Introduction
In Belgium, 2159 patients were newly diagnosed with a primary bladder cancer in 2008. These numbers have been stable over the last five years. The incidence and mortality rates increase sharply with age and about two-thirds of patients are ≥ 65 years old. The mean age of diagnosis is 73 years for men and 74 years for women. In the age group 45-59 years, the male/female ratio is 3:3, while in the age group 60-74 years the incidence rates in males are more than fivefold than the rates in females. The relative 5-year survival is 47% for men and 54% for women [
1]. Almost 400,000 bladder cancer cases occurred worldwide in 2008 [
2]. Moreover, in the USA and probably in most Western countries, bladder cancer is the most expensive cancer in terms of healthcare expenditure [
3] because of lifetime ongoing cystoscopies. Any reduction in the need for cystoscopies reduces the cost and even more important, improves quality of life.
Evidence supporting the use of selenium as a general cancer preventive agent includes proof from geographical, animal, in vitro, epidemiological and intervention studies. High selenium intake in Venezuela is associated with a reduced cancer risk [
4]. In animal models, antitumourigenic activity has been observed for metabolites of naturally occurring forms of selenium such as selenomethionine, selenocysteine and methylselenocysteine and inorganic selenium salts, such as selenite and selenate [
5,
6]. Recent in vitro studies have demonstrated that selenium may be an effective chemopreventive and anticancer agent with a broad spectrum against several human cancer cells (prostate, colon, bladder, lung, liver, ovarian, leukemia). In total, twenty-eight different selenium compounds have been reported to have anticancer, chemopreventive or apoptotic activities [
7‐
9]. Three case-control studies reported an increased risk of bladder cancer, associated with lower serum [
10] and toenail [
10‐
12] selenium concentrations. A meta-analysis of bladder cancer incidence in five observational studies [
13‐
17] found an inverse association with an overall risk estimate of 0.67 (95% CI 0.46 to 0.97) suggesting a protective effect of higher selenium levels against bladder cancer [
18]. The Nutritional Prevention of Cancer (NPC) study from Clark in 1996 was the first intervention study that showed a decrease in the incidence of prostate, lung and colorectal cancers in the selenium-supplemented group of older Americans. The effect seemed to be the strongest in the individuals with the lowest selenium status (< 123.2 ng/mL) [
19,
20]. Nevertheless, not all cancer prevention trials indicated reduced cancer risk by selenium supplementation. The selenium and Vitamin E Cancer Prevention Trial (SELECT) failed to show a benefit of selenium supplementation in reducing the risk of prostate cancer in a population of healthy men [
21]. The lack of positive effect of selenium supplementation on the prostate cancer incidence observed in this study may have been due to the different form of selenium used, selenomethionine, compared to Se-enriched yeast in the NPC study [
8,
22].
The review of Brinkman et al. [
23] and the later studies of Grossman et al. [
24], Altwein et al. [
25], Busby et al. [
26] and Amaral et al. [
27] suggest that selenium may be suitable for chemoprevention as well as for treatment. It is useful to perform a selenium trial in a country such as Belgium where the selenium intake is low due to low soil selenium content. In the Belgian case control study the mean blood selenium level in the cases was 78.77 ng/mL compared with 92.31 ng/mL in the controls [
10]. This contrasts with the patients enrolled in the NPC and the SELECT trials who had higher initial plasma levels of selenium (113 ng/mL and 135 ng/mL, respectively) [
28] and an intake of approximately 90 μg/day. As selenium is mainly excreted in the urine, it comes into direct contact and has prolonged exposure with the bladder mucosa, making the role as a potential chemoprevention agent biologically plausible, unlike the case in prostate cancer.
The objective of SELEBLAT is to investigate whether 200 μg/day Selenium-yeast, in addition to standard care, reduces the risk of recurrence for patients with non-invasive bladder cancer.
Discussion
This is the first report of a trial in tertiary prevention of selenium for bladder cancer. Other chemopreventive trials have been performed [
24,
36]. High dose vitamins improved significantly the time to tumour recurrence (p = 0.0014) [
37]. Retinoids have been tested with various results [
38‐
40]. Neither vitamin B6 (Pyridoxine) [
41,
42] nor difluoromethylornithine reduced the rate of recurrence [
43]. Lactobacillus casei powder significantly reduced the time to recurrence in the intervention group [
44,
45]. Several ongoing trials are investigating cyclooxygenase-2 inhibitors in the recurrence of bladder cancer. Preliminary analysis of the results of the study based at the MD Anderson Cancer Center did not demonstrate a difference in time to tumour recurrence between the two treatment arms [
24]. We are aware of one other selenium and vitamin E factorial trial (SELENIB) that is currently recruiting in the UK [
46]. This chemoprevention study, which is largerly similar to our study (as far as the selenium arm is concerned), aims to include 500 patients with non muscle-invasive bladder cancer. In this trial, patients receive a daily supplement of 200 μg high selenium yeast or placebo and 154 mg daily α-tocopherol or placebo for 5 years using a 2 × 2 factorial design. Two chemopreventive trials on prostate cancer with selenium yeast are currently ongoing [
47].
