Cryptorchidism concordance in monozygotic and dizygotic twin brothers, full brothers, and half-brothers

Transcription

1 MALE FACTOR Cryptorchidism concordance in monozygotic and dizygotic twin brothers, full brothers, and half-brothers Morten Søndergaard Jensen, M.D., a,b Gunnar Toft, Ph.D., a Ane Marie Thulstrup, Ph.D., a Tine Brink Henriksen, Ph.D., b Jørn Olsen, Ph.D., c Kaare Christensen, D.M.Sc., d and Jens Peter Bonde, D.M.Sc. a a Department of Occupational Medicine, Aarhus University Hospital, Aarhus, Denmark; b Perinatal Epidemiology Research Unit, Department of Pediatrics, Aarhus University Hospital, Skejby, Denmark; c Department of Epidemiology, School of Public Health, University of California, Los Angeles, Los Angeles, California; and d The Danish Twin Register, University of Southern Denmark, Odense, Denmark Objective: To study concordance rates of cryptorchidism (undescended testis) in pairs of boys with varying family structure, to evaluate the risk contribution from the intrauterine environment and genetic factors. Design: Population based study of 1,024,500 Danish boys born from January 1, 1973 to December 31, Classic twin method and computerized square dance design. Setting: Hospitals and outpatient clinics. Patient(s): Six groups of boy pairs: boys with no relation, paternal half-brothers, maternal half-brothers, full brothers, dizygotic twin brothers, and monozygotic twin brothers. Intervention(s): Observational study. Main Outcome Measure(s): Status on each individual regarding cryptorchidism and orchiopexy from the Danish National Patient Register. Result(s): Concordance rates of cryptorchidism in the groups were as follows: boys with no relation 3.2% (95% confidence interval 2.7%-3.6%), paternal half-brothers 3.4% (2.3%-4.7%), maternal half-brothers 6.0% (4.5%- 7.7%), full brothers 8.8% (8.3%-9.8%), dizygotic twin brothers 24.1% (16.0%-33.6%), and monozygotic twin brothers 27.3% (15.5%-41.2%). Conclusion(s): The concordance rate was higher in maternal than in paternal half-brothers, and much higher but of equal magnitude in both twin groups. The findings strongly support that the intrauterine environment and maternal inheritance are contributing to the occurrence of cryptorchidism. (Fertil Steril Ò 2010;93: Ó2010 by American Society for Reproductive Medicine.) Key Words: Cryptorchidism, orchiopexy, concordance rate, twins, half-siblings, genes, environment The etiology of cryptorchidism (undescended testis) remains largely unknown (1), although cryptorchidism is one of the most common abnormalities in newborn boys worldwide; in Denmark, a prevalence as high as 9% at birth has been reported (2). Adverse maternal lifestyle and environmental exposures during pregnancy are suspected to interfere with normal testicular descent and increase the risk of cryptorchidism in the male offspring (1, 3). Genetic factors are expected to cause cryptorchidism or to modify the risk of having the Received July 23, 2008; revised and accepted September 9, 2008; published online November 19, M.S.J. has nothing to disclose. G.T. has nothing to disclose. A.M.T. has nothing to disclose. T.B.H. has nothing to disclose. J.O. has nothing to disclose. K.C. has nothing to disclose. J.P.B. has nothing to disclose. The study was supported by The University of Aarhus Research Foundation (grant no ). Reprint requests: Morten Søndergaard Jensen, M.D., Department of Occupational Medicine, Aarhus University Hospital, Noerrebrogade 44, Building 2C, DK-8200 Aarhus C, Denmark (FAX: ; morten@sondergaard-jensen.dk). condition under certain environmental conditions by gene environment interaction (4, 5). Case control studies indicate that having a father or an older brother with cryptorchidism increases the risk by 4 to 5 times and 6 to 8 times, respectively (6 9). These finding were supported by data from a population-based register study, which further showed a higher recurrence risk in maternal