tatus,and % N % Among all Non-Hypertensive Participants Participants Non-Hypertensive all Among all Among H N 14 14.7 26 16.4 1.00 18 18.0 16 23.2 1.00 208 36.0 302209 34.1 36.2 1.00 309207 34.9 42.9 1.00 300 41.3 1.00 144 30.1 174 145 31.9 30.3 1.00 175 143 32.1 37.8 1.00 174 36.5 1.00 Cases Controls 5.25 5.25 5.25 9.30 Frequency-Adjusted Duration ofDurationExposure for Subject Frequency-Adjusted < >9.55>9.30 186 32.2>9.30 313 180 35.4 31.1 0.87 311 0.65 99 35.1 20.5 0.87 178 0.65 24.5 0.80 0.57 *Likelihood ratio ratio p-value test *Likelihood Frequency-Adjusted Duration ofDurationExposure (yrs) Frequency-Adjusted < ofDurationExposure for Subject Frequency-Adjusted s sex, smoking center, for age, adjusted All values contr levels exposure among on values based Tertile < Table 12: Risk of renal cell carcinoma and exposure 12: ofTable Risk renalcarcinoma cell Sunlight Exposure (low Cumulative exposure-unit-yrs) < >9.30 81 85.3 133 83.6 1.21 0.54 >5.25-9.55 184>5.25-9.30 31.8 270 189 30.5>5.25-9.30 32.7 0.97 265 0.73 177 29.9 36.6 1.05 248 0.79 34.2 0.99 0.74

243 0.797 *lrt 0.05 P-trend UCI 1.47 0.781.40 0.236 0.630.98 0.303 2.23 0.51 0.701 1.43 1.51 1.49 ------LCI OR % rtensive Male Participants Participants Male rtensive N ong male participants male ong % N 85 33.5 10386 30.7 33.9 0.93 10937 0.60 32.5 19.3 0.90 76 0.58 26.1 0.58 0.35 Cases Controls 0.03 0.03 0.03 P-trend UCI 0.96 0.96 0.95 1.44 0.35 49 79.0 33 75.0 0.67 0.20 1.091.24 98 38.61.11 135 92 40.3 36.2 0.93 122 79 0.60 36.4 41.1 0.99 112 0.64 38.5 0.95 0.60 ------LCI OR with Any IntensityHigh (yrs) Jobs s s withOnly Low Jobs Intensity (yrs) to occupational to sunlight hypertensive am status by controls , and body mass index mass body , and % N % Among Non-Hypertensive Male Participants Participants Non-Hypertensive Male Among Hype Among N 24 31.6 33 25.2 1.00 13 21.0 11 25.0 1.00 126 34.1 172141 29.1 38.1 1.00 199140 33.7 47.6 1.00 191 41.5 1.00 71 28.0 97 76 29.0 29.9 1.00 104 76 31.0 39.6 1.00 103 35.4 1.00 Cases Controls 6.00 6.30 6.30 13.50 < Frequency-Adjusted Duration ofDurationExposure for Subject Frequency-Adjusted < Frequency-Adjusted Duration ofDurationExposure (yrs) Frequency-Adjusted < ofDurationExposure for Subject Frequency-Adjusted status smoking center, for age, adjusted All values Tertile values based on exposure levels among male levels among exposure on values based Tertile p-value ratio test *Likelihood >16.10>13.50 108 29.2 203>13.50 111 34.3 30.0 0.67 217 0.47 59 36.7 20.1 0.69 119 0.49 25.9 0.64 0.42 >13.50 52 68.4 98 74.8 0.71 0.35 >6.00-16.10 136 36.8>6.30-13.50 216 118 36.5 31.9>6.30-13.50 0.78 175 0.56 95 29.6 32.3 0.89 150 0.64 32.6 0.77 0.54 Table 13: Risk of renal cell carcinoma and exposure 13: ofTablecarcinoma Riskrenal cell Sunlight Exposure (low exposure-unit-yrs) Cumulative <

244 *lrt P-trend UCI 10.38 0.82 0.618 2.24 2.24 2.27 2.28 0.272.27 0.527 0.282.15 0.645 0.50 0.911 ------LCI OR % N ong female participants female ong % Among Hypertensive Female Participants Participants Female Hypertensive Among N Cases Controls P-trend UCI 12.75 0.92 34 89.5 23 92.0 0.74 0.05 1.761.75 80 35.71.80 75 79 35.5 35.3 1.33 72 77 0.79 34.1 41.4 1.34 71 0.80 38.2 1.34 0.79 1.72 0.701.77 94 0.60 42.01.80 77 94 0.66 36.5 42.0 1.36 79 60 0.81 37.4 32.3 1.36 56 0.81 30.1 1.22 0.69 ------LCI OR s s withOnly Low Jobs Intensity (yrs) with Any IntensityHigh (yrs) Jobs to occupational to sunlight hypertensive am status by e controls e , and body mass index mass body , and % N % Among Non-Hypertensive Female Participants Participants Female Non-Hypertensive Among 2 10.5 5 17.9 1.00 4 10.5 2 8.0 1.00 N 84 40.4 12884 43.5 40.4 1.00 12983 43.9 43.9 1.00 126 47.4 1.00 50 22.3 59 51 28.0 22.8 1.00 60 49 28.4 26.3 1.00 59 31.7 1.00 Cases Controls 4.50 4.50 4.50 5.70 >5.70 17 89.5 23 82.1 1.14 0.10 Table 14: Risk of renal cell carcinoma and exposure 14: ofTablecarcinoma Riskrenal cell Sunlight Exposure (low exposure-unit-yrs) Cumulative < >4.50-5.80 55>4.50-5.70 26.4 74 51>4.50-5.70 25.2 24.5 1.10 69 50 0.69 23.5 26.5 1.09 67 0.68 25.2 1.11 0.68 Tertile values based on exposure levels among femal levels among exposure on values based Tertile p-value ratio test *Likelihood All values adjusted for age, center, smoking status smoking center, for age, adjusted All values >5.80>5.70 69 33.2>5.70 92 73 31.3 35.1 1.09 96 56 0.69 32.7 29.6 1.13 73 0.72 27.4 1.11 0.68 Frequency-Adjusted Duration ofDurationExposure (yrs) Frequency-Adjusted < ofDurationExposure for Subject Frequency-Adjusted < ofDurationExposure for Subject Frequency-Adjusted <

245 The lack of association among participants with any high intensity occupational UV

exposure may have been influenced by other carcinogenic co-exposures in our

population, particularly pesticide exposures, which has previously been reported to

increase kidney cancer risk among subjects in this study [324] . Therefore, we reanalyzed

the association between occupation UV exposure and renal cancer by first controlling for

pesticide exposures and subsequently by excluding participants with pesticide exposures.