Selenized yeast was selected for use in this trial because of its availability and well-characterized safety profile. In both the SELECT trial (prostate cancer prevention) and the NPC trial (basal cell or squamous cell skin cancer prevention), selenium supplementation was not significantly associated with any of the cardiovascular disease endpoints during 7.6 years of follow-up (all cardiovascular diseases, myocardial infarction, stroke, all cardiovascular disease mortality) [
48]. In secondary analyses of data from the NPC study, 58 subjects in the selenium-supplemented group (n = 600) and 39 subjects in the placebo group (n = 602) developed type 2 diabetes (hazard ratio 1.55; 95% CI, 1.03-2.33) [
49]. In the SELECT trial a statistically non-significant increased risk of type 2 diabetes was observed in the selenium group (RR, 1.07; 99% CI, 0.94-1.22) but not in the selenium vitamin E group [
21]. A secondary analysis was carried out using data from a pre-existing randomized clinical trial designed to investigate the effects of selenium yeast on prostate cancer progression (Watchful Waiting Trial) [
50]. There were no statistically significant differences in glucose levels during the course of the trial in men supplemented with selenium as compared with those on placebo. Laclaustra et al. reports that in U.S. adults, high serum selenium concentrations were associated with higher prevalence of diabetes and higher fasting plasma glucose and glycosylated hemoglobin levels. Mean serum selenium level in this trial was 137.1 ng/mL [
51]. In the New England case-control study, independently of the selenium serum level, an increase bladder cancer risk was associated with a history of diabetes (adjusted odds ratio = 2.2, 95% CI, 1.3 to 3.8). The risk may be greater among patients taking oral hypoglycemics and those with diabetes of longer duration [
52]. Total cholesterol, triglycerides, LDL cholesterol, and fasting serum glucose concentrations significantly increased with serum selenium concentration in the Taiwanese elderly. The mean serum selenium concentration was 89.76 ng/mL [
53]. In the UK, the PRECISE Pilot trial randomized 501 elderly volunteers of relatively low selenium status [mean (SD) plasma selenium 88.8 (19.2 ng/mL)] to a six-month treatment with 100, 200 or 300 mg selenium/day as high-selenium yeast or placebo yeast [
36]. Supplementation at 100 and 200 mg selenium/day lowered total serum cholesterol and non-HDL cholesterol [
54]. The effects of a long-term selenium supplement on blood pressure are inconsistent. Laclaustra et al. reported that blood pressure increases if the plasma selenium level is higher than 160 ng/mL [
55]. In a cross-sectional study with selenium in Belgium, the Flemish Study on Environment Genes and Health Outcomes (FLEMENGHO), 20 ng/mL higher plasma selenium level was associated with a clinically not relevant lower blood pressure with an effect sizes of 2.2 mmHg systolic (95% CI -0.57 to -5.05; p = 0.009) and 1.5 mmHg diastolic (95% CI -0.56 to -2.44; p = 0.017) in men, but not in women [
56]. As in Belgium the daily intake of selenium is low, we think that patients taking selenium for three years can only benefit from this treatment, including of blood pressure reduction.
If it is possible to increase the time to recurrence with selenium, patients have to undergo fewer cystopscopies. This means a reduction not only in suffering for the patient, but also in costs for society.
We are aware that the selenium status at baseline and the genetic variation of tested individuals may represent additional reasons for positive or negative results [
8] or risk of disease [
54]. There are 25 selenoproteins and three selenium-containing enzymes (glutathione peroxidases, thioredoxin reductases and iodothyronine deiodinases), the first two being linked with antioxidant activity, and the latter involved with thyroid hormone metabolism. Rayman [
57] suggests that only those persons at risk, those with single nucleotide polymorphisms in selenoproteins, GPx1, GPx4, SEPS1, Sep15, SEPP1 and TXNRD1, should be treated with selenium at doses which optimize the plasma selenium levels in order to activate specific selenoproteins. An intake of 40 μg per day is required to maintain the plasma glutathione peroxidase (GPx) activity at plateau [
58]. We will be investigating those additional genetic analyses in a subsequent sub-project.