than in paternal half-brothers (10). Male twins had higher concordance rates than other full brothers, indicating an added effect of environmental risk factors (10). Here, we study concordance rates (i.e., the probability that a pair of individuals will both have cryptorchidism, given that one of the individuals has the condition) in pairs of boys with varying family relation. We compare concordance rates in monozygotic and dizygotic twins by the classic twin method and in half-brothers by a computerized square dance design (11 13). The impact of pregnancy interval is studied, and if environmental risk factors with short duration are important we expect that brothers with a small age difference will 124 Fertility and Sterility â Vol. 93, No. 1, January /10/$36.00 Copyright ª2010 American Society for Reproductive Medicine, Published by Elsevier Inc. doi: /j.fertnstert

2 FIGURE 1 Establishing six groups of boy pairs in a study population of 1,024,500 Danish boys born ,079,780 (100 %) boys born and alive in July 2007 No relation 218,713 pairs Randomly generated pairs of unrelated boys with max. 10 years age difference Singletons 1,024,500 (95%) boys with known parents Maternal halfbrothers 31,544 pairs Half-brothers with the same mother and different fathers Dizygotic 2,962 pairs Paternal halfbrothers 34,282 pairs 55,280 (5 %) discarded boys with one or two unknown parents Half-brothers with the same father and different mothers Twins Monozygotic 1,669 pairs Full brothers 206,992 pairs Full brothers with the same father and mother, not twins Jensen et al. Cryptorchidism concordance in brothers. Fertil Steril have higher concordance rates than brothers with a larger age difference (14). MATERIALS AND METHODS We performed a population-based study on the entire population of 1,079,780 Danish boys born from January 1, 1973 to December 31, 2004 and still alive in July The boys were identified in the Danish Civil Registration System, where each person is given a unique 10-digit number (personal identifier) at birth. The record of each boy also contains personal identifiers of both parents (i.e., the father and mother of each boy are uniquely identifiable), and 1,024,500 boys (95%) had identifiers on both parents. They constitute the study population, from which 496,162 pairs of boys were drawn to six different groups (Fig. 1). Randomly selecting two boys from the population who had different fathers and mothers and maximum 10 years age difference generated 218,713 pairs of boys with no relation. Each boy could only be selected once to this group but were eligible for other groups. Two groups of half-brothers were generated without restrictions in age difference: the maternal half-brothers were 31,544 pairs of boys with the same mother and different fathers (maternal or uterine half-brothers), and the paternal half-brothers were 34,282 pairs of boys with the same father and different mothers (paternal or agnate half-brothers). In the full brothers group 206,992 pairs of brothers with the same father and mother were selected with a minimum age difference of 6 months to avoid twins. In families with three or more boys only one pair of half- or full brothers was randomly selected to obtain independence of observations. Groups of 1,696 monozygotic and 2,962 dizygotic twin pairs were identified (more than 95% of all twins born in the study period) in the National Danish Twin Register (15). Zygosity diagnosis was based on four questions on similarity, and the accuracy of this method has been shown to be approximately 95% compared with DNA testing, which has a probability of monozygosity determination of more than 99% (15, 16). Case ascertainment was accomplished by linkage to the Danish National Patient Register using the unique personal identifier. All diagnoses and surgical procedures used in this study of the 992,324 boys in the years were extracted from the register. In this study we use two