No association between occupational UV exposure and renal cancer was observed in

either analysis (results not shown).

In addition to UV exposure, this study also examined the association between RCC risk

and dietary intake. Given the gender differences observed between RCC risk and

occupational UV exposure, we also explored whether gender differences could explain

some of the association seen between RCC risk and dietary intake frequency of vitamin D

(based on liver, egg, and fish) and calcium (based on milk, yogurt, and cheese) rich

foods. No gender differences were observed between intake frequency of dietary vitamin

D rich foods or dietary intake frequency of calcium rich foods and renal cancer risk when analyses were stratified by sex (results not shown). Furthermore, dietary intake frequency of vitamin D and calcium rich foods were assessed to determine if these variables modified the association between occupational UV exposure and renal cancer risk. Neither intake frequency of vitamin D rich foods nor intake frequency of calcium rich foods modified the association between occupational UV exposure and cancer risk among all participants, male participants or female participants (results not shown).

However, a significant linear trend was observed with increasing cumulative (p-trend=

246 0.04) and frequency-adjusted duration (p-trend= 0.04) of occupational UV exposure among females with high (at or above the median) total calcium intake frequency compared to females with low (below the median) intake frequency (Table 15). *lrt 0.091 0.061 0.04 0.04 P-trend UCI 2.39 2.27 0.112.10 0.184 0.14 0.087 2.35 2.09 2.06 2.23 ------LCI OR % N ium rich foods among female participants female rich ium foods among % Among Female Participants with Intake Participants High Calcium Female Among N Cases Controls P-trend 1.49 0.53 111 39.9 97 32.4 1.55 1.02 1.631.72 81 29.11.63 80 79 26.8 28.4 1.35 77 0.87 77 25.8 33.2 1.33 76 0.86 28.1 1.43 0.91 UCI 1.43 0.43 110 39.6 941.56 31.4 0.58 1.57 1.03 70 30.2 69 25.6 1.43 0.90 ------lcium Intake lcium LCI withHigh Any Jobs Intensity (yrs) swith Only Low Jobs Intensity (yrs) to occupational to sunlight dietary intake byof calc e controls e cheese, and yogurt intake yogurt and cheese, OR on median intake among controls among onintake median , body mass index, and self-reported hypertension self-reported index, andmass body , % N % 1 9.1 6 25.0 1.00 5 10.9 1 3.4 1.00 N 47 30.5 62 30.147 30.5 1.00 64 31.147 32.9 1.00 60 33.0 1.00 87 31.3 125 41.8 88 31.7 1.00 125 41.8 85 36.6 1.00 125 46.3 1.00 Cases Controls Among Female Participants with Low Participants Ca Female Among 4.50 4.50 4.50 5.70 Frequency-Adjusted Duration of DurationExposure Subject for Frequency-Adjusted smoking status center, for adjustedage, valuesAll femal among exposure levels based on values Tertile based intake (low, high) food rich calcium Dietary milk, based foodson rich calcium intake of Dietary < testratio p-value *Likelihood < Frequency-Adjusted Duration of DurationExposure Subject for Frequency-Adjusted >5.80 53>5.70 34.4 75 36.4 56>5.70 36.4 0.79 0.44 78 37.9 46 32.2 0.84 0.47 60 33.0 0.84 0.45 Frequency-Adjusted Duration of DurationExposure (yrs) Frequency-Adjusted < >4.50-5.80 54 35.1>4.50-5.70 69 33.5 51 33.1 0.92>4.50-5.70 0.52 64 31.1 50 35.0 0.97>5.70 0.54 62 34.1 10 0.90 90.9 0.50 18 75.0 NA 41 89.1 28 96.6 0.10 0.00 Table 15: Risk of renal cell carcinoma and exposure and 15: TableRisk of carcinoma cell renal Sunlight Exposure (lowCumulative exposure-unit-yrs) <

247 Comprehensive analysis of 139 single nucleotide polymorphisms (SNPs) across eight

vitamin D pathway genes ( CYP24A1, GC, RXRA, RXRB, STAT1, THRAP4, THRAP5, and

VDR ) identified several (N= 13) SNPs that were significantly associated with RCC risk.

Although no association was observed between the common vitamin D receptor ( VDR)

polymorphisms ( FokI, TaqI , and BsmI ) and RCC risk in our study, most epidemiological

studies focus on these polymorphisms. Therefore, haplotype analyses for the common

VDR polymorphisms in our study were considered in relation to RCC risk. No association was observed between common VDR haplotypes and RCC risk (results not shown).

Additionally, dietary intake frequency of calcium and vitamin D rich foods were also

examined to determine if intake frequency variables modified the association between

haplotypes and RCC risk. No association was observed between dietary intake frequency

of vitamin D or calcium rich foods and VDR haplotypes (results not shown).

Across the eight vitamin D pathway genes, two genes (retinoid-X-receptor-alpha ( RXRA )

and VDR ) were significantly associated with RCC risk. The global significance of

association for these genes in the vitamin D pathway were analyzed using a minimum-p-

value permutation (Min-P) test because it corrected for multiple testing while accounting

for correlations between SNPs within a gene. Since a cut off of 10% for the Min-P test

may have been too conservative to evaluate haplotypes and RCC risk, we also applied

Haplowalk methods to the remaining vitamin D pathway genes.

HaploWalk methods identified a single region of interest across the group specific ( GC ) vitamin D binding protein, centered around introns 11 and 12, that increased RCC risk

248 (Figure 7). Haplotype analyses (shown in Table 16) revealed that among subjects with

the group specific CCAT haplotype, a significant increase in RCC risk (OR= 1.31; 95%

CI= 1.01-1.70) was observed compared to patients with the most common referent haplotype, ACGT ; however, this association was not statistically significant after adjustment of sex, age, smoking status, and study center (OR= 1.28; 95% CI= 0.98-1.68).