The SELEBLAT study is well-designed. Selenium-yeast is a safe and cheap medicine. A computerized program performs our randomization. Neither the investigators, nor the research nurses have access to the randomization numbers. All study personnel is blinded for the medication. Both treatment groups are regularly compared on similarity of prognostic characteristics such as age, gender, smoking, grade of tumour, baseline selenium level and co-morbidity. We will report reasons and numbers of those patients lost to follow up. Both an intention to treat and a per protocol analysis will be performed. For hypothesis building subgroup analyses and analyses of interactions will be performed, as we have demographic and clinical data and questionnaires results on diet and Quality of Life. Standard treatment will be compared between the intervention and the placebo group. Additional intake of selenium by food supplements will be analyzed and compliance will be reported. Our primary outcome measures are both clinical and histologically confirmed recurrences as frequently patients with clinical recurrence are only cauterized without taking a biopsy. Cauterization only reduces the risk of perforation of the bladder. All subsequent patients undergoing TUR operation are included meeting the inclusion criteria. We decided to include both primary and second primary tumours, because the incidence of bladder cancer is low, while the prevalence and the rate of tumour recurrence is high [
24]. Recurrence within the first three months after TUR is considered as residual untreated disease. Although bladder cancer and prostate cancer occur together in 10% of the cases [
59,
60], we decided to exclude patients with prostate cancer too. We deliberately did not exclude any patient for presence of co-morbidity or age aiming for a study population, which represents as far as possible the normal population diagnosed with bladder cancer. We intend to reach the calculated sample size by pooling of our results with the results from the Selenib study (UK), which is similar in design. Progress of the study can be followed on the website:
http://www.genepid.bham.ac.uk/Seleblat_Recruitment.shtml The first results of our study can be expected for 2014.
Acknowledgements
With thanks to all participating urologists from the 15 recruiting hospitals for their continued support: Peter Dekuyper, Henri Minnaert, Guy Renders, Karl Lessage, Pieter Verleyen, Peter Vossaert, Patrick Werbrouck, Dirk Michielsen, Brigitte Winnepenninckx, Pieter Mattelaer, Diederik Ponette, Dirk Herremans, Peter Martens, Jo Stragier, Koen Van Renterghem, Patrick Willemen, Hubert Claes, Benoit Hermans, Jolanda Verheezen, Koen Ackaert, Christophe Ghysel, Dieter Ost and Hans Goethuys. We would also like to recognize the invaluable contribution made by the research nurses: An Smeets, Kim De Mets, Nathalie Lettens, An Sools, Sigrid Steegmans, Karen Joosten, Christel Janssens, Lieve Valkeneers, Els Goossens, Ria Hulsbosch, Valentina Butoescu, Ann Ventriglia and administrative staff: Marina Devis, Simon Maebe. We acknowledge the members of the data and safety monitoring board: Geert Robaeys, Carla Truyers, Willem Oosterlinck, Dirk Vanden Berghe and the staff of the Vesalius Research Centrum, Diether Lambrechts and Gilian Peuteman.
Authors' contributions
MG: study design, conception and design of the article, analysis and interpretation of data, drafting the article. FB: study design, conception and design of the article, analysis and interpretation of data, drafting the article. SJ: study design, revising the article critically for important intellectual content. KA: local investigator, recruitment of patients, revising the article critically for important intellectual content. FA: local investigator, recruitment of patients, revising the article critically for important intellectual content. IB: local investigator, recruitment of patients, revising the article critically for important intellectual content. JB: local investigator, recruitment of patients, revising the article critically for important intellectual content. AB: local investigator, recruitment of patients, revising the article critically for important intellectual content. JD: local investigator, recruitment of patients, revising the article critically for important intellectual content. KD: local investigator, recruitment of patients, revising the article critically for important intellectual content. LG: local investigator, recruitment of patients, revising the article critically for important intellectual content. EK: study design, revising the article critically for important intellectual content. BT: local investigator, recruitment of patients, revising the article critically for important intellectual content. SV: recruitment of patients, revising the article critically for important intellectual content. BV: recruitment of patients, revising the article critically for important intellectual content. FV: local investigator, recruitment of patients, revising the article critically for important intellectual content. KV: local investigator, recruitment of patients, revising the article critically for important intellectual content. HV: study design, revising the article critically for important intellectual content. MZ: study design, conception and design of the article, analysis and interpretation of data, drafting the article. All authors read and approved the final manuscript.