types of endpoints: all boys with a diagnosis of cryptorchidism (referred to as cases or diagnosis cases ) and a subset of the diagnosed cases that also underwent orchiopexy (referred to as treated cases or orchiopexy cases ). Orchiopexy indicates that the disease (cryptorchidism) was persistent and required surgery, whereas diagnosing cryptorchidism need not lead to clinical actions. Diagnosed cases were boys with the International Classification of Diseases (ICD)-8 codes 75210, 75211, and or the ICD-10 codes Q53, Q53.1, Q53.1A, Q53.2, Q53.2A, and Q53.9. The ICD-9 codes were never used in official registers in Denmark. Orchiopexy cases were diagnosed boys also given one of the codes KKFH00, KKFH01, and KKFH10 in the Nordic Classification of Surgical Procedures or one of the codes 5564 or in the Surgery and Treatment Classification of the Danish National Board of Health. The National Danish Medical Birth Register provided the birth weights and gestational ages of the sons, using the personal identifier. The Danish Data Inspectorate approved the study. Statistical Analysis The analyses were performed using the Stata 9 statistical software package (StataCorp, College Station, TX). The presented concordance rates were calculated as the number of concordant pairs divided by half the case prevalence. For the full brothers diagnosis cases the figures were [497/ (11,309/2)] 100 ¼ 8.8%. The 95% binomial exact confidence intervals (CIs) were presented. Concordance rate ratios (CRRs) were calculated as odds ratios using logistic regression with full brothers as the reference group. In the full brothers concordance rates were stratified by pregnancy interval. The reference pregnancy interval (0 3 years) was compared with 3 6 years and >6 years, respectively. A trend test estimating the reduction in concordance rate for each year of increase in the pregnancy interval was performed by logistic regression. Low birth weight and preterm birth increases the risk of cryptorchidism, and at the same time the recurrence rates of these obstetric complications are higher in maternal than in paternal half siblings (6, 17 19). We restricted our cohort to pairs of term boys (gestational age R37 weeks) with normal birth weight (R2,500 g) to minimize the excess concordance of cryptorchidism in Fertility and Sterility â 125

3 TABLE 1 Estimated degree of shared intrauterine environment and proportion of shared genes in a pair of boys in each group. Characteristic Group Shared intrauterine environment (timing) Proportion of shared genes (origin) a No relation No 0% Paternal half-brothers No 25% (paternal) Maternal half-brothers Yes (different) 25% (maternal) Full brothers Yes (different) 50% (balanced) Dizygotic twins b Yes (same) 50% (balanced) Monozygotic twins b Yes (same) 100% (balanced) a Paternal/maternal: brothers share 50% from common father/mother; balanced: equal genetic contribution from common parents. b Twins share the uterus but have their own blood supply. Genetic expression in monozygotic twin is not 100% identical. Jensen et al. Cryptorchidism concordance in brothers. Fertil Steril maternal half-brothers caused by the higher recurrence rate of low birth weight and short gestational age. Furthermore we truncated data toward longer follow-up in 1-year intervals, performing iterative logistic regressions to take into account bias by diagnostic delay. This was not done in the twin groups because of limited power. The estimated degree of shared intrauterine environment and proportion of shared genes in a pair of boys in each of the six groups are presented in Table 1. RESULTS The case prevalence in twins was not significantly different from that in singletons (Table 2). Concordance rates of diagnosed cryptorchidism in the groups were as follows: no relation 3.2% (95% CI 2.7%-3.6%), paternal half-brothers 3.4% (2.3%-4.7%), maternal half-brothers 6.0% (4.5%-7.7%), full brothers 8.8% (8.3%-9.8%), dizygotic