The R unadjusted and adjusted global p-values for this region were 0.05 and 0.10, respectively. Additionally through HapoWalk methods, a strong signal was noticed from

region -10041 to -5772 across the STAT1 gene (Figure 8), where participants with the

GCCG haplotype had a significantly increased risk of RCC (OR= 1.21, 95% CI= 1.01-

1.45) compared to participants with the common referent haplotype, GTGG (Table 16).

The R adjusted global p-value for this region was 0.17.

249

250 Table 16: Haplotype associations with genes in the vitamin D pathway Unadjusted Adjusted Haplotypes Cases(%) Controls(%) OR (LCI-UCI) P-value *OR *(LCI-UCI) *p-value Region 1 GC : 3'-rs222035, rs1491709, rs705117, rs17467825-5' A-C-G-T 57.1 56.7 1.00 1.00 C-C-G-C 29.4 30.5 0.96 (0.82-1.11) 0.57 0.94 (0.87-1.06) 0.30 C-C-A-T 8.3 6.2 1.31 (1.01-1.70) 0.04 1.28 (0.98-1.66) 0.07

Region 1 STAT1 : 3'-rs7558921, rs6751855, rs1467199, rs13029532-5' G-T-G-G 57.7 62.1 1.00 1.00 G-C-G-G 11.4 10.7 1.12 (0.91-1.39) 0.30 1.13 (0.91-1.41) 0.26 C-C-G-T 10.0 8.8 1.20 (0.95-1.50) 0.12 1.20 (0.96-1.51) 0.11 G-C-C-G 18.0 15.5 1.22 (1.03-1.46) 0.003 1.21 (1.01-1.45) 0.04

*Adjusted for age, sex, study center, and smoking habit (ever, never) GC chr4 region1: 72840538-72824381 STAT1 chr2 region1: 191596469-191584146

251

To determine if gene-gene interactions existed between the thirteen significant SNPs

(plus the three common VDR SNPs) in our study and RCC risk, multiplicative tests

comparing regression models with and without interaction terms, using a likelihood ratio

252 test (LRT), were performed. The rs4760648 (p-interaction= 0.01) and the rs88641 (p-

interaction= 0.004) VDR polymorphisms significantly interacted with the rs6751855

STAT1 gene (Table 17). For the rs4760648 VDR marker, those with the AA allele had the greatest reduction in RCC risk if they also had the TT STAT1 rs6751855 genotype (OR=

0.48; 95% CI= 0.30-0.79). However, the greatest increase in RCC risk was observed

among participants with the TT VDR rs88641 allele and the TT STAT1 rs6751855

genotype (OR= 2.22; 95% CI= 1.01-4.88). A significant gene-gene interaction was also

observed between the GC rs16847050 polymorphisms and the VDR rs2853564

polymorphism (p-interaction= 0.05). Results in Table 17 show an increase in RCC risk

among participants with the T VDR rs2853564 allele and GG rs16847015 GC genotype

(OR= 1.34; 95% CI= 1.07-1.68). Interestingly all gene-gene interactions were observed

among SNPs that were identified by HaploWalk and haplotype methods. Although these

results are striking, results must be considered cautiously, considering our lack in power,

and issues related to multiple testing, small numbers, and chance.

253 0.01 0.05 *lrt 0.004 UCI 3.41 4.61 1.30 2.14 1.36 1.73 1.89 3.63 5.47 67.79 12.90 3.09 ------LCI OR CC AA % ntrols ntrols N % N UCI 1.691.041.27 36 13.01.04 137 49.5 104 37.5 891.99 211 21.0 49.91.22 241 123 87.0 29.14.45 0.82 1.41 0.51 334 87 0.93 0.74 79.01.68 37.0 65 0.40 27.7 9 118 1.18 50.0 0.80 5.6 51 74 21.6 1.33 33.2UCI 8 0.93 2.36 59 4.5 1.54 2.12 26.12.04 N 2.003.00 2.28 % 0.74 1.52 52.05 33.3 9 N 60.0 1 % 2 6.7 16.7 10 4 OR 66.7 33.3 LCI 6 - 7.47 UCI 50.0 10 0.82 3.78 83.3 1.11 0.36 *lrt 0.04 1.82 0.72 ------LCI OR -23) [-100041T>C] -23) -15) [-17595G>A] -15) CT AG GC % STAT1 N rs16847050 ( rs16847050 rs6751855 rs6751855 ( % N UCI 1.080.790.95 137 43.4 128 40.5 209 51.12.33 265 128 83.9 31.34.88 0.73 0.52 337 0.80 82.4 0.50 2.36 89 21.2 15 0.75 0.54 4.7 145 104 25.5 25.9UCI 13 1.14 3.2 0.81 158 28.6 1.990.89 1.70 1.21 N 0.87 1.89 %1.68 87 N 47.5 26 % 14.2 105 113 48.4 OR 61.7 29 LCI 13.4 1.45 1.03 134 61.8 1.70 0.96 1.50 1.10 ------LCI genes and RCC risk and RCC genes OR it TT GG % N % N N % N % OR LCI 21 13.0 34 19.2 1.00 51 16.1 72 17.6 1.160.80 91 56.549 30.4 76 42.9 67 37.9 0.76 0.54 0.48 0.30 15 5.4 14 3.3 2.22 1.01 85 14.8 109 13.5 1.34 0.95 21036.5 34943.4 1.00 70 38.3 83 38.2 1.450.99 140 87.0 143 80.8111 27.5 0.68 0.49 124 22.1126 33.0 1.67 1.19 138 25.4 1.72 1.24 281 48.8 347 43.1366 63.5 1.34 1.05 456 56.6 1.34 1.07 15137.5 28550.8 1.00 21250.5 25344.5 1.511.14 Cases Controls Cases Controls Cases Controls Cases Co Cases Controls Cases Co CC GG AG AA AG/AA CT TT CT/TT CT TT CT/TT CC ) ) ) SNP SNP VDR-39 VDR-42 VDR-92 rs2853564 (IVS2-1930C>T) ( Table 17: Gene-gene analysespathway for D vitamin Table Gene-gene 17: VDR rs4760648 (IVS2-4108G>A) ( VDR hab smoking and center sex, study forAdjusted age, *Likelihood ratio test p-value ratio test *Likelihood rs886441 (IVS4-4004C>T) (

254 In the third manuscript, VDR and RXRA polymorphisms that were significantly associated with RCC risk were examined in relation to dietary intake frequency of total vitamin D

(based on liver, fish and eggs) and total calcium (based on milk, cheese, and yogurt).

However, the association between dietary intake frequency of total vitamin D and total calcium, RCC risk, and all VDR and RXR SNPs were also assessed. Results presented in

Tables 18 & 19 show SNP variants assessed using a dominant model (i.e. Wald chi- square test for the presence or absence of the variant allele (0, 1)), where participants with the wild type allele were compared to those with at least one variant allele.

Therefore the referent group represents participants with both low (<33%) dietary intake frequency and possession of the wild type allele.

The joint effects between RXR and VDR vitamin D pathway genes, dietary intake frequency of total vitamin D and risk of RCC are presented in Table 18. Reduced RCC risk was linearly observed with increasing intake frequency of total vitamin D among subjects with the wild type allele for VDR rs3819545, rs2189480, and rs11574077 SNPs.

A statistically significant interaction was observed for rs11574077 (p-interaction= 0.035).

However, with increasing intake frequency of total vitamin D reduced RCC risk was

observed among participants with at least one variant allele for the VDR rs3782905 SNP.

This association was also observed among particapnts with at least one variant allele for

two RXRA (rs11103473 and rs11185662) and one RXRB (rs2269346) SNPs. A significant

interaction was shown between RCC risk, dietary intake frequency of total vitamin D and

the RXRB rs2269346 (p-interaction= 0.040) SNP.

255 Table 18: Joint effect of vitamin D pathway genes and dietary intake frequency of total vitamin D on renal cancer risk Wild Type Allele > 1 Variant Allele Cases Controls Cases Controls N % N % OR LCI - UCI P-trend N % N % OR LCI - UCI P-trend *LRT Main Effect ( VDR -40): rs3782905 (IVS4+6584C>G) CC 373 48.0 514 49.7 1.00 CG/GG 404 52.0 521 50.3 1.06 0.87 - 1.28 0.56 Dietary Intake Frequency of Total Vitamin D Low (<33%) 131 35.1 185 36.0 1.00 153 38.0 160 30.7 1.29 0.93 - 1.79 Medium (33-66%) 132 35.4 167 32.5 1.10 0.79 - 1.55 147 36.5 186 35.7 1.10 0.79 - 1.52 High (>66%) 110 29.5 162 31.5 1.02 0.71 - 1.47 0.88 103 25.6 175 33.6 0.89 0.62 - 1.27 0.03 0.250 Main Effect ( VDR -101): rs3819545 (IVS4-6046T>C) TT 275 35.4 337 32.6 1.00 CC/CT 502 64.6 698 67.4 0.91 0.75 - 1.12 0.38 Dietary Intake Frequency of Total Vitamin D Low (<33%) 107 38.9 105 31.2 1.00 177 35.3 240 34.4 0.76 0.54 - 1.07 Medium (33-66%) 108 39.3 124 36.8 0.85 0.57 - 1.27 171 34.1 229 32.8 0.78 0.54 - 1.11 High (>66%) 60 21.8 108 32.0 0.59 0.38 - 0.93 0.02 153 30.5 229 32.8 0.74 0.51 - 1.08 0.99 0.155 Main Effect ( VDR -32): rs2189480 (IVS4-4868C>A) CC 294 37.8 390 37.7 1.00 AA/AC 483 62.2 645 62.3 0.99 0.81 - 1.20 0.91 Dietary Intake Frequency of Total Vitamin D Low (<33%) 111 37.8 121 31.0 1.00 173 35.9 224 34.7 0.85 0.60 - 1.19 Medium (33-66%) 112 38.1 137 35.1 0.92 0.63 - 1.34 167 34.6 216 33.5 0.85 0.60 - 1.20 High (>66%) 71 24.1 132 33.8 0.61 0.40 - 0.93 0.03 142 29.5 205 31.8 0.85 0.59 - 1.23 0.93 0.130 Main Effect ( VDR -94): rs11574077 (IVS5-1456A>G) AA 701 90.2 938 90.6 1.00 AG/GG 76 9.8 97 9.4 0.94 0.68 - 1.30 0.71 Dietary Intake Frequency of Total Vitamin D Low (<33%) 258 36.9 300 32.0 1.00 26 34.2 45 46.4 0.52 0.30 - 0.91 Medium (33-66%) 253 36.1 326 34.8 0.90 0.70 - 1.16 26 34.2 27 27.8 0.97 0.54 - 1.75 High (>66%) 189 27.0 312 33.3 0.76 0.58 - 1.00 0.03 24 31.6 25 25.8 1.12 0.60 - 2.08 0.03 0.035 Main Effect ( RXRA 19): rs11103473 (IVS1-5849T>A) TT 316 40.7 433 41.8 1.00 AA/AT 461 59.3 602 58.2 1.09 0.90 - 1.33 0.37 Dietary Intake Frequency of Total Vitamin D Low (<33%) 108 34.2 148 34.2 1.00 176 38.3 197 32.7 1.28 0.91 - 1.79 Medium (33-66%) 108 34.2 140 32.3 1.04 0.72 - 1.52 171 37.2 213 35.4 1.17 0.83 - 1.64 High (>66%) 100 31.6 145 33.5 1.02 0.69 - 1.50 0.61 113 24.6 192 31.9 0.92 0.64 - 1.34 0.02 0.660 Main Effect ( RXRA -52): rs11185662 (IVS1-31359T>C) TT 453 58.3 601 58.1 1.00 CC/CT 324 41.7 434 41.9 0.99 0.82 - 1.21 0.96 Dietary Intake Frequency of Total Vitamin D Low (<33%) 157 34.7 201 33.4 1.00 127 39.2 144 33.2 1.17 0.84 - 1.63 Medium (33-66%) 155 34.3 198 32.9 1.04 0.76 - 1.43 124 38.3 155 35.7 1.03 0.74 - 1.43 High (>66%) 140 31.0 202 33.6 0.99 0.71 - 1.38 0.94 73 22.5 135 31.1 0.75 0.51 - 1.11 0.04 0.230 Main Effect ( COL11A2 -02): rs2269346 (IVS1+1038G>A) GG 715 92.0 941 90.9 1.00 AG/GG 62 8.0 94 9.1 0.88 0.62 - 1.24 0.47 Dietary Intake Frequency of Total Vitamin D Low (<33%) 264 37.0 304 32.3 1.00 20 32.3 41 43.6 0.55 0.30 - 0.98 Medium (33-66%) 256 35.9 319 33.9 0.93 0.73 - 1.19 23 37.1 34 36.2 0.76 0.42 - 1.37 High (>66%) 194 27.2 318 33.8 0.76 0.58 - 1.00 0.04 19 30.6 19 20.2 1.33 0.67 - 2.61 0.02 0.040 Dietary effects adjusted for sex, age, center, hypertensive status, BMI, smoking status, and years of occupational sunlight exposure Total vitamin D adjusted main effects: 33-66% intake (OR= 1.00; 95% CI= 0.82-1.21); >66% intake (OR= 0.85; 95% CI= 0.69-1.04); p-trend= 0.13 Total vitamin D unadjusted main effects: 33-66% intake (OR= 0.94; 95% CI= 0.78-1.13); >66% intake (OR= 0.78; 95% CI= 0.64-0.95); p-trend= 0.01 Dietary intake frequency of total vitamin D based on dietary liver, egg, and fish consumption Dietary intake frequency of total vitamin D categorized in tertiles based on controls * Likelihood Ratio Test