twin brothers 24.1% (16.0%-33.6%), and monozygotic twin brothers 27.3% (15.5%-41.2%). A similar pattern of concordance rates was observed when treated cases were studied (Table 2). The concordance rates in the two specific twin groups were not statistically different from each other but significantly higher than in full brothers. Compared with brothers with the same parents (full brothers), changing a mother (paternal halfbrothers) reduced the cryptorchidism concordance rate significantly more than changing a father (maternal halfbrothers): CRR ¼ 0.41 (95% CI ) vs. CRR ¼ 0.73 ( ), CRR(maternal)/CRR(paternal) ¼ 1.76 ( ). The equivalent orchiopexy concordance rate ratios were 0.34 ( ) vs ( ), CRR(maternal)/ CRR(paternal) ¼ 1.81 ( ). The risk reductions comparing full brothers with maternal or paternal half-brothers were all statistically significant (Table 2). In full brothers concordance rates of diagnosed cryptorchidism were stratified by pregnancy interval with 0 3 year as reference; CRR (3 6 year) ¼ 0.89 ( ) and CRR (R6 years) ¼ 0.78 ( ). The equivalent rates for orchiopexy concordance were CRR (3 6 years) ¼ 0.89 ( ) and CRR (R6 years) ¼ 0.73 ( ). The trend estimates for each year of increase in pregnancy interval were 0.97 ( ) for diagnosed cases and 0.97 ( ) for treated cases. The singleton CCRs presented in Table 2 did not change substantially when our cohort was restricted to term boys with normal birth weight (n ¼ 303,449 pairs), and CCR estimates were also largely unaffected with data truncation for up to at least 20 years follow-up of all boys (n ¼ 147,049 pairs), at which more than 99% of all diagnoses were registered. DISCUSSION Maternal factors (genetic or environmental) contributed significantly more than paternal factors to the risk of having cryptorchidism, but our results also indicate an independent contribution by the fathers. The concordance rates among dizygotic and monozygotic twins were substantial (approximately 25%) and of equal magnitude, indicating the importance of the intrauterine environment. Overall the findings suggest that the search for the etiology of cryptorchidism should include maternal genes, the intrauterine environment, and especially maternal gene environment interaction. The prevalence in the twin population was not significantly higher than in singletons, which would be expected because of lower birth weight and shorter gestational period on average for twins (6, 17, 18). Spontaneous testicular descent after birth is common in preterm boys, and the reporting to the register is usually not done before the age of 1 year to take spontaneous descent into consideration. At that time 126 Jensen et al. Cryptorchidism concordance in brothers Vol. 93, No. 1, January 2010

4 Fertility and Sterility â 127 TABLE 2 Concordance rates and CRRs of cryptorchidism diagnosis and orchiopexy in singletons and twins. Variable No relation Paternal half-brothers Singletons Maternal half-brothers Twins Full brothers Dizygotic Monozygotic CRR maternal /CRR paternal (95% CI) Pairs (n) 218,713 34,282 31, ,992 2,962 1,669 Diagnosis Case prevalence, 12,538 (2.9) 2,021 (2.9) 1,846 (2.9) 11,309 (2.7) 174 (2.9) 88 (2.6) Concordant pairs, 198 (0.09) 34 (0.10) 55 (0.17) 497 (0.24) 21 (0.71) 12 (0.72) Concordance rate, a 3.2 ( ) 3.4 ( ) 6.0 ( ) 8.8 ( ) 24.1 ( ) 27.3 ( ) 1.76 % (95% CI) CRR (95% CI) 0.38 ( ) 0.41 ( ) 0.73 ( ) 1.00 (ref.) 2.97 ( ) 3.01 ( ) ( ) Orchiopexy Case prevalence, 7,878 (1.8) 1,230 (1.8) 1,155 (1.8) 7,005 (1.7) 108 (1.8) 60 (1.8) Concordant pairs, 70 (0.03) 15 (0.04) 25 (0.08) 264 (0.13) 9 (0.30) 8 (0.48) Concordance rate, a 1.8 ( ) 2.4 ( ) 4.3 ( ) 7.5 ( ) 16.7 ( ) 26.7 ( ) 1.81 % (95% CI) CRR (95% CI) 0.25 ( ) 0.34 ( ) 0.62 ( ) 1.00 (ref.) 2.39 ( ) 3.77 ( ) ( ) Note: ref. ¼ reference; CRR ¼ Concordance Rate Ratio. a Concordance rate: [concordant pairs/(case prevalence/2)] 100%. Jensen et al. Cryptorchidism concordance in brothers. Fertil Steril 2010.