The joint effects between RXR and VDR vitamin D pathway genes, dietary intake frequency of total calcium and risk of RCC are presented in Table 19. With increasing

256 dietary intake frequency of total calcium, inreased risk was primarly observed among

participants with the wild type alleles. Increased RCC risk was seen with increasing

dietary intake frequency of total calcium among participants with the wild type VDR rs2239186, rs12717991 and rs2107301 SNPs, wild type RXRA rs3132288, rs9409929,

rs877954, rs3132294, rs4240705, rs3132296, rs11103473, rs11185662, and rs4917348

SNPs, and wild type RXRB rs9277936 and rs1547387 SNPs. A significant interaction was

shown for the VDR rs2107301 (p-interaction= 0.026) SNP. While a borderline

statistically significant interaction was observed for RCC risk, dietary intake frequency of

total calcium and RXR SNPs rs4917348 (p-interaction= 0.053) and rs9277936 (p-

interaction= 0.052).

257 Table 19: Joint effect of vitamin D pathway genes and dietary intake frequency of total calcium on renal cancer risk Wild Type Allele >1 Variant Allele Cases Controls Cases Controls N % N % OR LCI - UCI P-trend N % N % OR LCI - UCI P-trend *LRT Main Effect ( VDR -35): rs2239186 (IVS4+3341T>C) TT 447 57.5 581 56.1 1.00 CT/CC 330 42.5 454 43.9 0.96 0.79 - 1.17 0.70 Dietary Intake Frequency of Total Calcium Low (<33%) 12628.2 20034.4 1.00 10030.4 14531.9 1.140.81 - 1.61 Medium (33-66%) 163 36.5 201 34.6 1.29 0.94 - 1.77 11133.7 15433.9 1.140.81 - 1.62 High (>66%) 158 35.3 180 31.0 1.34 0.95 - 1.90 0.05 118 35.9 155 34.1 1.20 0.84 - 1.72 0.86 0.47 Main Effect ( VDR -97): rs12717991 (IVS4-166G>A) GG 293 37.7 336 32.5 1.00 AG/AA 484 62.3 699 67.5 0.83 0.68 - 1.01 0.06 Dietary Intake Frequency of Total Calcium Low (<33%) 86 29.4 12938.4 1.00 14029.0 21630.9 1.020.72 - 1.46 Medium (33-66%) 107 36.5 111 33.0 1.41 0.95 - 2.09 16734.6 24434.9 1.070.75 - 1.51 High (>66%) 100 34.1 96 28.6 1.51 0.98 - 2.32 0.03 176 36.4 239 34.2 1.11 0.77 - 1.61 0.85 0.357 Main Effect ( VDR -81): rs2107301 (IVS5+3260C>T) CC 384 49.4 481 46.5 1.00 CT/TT 393 50.6 554 53.5 0.91 0.75 - 1.10 0.33 Dietary Intake Frequency of Total Calcium Low (<33%) 10326.8 18037.4 1.00 12331.4 16529.8 1.370.97 - 1.94 Medium (33-66%) 144 37.5 153 31.8 1.63 1.15 - 2.31 13033.2 20236.5 1.160.82 - 1.63 High (>66%) 137 35.7 148 30.8 1.60 1.10 - 2.32 0.01 139 35.5 187 33.8 1.27 0.89 - 1.83 0.51 0.026 Main Effect ( LOC642985 -01): rs3132288 (Ex2C>T) CC 262 33.7 332 32.1 1.00 CT/TT 515 66.3 703 67.9 0.91 0.75 - 1.12 0.40 Dietary Intake Frequency of Total Calcium Low (<33%) 66 25.2 11634.9 1.00 16031.1 22932.6 1.260.87 - 1.83 Medium (33-66%) 91 34.7 113 34.0 1.46 0.96 - 2.23 18335.6 24234.4 1.310.90 - 1.90 High (>66%) 105 40.1 103 31.0 1.78 1.15 - 2.75 0.01 171 33.3 232 33.0 1.24 0.84 - 1.85 0.85 0.073 Main Effect ( RXRA -49): rs9409929 (*12052A>G) AA 302 38.9 415 40.1 1.00 AG/GG 475 61.1 620 59.9 1.02 0.84 - 1.24 0.84 Dietary Intake Frequency of Total Calcium Low (<33%) 81 26.8 14334.5 1.00 14530.6 20232.6 1.270.89 - 1.82 Medium (33-66%) 102 33.8 144 34.7 1.27 0.87 - 1.87 17236.3 21134.0 1.390.97 - 1.97 High (>66%) 119 39.4 128 30.8 1.61 1.08 - 2.40 0.02 157 33.1 207 33.4 1.26 0.87 - 1.84 0.90 0.154 Main Effect ( RXRA -48): rs877954 (IVS9+355A>G) GG 349 44.9 448 43.3 1.00 AG/GG 428 55.1 587 56.7 0.96 0.79 - 1.16 0.68 Dietary Intake Frequency of Total Calcium Low (<33%) 93 26.6 16035.7 1.00 13331.1 18531.5 1.330.94 - 1.89 Medium (33-66%) 133 38.1 157 35.0 1.52 1.06 - 2.18 14133.0 19833.7 1.240.87 - 1.76 High (>66%) 123 35.2 131 29.2 1.59 1.07 - 2.35 0.04 153 35.8 204 34.8 1.31 0.91 - 1.90 0.98 0.095 Main Effect ( RXRA -33): rs3132294 (IVS8+278T>C) TT 442 56.9 583 56.3 1.00 CT/CC 335 43.1 452 43.7 1.01 0.83 - 1.23 0.85 Dietary Intake Frequency of Total Calcium Low (<33%) 12227.6 20535.2 1.00 10431.1 14031.0 1.330.94 - 1.88 Medium (33-66%) 160 36.2 209 35.8 1.30 0.95 - 1.79 11434.1 14632.3 1.330.94 - 1.88 High (>66%) 160 36.2 169 29.0 1.54 1.08 - 2.18 0.04 116 34.7 166 36.7 1.19 0.83 - 1.71 0.72 0.089 Main Effect ( RXRA -38): rs4240705 (IVS5-2122G>A) GG 328 42.2 425 41.1 1.00 AG/AA 449 57.8 610 58.9 0.97 0.80 - 1.17 0.74 Dietary Intake Frequency of Total Calcium Low (<33%) 89 27.1 15436.2 1.00 13730.6 19131.3 1.310.92 - 1.86 Medium (33-66%) 124 37.8 150 35.3 1.47 1.02 - 2.13 15033.5 20533.6 1.270.89 - 1.80 High (>66%) 115 35.1 121 28.5 1.59 1.06 - 2.37 0.02 161 35.9 214 35.1 1.31 0.90 - 1.89 Main Effect ( RXRA -34): rs3132296 (IVS4+1666C>T) CC 359 46.2 488 47.1 1.00 CT/TT 418 53.8 547 52.9 1.07 0.89 1.30 0.46 Dietary Intake Frequency of Total Calcium Low (<33%) 96 26.7 16934.6 1.00 13031.2 17632.2 1.400.99 - 1.98 Medium (33-66%) 135 37.6 171 35.0 1.43 1.01 - 2.03 13933.3 18433.6 1.360.96 - 1.92 High (>66%) 128 35.7 148 30.3 1.49 1.01 - 2.18 0.03 148 35.5 187 34.2 1.42 0.98 - 2.05 0.89 0.231 Main Effect ( RXRA -19): rs11103473 (IVS1-5849T>A) TT 316 40.7 433 41.8 1.00 AT/AA 461 59.3 602 58.2 1.09 0.90 - 1.33 0.37 Dietary Intake Frequency of Total Calcium Low (<33%) 85 26.9 15836.5 1.00 14130.7 18731.1 1.511.06 - 2.14 Medium (33-66%) 117 37.0 153 35.3 1.45 1.00 - 2.11 15734.1 20233.6 1.491.05 - 2.12 High (>66%) 114 36.1 122 28.2 1.68 1.12 - 2.52 0.02 162 35.2 213 35.4 1.45 1.00 - 2.11 0.89 0.069 Main Effec ( RXRA- 52): rs11185662 (IVS1-31359T>C) TT 453 58.3 601 58.1 1.00 CT/CC 324 41.2 434 42.9 0.99 0.82 - 1.21 0.95 Dietary Intake Frequency of Total Calcium Low (<33%) 13429.6 21235.3 1.00 92 28.4 13330.6 1.120.79 - 1.59 Medium (33-66%) 155 34.3 209 34.8 1.19 0.87 - 1.62 11936.7 14633.6 1.240.89 - 1.75 High (>66%) 163 36.1 180 30.0 1.37 0.96 - 1.91 0.04 113 34.9 155 35.7 1.15 0.80 - 1.65 0.73 0.429 Main Effect ( RXRA -58): rs4917348 (-4332A>G) AA 516 66.4 698 67.4 1.00 AG/GG 261 33.6 337 32.6 1.05 0.86 - 1.29 0.63 Dietary Intake Frequency of Total Calcium Low (<33%) 15029.1 24535.1 1.00 76 29.1 10029.7 1.250.86 - 1.82 Medium (33-66%) 171 33.2 240 34.4 1.15 0.86 - 1.55 10339.5 11534.1 1.431.01 - 2.05 High (>66%) 194 37.7 213 30.5 1.42 1.03 - 1.96 0.02 82 31.4 122 36.2 1.08 0.74 - 1.59 0.31 0.053 Main Effect ( RING1 -07): rs9277936 (*3071A>T) AA 377 48.5 543 52.5 1.00 AT/TT 400 51.5 492 47.5 1.14 0.94 - 1.38 0.15 Dietary Intake Frequency of Total Calcium Low (<33%) 10628.2 20036.8 1.00 12030.0 14529.5 1.501.06 - 2.12 Medium (33-66%) 126 33.5 181 33.3 1.29 0.92 - 1.81 14837.0 17435.4 1.531.09 - 2.14 High (>66%) 144 38.3 162 29.8 1.59 1.11 - 2.28 0.04 132 33.0 173 35.2 1.36 0.94 - 1.95 0.77 0.052 Main Effect ( SLC39A7 -01): rs1547387 (Ex3-8C>G) CC 583 75.0 779 75.3 1.00 CG/GG 194 25.0 256 24.7 1.01 0.81 - 1.26 0.83 Dietary Intake Frequency of Total Calcium Low (<33%) 16828.9 27735.6 1.00 58 29.9 68 26.6 1.420.94 - 2.14 Medium (33-66%) 203 34.9 261 33.5 1.27 0.96 - 1.67 71 36.6 94 36.7 1.210.83 - 1.77 High (>66%) 211 36.3 241 30.9 1.38 1.02 - 1.87 0.03 65 33.5 94 36.7 1.11 0.74 - 1.66 0.19 0.183 Dietary effects adjusted for sex, age, study center, self-reported hypertensive status, BMI and smoking status Total calcium adjusted main effects: 33-66% intake (OR= 1.18, 95% CI= 0.95-1.47); high intake (OR= 1.12, 95% CI= 0.88-1.42); p-trend= 0.34 Total calcium unadjusted main effects: 33-66% intake (OR= 1.14, 95% CI= 0.94-1.39); high intake (OR= 1.25, 95% CI= 1.03-1.51); p-trend= 0.02 Dietary intake frequency of total calcium based on dietary cheese, yogurt, and milk consumption Dietary intake frequency of total calcium categorized in tertiles based on controls * Likelihood Ratio Test