5 the prevalences in twins and singletons have apparently equalized. Another cohort study of familial aggregation of cryptorchidism in Denmark was recently published, which used a different analytic approach but reached the same conclusion regarding maternal and paternal half-brothers (10). The study did not include data on birth weight, gestational age, and twin zygosity, and it did not evaluate the effect of pregnancy interval. The finding of a significantly higher recurrence risk in brothers of a cryptorchidism case than in the sons of fathers with cryptorchidism was of particular interest (10). The results from our twin analysis and from the comparison of brothers and sons of cryptorchidism cases point toward important environmental risk factors. Our pregnancy interval analysis showed low (or no) risk reduction with increasing interval, suggesting that the environmental or lifestyle factors of major importance are stable over time, for example epigenetic change or environmental exposures with long half-lives. Important maternal genetic factors also fit with this observation (14). The recurrence risk of low birth weight and preterm birth is higher in maternal than in paternal half sibs (19), which may explain some of the difference in cryptorchidism concordance between the two groups given that these obstetric complications also correlate with the risk of cryptorchidism (6, 17, 18). However, our results did not change substantially in an analysis restricted to term boys with normal birth weight, indicating that low birth weight and preterm birth may not be strong intermediate factors. The difference in cryptorchidism concordance between maternal and paternal half-brothers may then be ascribed to the intrauterine environment and to genetic factors. The proportion of shared genes is the same (25%), but they descend from the mother or father respectively (Table 1). In this regard genomic imprinting (parent-of-origin effect) must be remembered, as demonstrated in the Angelman and Prader-Willi syndromes (20, 21). The difference is also clearly exemplified when looking at the sex chromosomes. Theoretically, paternal half-brothers have a copy of their fathers Y chromosome but different X chromosomes, whereas 50% of maternal halfbrothers share the X chromosome from their common mother and have different Y chromosomes. On the basis of our findings the X chromosome should receive special attention because it contains the androgen receptor gene (AR) and a number of genes whose products are expressed in spermatogonia (22, 23). The inguinoscrotal phase of testicular descent is known to be androgen dependent (24), and elongation of GGN repeat in the AR has been correlated with poorer function of the receptor and increased risk of cryptorchidism (25 27). This may explain some of the excess risk among maternal half-brothers. Our study is a population-based cohort study of the entire Danish male population born , with twins and their zygosity ascertained in the nationwide Danish Twin Register. Case ascertainment was based on routine data from a nationwide register and was of equal quality in all the studied groups, which makes differentiated ascertainment unlikely. It is likely, however, that a family history of cryptorchidism leads to a more intense search for the condition in the newborn, and concordance rates in brothers of probands may be exaggerated owing to this diagnostic bias. The validity of the cryptorchidism diagnosis has to our knowledge never been validated, but the treated cases we studied are expected to have a diagnostic specificity close to 100%. We found no significant differences in concordance rates among dizygotic and monozygotic twins, but this conclusion is only meaningful if the zygosity is properly determined. The applied method of diagnosing zygosity is approximately 95% accurate when DNA testing is used as the gold standard (16). One problem inherent to the study of concordance rates is to determine when a couple is discordant, particularly when there is a diagnostic delay. To accommodate this we had at least 3 years of follow-up of our cohort. Furthermore, among singletons we restricted the cohort to pairs with up to at least 20 years of follow-up without seeing any change in the relative measures (CRRs). This was not possible for twins because of limited twin pairs. In theory the difference in concordance rates between full brothers and dizygotic twin brothers can be attributed to the intrauterine environment (Table 1), but we are hesitant to make that conclusion: first, twin pregnancies differ from those of singletons in several respects, including birth weight and gestational length (28). Secondly, diagnostic bias is more likely in twins than in singleton brothers. Third, during the past 2 decades treatment of infertility has become widespread, resulting in an artificially increased