258 The joint effects between all RXR and VDR vitamin D pathway genes were also assessed in relation to occupational UV exposure and RCC risk among males (Table 20). With increasing cumulative UV exposure, reduced RCC risk was linearly observed among males with at least one variant VDR rs10783219, rs2239186, rs12717991, or rs2107301 allele. However, this association was seen among participants with the wild type rs2239182 VDR allele (p-trend= 0.02). Across the RXR genes, decreased RCC risk was shown with increasing cumulative UV exposure among male subjects with the wild type rs4917348 RXRA (p-trend= 0.03) and wild type RXRB rs2855459 (p-trend= 0.03) SNPs.

259 Table 20: Joint effect of vitamin D pathway genes and cumulative occupational UV exposure on renal cancer risk among males Wild Type Allele >1 Variant Allele Cases Controls Cases Controls N % N % OR LCI - UCI P-trend N % N % OR LCI - UCI P-trend *LRT Main Effect ( VDR -105): rs10783219 (IVS1-1747A>T) AA 180 38.1 223 34.4 1.00 AT/TT 292 61.9 425 65.6 0.85 0.66 - 1.10 0.22 Cumulative exposure high risk allele (low exposure-yrs) <6.00 44 25.7 67 30.6 1.00 93 32.9 121 28.9 1.20 0.74 - 1.94 >6.00-16.10 65 38.0 84 38.4 1.08 0.64 - 1.83 11139.2 15236.4 1.03 0.64 - 1.66 >16.10 62 36.3 68 31.1 1.17 0.68 - 2.01 0.74 79 27.9 14534.7 0.66 0.40 - 1.10 0.01 0.06 Main Effect ( VDR -35): rs2239186 (IVS4+3341T>C) TT 284 60.2 369 56.9 1.00 CT/CC 188 39.8 279 43.1 0.92 0.72 - 1.18 0.53 Cumulative exposure high risk allele (low exposure-yrs) <6.00 77 28.1 109 30.1 1.00 60 33.3 79 28.7 1.12 0.71 - 1.77 >6.00-16.10 111 40.5 131 36.2 1.05 0.70 - 1.59 65 36.1 10538.2 0.87 0.55 - 1.36 >16.10 86 31.4 122 33.7 0.80 0.52 - 1.23 0.59 55 30.6 91 33.1 0.73 0.45 - 1.17 0.02 0.82 Main Effect ( VDR -97): rs12717991 (IVS4-166G>A) GG 182 38.6 212 32.7 1.00 AG/AA 290 61.4 436 67.3 0.81 0.63 - 1.05 0.11 Cumulative exposure high risk allele (low exposure-yrs) <6.00 51 29.5 63 30.1 1.00 86 30.6 125 29.2 0.85 0.53 - 1.37 >6.00-16.10 64 37.0 75 35.9 0.85 0.50 - 1.44 11239.9 16137.6 0.83 0.52 - 1.32 >16.10 58 33.5 71 34.0 0.80 0.47 - 1.37 0.94 83 29.5 14233.2 0.59 0.36 - 0.96 0.03 0.57 Main Effect ( VDR -81): rs2107301 (IVS5+3260C>T) CC 240 50.8 301 46.5 1.00 CT/TT 232 49.2 347 53.5 0.86 0.67 - 1.10 0.23 Cumulative exposure high risk allele (low exposure-yrs) <6.00 64 28.1 91 30.8 1.00 73 32.3 97 28.4 1.02 0.65 - 1.61 >6.00-16.10 85 37.3 109 36.9 0.94 0.60 - 1.48 91 40.3 12737.1 0.94 0.60 - 1.46 >16.10 79 34.6 95 32.2 0.91 0.57 - 1.46 0.92 62 27.4 11834.5 0.60 0.37 - 0.96 0.01 0.25 Main Effect ( VDR -84): rs2239182 (IVS5+3419A>G) AA 140 29.7 174 26.9 1.00 AG/GG 332 70.3 474 73.1 0.84 0.64 - 1.11 0.22 Cumulative exposure high risk allele (low exposure-yrs) <6.00 41 30.6 46 26.7 1.00 96 30.0 142 30.5 0.73 0.44 - 1.22 >6.00-16.10 54 40.3 64 37.2 0.94 0.52 - 1.68 12238.1 17237.0 0.67 0.40 - 1.11 >16.10 39 29.1 62 36.0 0.53 0.29 - 0.99 0.02 102 31.9 151 32.5 0.60 0.35 - 1.01 0.46 0.49 Main Effect ( RXRA -58): rs4917348 (-4332A>G) AA 307 65.0 451 69.6 1.00 AG/GG 165 35.0 197 30.4 1.16 0.89 - 1.51 0.21 Cumulative exposure high risk allele (low exposure-yrs) <6.00 93 31.7 126 28.5 1.00 44 27.3 62 31.8 0.88 0.54 - 1.45 >6.00-16.10 112 38.2 171 38.7 0.85 0.58 - 1.25 6439.8 6533.3 1.200.76 - 1.91 >16.10 88 30.0 145 32.8 0.70 0.46 - 1.05 0.03 53 32.9 68 34.9 0.88 0.54 - 1.43 0.95 0.58 Main Effect ( COL11A2 -07): rs2855459 (IVS4-61C>T) CC 372 78.8 523 80.7 1.00 CT/TT 100 21.2 125 19.3 1.10 0.80 - 1.49 0.64 Cumulative exposure high risk allele (low exposure-yrs) <6.00 115 31.9 158 30.8 1.00 22 23.4 30 24.2 0.91 0.48 - 1.72 >6.00-16.10 136 37.8 183 35.7 0.97 0.68 - 1.38 4042.6 5342.7 1.000.60 - 1.67 >16.10 109 30.3 172 33.5 0.75 0.52 - 1.09 0.04 32 34.0 41 33.1 0.88 0.50 - 1.55 0.71 0.67 Occupational UV exposure effect for males adjusted for age, center, hypertension status, BMI, smoking status, and intake frequency of total vitamin D Occupational UV exposure effect among males: >6.00-16.10 (OR= 0.83; 95% CI= 0.64-1.08); >16.10 (OR= 0.76; 95% CI= 0.58-1.00); p-trend=0.05 Years of occupational UV exposure for male subjects (range= 0.2-76.9 yrs, mean= 12.2 yrs, median= 7.8 yrs) Tertile values based on exposure levels among all male controls * Likelihood Ratio Test