dizygotic twinning rate in infertile couples, which could lead to overestimation of the true concordance rate if poor reproductive capacity is a risk factor for cryptorchidism. This association has been suggested in more than one study (17, 29). We did not have enough power to stratify the twin population into periods before and after assisted reproduction became a common treatment. In conclusion, mothers contributed significantly more than fathers to the risk of having a son with cryptorchidism. The concordance rates among dizygotic and monozygotic twins were substantial and of equal magnitude, pointing toward environmental etiology. When studying the impact of pregnancy interval the risk factors seem stable over longer periods. Our findings suggest that risk factors should be sought in the maternal genes (most obviously the X chromosome) and in the intrauterine environment. REFERENCES 1. Virtanen HE, Toppari J. Epidemiology and pathogenesis of cryptorchidism. Hum Reprod Update 2008;14: Boisen KA, Kaleva M, Main KM, Virtanen HE, Haavisto AM, Schmidt IM, et al. Difference in prevalence of congenital cryptorchidism in infants between two Nordic countries. Lancet 2004;363: Skakkebaek NE, Rajpert-De Meyts E, Main KM. Testicular dysgenesis syndrome: an increasingly common developmental disorder with environmental aspects. Hum Reprod 2001;16: Jensen et al. Cryptorchidism concordance in brothers Vol. 93, No. 1, January 2010

6 4. Bogatcheva NV, Agoulnik AI. INSL3/LGR8 role in testicular descent and cryptorchidism. Reprod Biomed Online 2005;10: Giwercman A, Rylander L, Hagmar L, Giwercman YL. Ethnic differences in occurrence of TDS genetics and/or environment? Int J Androl 2006;29: Weidner IS, Moller H, Jensen TK, Skakkebaek NE. Risk factors for cryptorchidism and hypospadias. J Urol 1999;161: Berkowitz GS, Lapinski RH. Risk factors for cryptorchidism: a nested case-control study. Paediatr Perinat Epidemiol 1996;10: Czeizel A, Erodi E, Toth J. Genetics of undescended testis. J Urol 1981;126: Elert A, Jahn K, Heidenreich A, Hofmann R. [The familial undescended testis]. Klin Padiatr 2003;215: Schnack TH, Zdravkovic S, Myrup C, Westergaard T, Wohlfahrt J, Melbye M. Familial aggregation of cryptorchidism a nationwide cohort study. Am J Epidemiol 2008;167: Boomsma D, Busjahn A, Peltonen L. Classical twin studies and beyond. Nat Rev Genet 2002;3: Olsen J, Schmidt MM, Christensen K. Evaluation of nature-nurture impact on reproductive health using half-siblings. Epidemiology 1997;8: Basso O. Options and limitations in studies of successive pregnancy outcomes: an overview. Paediatr Perinat Epidemiol 2007;21(Suppl 1): Wilcox AJ. The analysis of recurrence risk as an epidemiological tool. Paediatr Perinat Epidemiol 2007;21(Suppl 1): Kyvik KO, Green A, Beck-Nielsen H. The new Danish Twin Register: establishment and analysis of twinning rates. Int J Epidemiol 1995;24: Christiansen L, Frederiksen H, Schousboe K, Skytthe A, Wurmb- Schwark N, Christensen K, et al. Age- and sex-differences in the validity of questionnaire-based zygosity in twins. Twin Res 2003;6: Akre O, Lipworth L, Cnattingius S, Sparen P, Ekbom A. Risk factor patterns for cryptorchidism and hypospadias. Epidemiology 1999;10: Jones ME, Swerdlow AJ, Griffith M, Goldacre MJ. Prenatal risk factors for cryptorchidism: a record linkage study. Paediatr Perinat Epidemiol 1998;12: Basso O, Olsen J, Christensen K. Low birthweight and prematurity in relation to paternal factors: a study of recurrence. Int J Epidemiol 1999;28: Hall JG. Genomic imprinting. Curr Opin Genet Dev 1991;1: Hall JG. Genomic imprinting and its clinical implications. N Engl J Med 1992;326: Lubahn DB, Joseph DR, Sullivan PM, Willard HF, French FS, Wilson EM. Cloning of human androgen receptor complementary DNA and localization to the X chromosome. Science 1988;240: Wang PJ, McCarrey JR, Yang F, Page DC. An abundance of X-linked genes expressed in spermatogonia. Nat Genet 2001;27: Klonisch T, Fowler PA, Hombach-Klonisch S. Molecular and genetic regulation of testis descent and external genitalia development. Dev Biol 2004;270: Lundin KB, Giwercman A, Dizeyi N, Giwercman YL. Functional in vitro characterisation of the androgen receptor GGN polymorphism. Mol Cell Endocrinol 2007;264: Aschim EL, Nordenskjold A, Giwercman A, Lundin KB, Ruhayel Y, Haugen TB, et al. Linkage between cryptorchidism, hypospadias, and GGN repeat length in the androgen receptor gene. J Clin Endocrinol Metab 2004;89: Radpour R, Rezaee M, Tavasoly A, Solati S, Saleki A. Association of long polyglycine tracts (GGN repeats) in exon 1 of the androgen receptor gene with cryptorchidism and penile hypospadias in Iranian patients. J Androl 2007;28: Hall JG. Twinning. Lancet 2003;362: Jensen MS, Toft G, Thulstrup AM, Bonde JP, Olsen J. Cryptorchidism according to maternal gestational smoking. Epidemiology 2007;18: Fertility and Sterility â 129