Given the complex relationship between vitamin D, calcium and cancer risk, one might expect that subjects with the low risk alleles would observe a greater decrease in renal cancer risk with increasing consumption frequency of dietary vitamin D and occupational sunlight exposure and even possibly with increasing consumption frequency of dietary

260 calcium. However, the results presented here can only suggest that certain SNPs across

RXR and VDR genes in the vitamin D pathway modify the association between total vitamin D and total calcium intake frequency, along with occupational UV exposure and

RCC risk. Again, results need to be considered carefully due to chance, multiple testing, low numbers, and lack of power.

Since the most common form of renal cell carcinoma is of clear cell type, all significant results from the three manuscripts and the appendix were reanalyzed using only cases histologically diagnosed with this subtype of renal cancer. Results were not affected

(change in OR by at least 10%) when cases were confined to those with clear cell subtype

(results not shown).

Lastly, the relative importance of diet, sunlight and genotype were examined in relation to developing RCC by calculating population attributable risks (PARs). Each estimate should be considered cautiously, considering each exposure variable has its own limitations, which were previously mentioned, and results are most likely overestimated given that results are presented unadjusted. The PARs for cumulative occupational UV exposure among all participants and by gender are shown in Table 21. Noteworthy results indicate approximately 12% of RCC among female participants can be attributed to occupational UV exposure above the median (>5.3 low-exposure-unit-years). No gender differences were observed for calculated PARs among participants with dietary intake frequency of vitamin D rich foods (Table 22); regardless of gender, approximately 8% reduction in RCC can be attributed to high (above the median) intake frequency of

261 vitamin D rich foods. Table 23 shows the PARs for significant SNPs across the VDR

gene, including the common VDR polymorphisms, for all participants genotyped.

Noteworthy results were observed for participants with the VDR rs4760648 and rs2853563 polymorphisms. According to calculated PARs from Table 23, approximately

18% reduction in RCC can be attributed to possession of heterozygous/homozygous variant VDR rs4760648 alleles, while approximately 12% of RCC can be attributed to

possession of the heterozygous/homozygous variant VDR rs2853564 alleles. Interestingly

again, both polymorphisms were identified by HaploWalk and haplotype methods.

262 Table 21: Population attributable risk among participants occupationally exposed to sunlight ALL Participants Cumulative Exposure (low-exposure-unit-yrs) >6.5 <6.5 RCC Cases 507 550 1057 Controls 712 720 1432 1219 1270 2489 PAR= -3.5% Male Participants Cumulative Exposure (low-exposure-unit-yrs) >8.0 <8.0 RCC Cases 303 322 625 Controls 469 458 927 772 780 1552 PAR= -4.3% Female Participants Cumulative Exposure (low-exposure-unit-yrs) >5.3 <5.3 RCC Cases 214 218 432 Controls 216 289 505 430 507 937 PAR= 11.8% Cumulative exposure is categorized into high (above the median) and low (at or below the median) based on all controls, all male controls, and all female controls

263 Table 22: Population attributable risk for dietary intake frequency of total vitamin D ALL Participants Intake Frequency of Total Vitamin D >50% <50% RCC Cases 486 606 1092 Controls 719 757 1476 1205 1363 2568 PAR= -8.2%

Male Participants Intake Frequency of Total Vitamin D >50% <50% RCC Cases 292 351 643 Controls 472 480 952 764 831 1595 PAR= -8.3%

Female Participants Intake Frequency of Total Vitamin D >50% <50% RCC Cases 194 255 449 Controls 247 277 524 441 532 973 PAR= -7.4% Dietary intake frequency is categorized into high (above the median) and low (at or below the median) based on all controls, all male controls, and all female controls

264 Table 23: Population attributable risk among VDR genotyped participants VDR- 107 rs11574027 (IVS2+6247G>T) GT GG RCC Cases 25 752 777 Controls 17 1018 1035 42 1770 1812 PAR= 1.6%

VDR -42 rs4760648 (IVS2-4108G>A) AG/AA GG RCC Cases 488 289 777 Controls 705 326 1031 1193 615 1808 PAR= -17.6%

VDR -39 rs2853564 (IVS2-1930C>T) CT/TT CC RCC Cases 490 286 776 Controls 600 434 1034 1090 720 1810 PAR= 12.2%

VDR -36 rs2254210 (IVS3-816C>T) CT/TT CC RCC Cases 453 324 777 Controls 550 483 1033 1003 807 1810 PAR= 10.8%

VDR -04 rs10735810 (Ex4+4T>C) CT/CC TT RCC Cases 525 286 811 Controls 692 339 1031 1217 625 1842 PAR= -7.3%

VDR -92 rs886441 (IVS4-4004C>T) CT/TT CC RCC Cases 313 464 777 Controls 361 672 1033 674 1136 1810 PAR= 8.2%

VDR -08 rs1544410 (IVS10+283G>A) AG/AA GG RCC Cases 461 315 776 Controls 623 412 1035 1084 727 1811 PAR= -2.0%

265