- Email: [email protected]
How twin studies help to understand inflammatory joint disease
G Model
ARTICLE IN PRESS
BONSOI-4340; No. of Pages 7
Joint Bone Spine xxx (2016) xxx–xxx
Available online at
ScienceDirect www.sciencedirect.com
Review
How twin studies help to understand inflammatory joint disease Nathalie C. Lambert Inserm UMRs1097, parc scientifique de Luminy, 163, avenue de Luminy, case 939, Bâtiment TPR2 Inserm, Entrée A, 1er étage, 13288 Marseille cedex 09, France
a r t i c l e
i n f o
Article history: Accepted 6 January 2016 Available online xxx Keywords: Inflammatory joint disease Twins Monozygotic Dizygotic Heritability
a b s t r a c t Inflammatory joint disease (IJD) is a group of conditions that target the joints and periarticular structures. The contribution of genetic factors to these conditions is often less than 50%, suggesting a major role for environmental influences. Twin studies are the best means of assessing the role for genetic factors in IJD. Conclusive evidence has been provided by a few studies in vast samples of monozygotic and dizygotic twins, with only one twin in each pair having IJD. These studies have been most successful in ankylosing spondylitis and psoriatic arthritis. The other IJDs have proven more difficult to evaluate. This review demonstrates that genetic and environmental factors are inextricably linked and that ascribing IJDs to one or the other is misguided. Awareness of the limitations and possible sources of bias in twin studies is important when seeking to understand the development of these complex diseases. © 2016 Published by Elsevier Masson SAS on behalf of Société française de rhumatologie.
1. Introduction Inflammatory joint diseases (IJD) are conditions that target the joints and periarticular structures (bone, muscles, and tendons). The most common IJD is spondyloarthritis, i.e., ankylosing spondylitis (AS) and psoriatic arthritis (PsA). Features shared by the various forms of spondyloarthritis include inflammation of the spine and pelvis, peripheral arthritis and, to variable degrees, involvement of other organs. The second in frequency is rheumatoid arthritis (RA), which is the most common autoimmune joint disease. Polymyalgia rheumatica (PMR) is less common and chiefly affects individuals older than 60 years of age. The IJD family also includes crystal-deposition and autoinflammatory arthritides such as gout and articular chondrocalcinosis, which are relatively common. Finally, some arthritides related to microorganisms lie in the gray zone between infection and inflammation: a viral or bacterial agent can trigger an inflammatory response, as occurs in reactive arthritis. The immune response is often disproportionate, similar to that seen in rheumatic fever. Finally, the joints are sometimes involved in various systemic autoimmune disorders such as scleroderma, dermatomyositis, and systemic lupus erythematosus (SLE). Many genetic factors are involved in these diseases. The human leukocyte antigen (HLA) genes have long been identified as the main risk factors for most IJDs. For instance, HLA-B*27 is associated with AS [1,2] and some of the HLA-DRB1 alleles (e.g., *0101, *0401,
E-mail address: [email protected]
*0404, *0405, and *1001) that carry an amino acid motif known as the shared epitope are associated with RA [3]. Genome-wide association studies indicate that numerous other genes are risk factors; however, they convey far lower levels of risk, and some polymorphisms are associated with more than one IJD. Thus, PTPN22, which encodes the protein tyrosine phosphatase non-receptor type 22, is associated with many autoimmune diseases [4]. Recent research indicates an association linking PsA to variants in RUNX3, a previously identified AS-susceptibility gene [5]. In most IJDs, concordance between monozygotic twins is lower than 50%. This fact supports a strong role for environmental factors. Twin studies are the best means of assessing the contribution of genetic factors to the development of a disease. 2. Twins: identical versus fraternal Dizygotic or fraternal twins are produced when two ova are independently fertilized by two sperm cells. Each twin develops in its own amniotic sac and has its own placenta: the pregnancy is dichorionic and diamniotic (Fig. 1). Very rarely, dizygotic twins are produced during different ovulations and therefore have different gestational ages. About three-quarters of all twin pairs are dizygotic. Monozygotic twins are formed from a single ovum fertilized by a single sperm cell and account for about one-quarter of all twins. Depending on the stage at which the zygote divides into two embryos, there may be two placentas and amniotic sacs (dichorionic diamniotic, 29%), a single placenta but two amniotic sacs (monochorionic diamniotic, 70%), or a single placenta and a single
http://dx.doi.org/10.1016/j.jbspin.2016.02.008 1297-319X/© 2016 Published by Elsevier Masson SAS on behalf of Société française de rhumatologie.
Please cite this article in press as: Lambert NC. How twin studies help to understand inflammatory joint disease. Joint Bone Spine (2016), http://dx.doi.org/10.1016/j.jbspin.2016.02.008
G Model BONSOI-4340; No. of Pages 7
ARTICLE IN PRESS N.C. Lambert / Joint Bone Spine xxx (2016) xxx–xxx
2
Fig. 1. Development of monozygotic and dizygotic twins.
amniotic sac (monochorionic monoamniotic, 1%) (Fig. 1). Late separation of the zygote is an exceedingly rare event that results in conjoined twins. Twins may differ in their development, with one weighing more than the other. In vanishing twin syndrome, one twin fails to develop and disintegrates [6]. Twinning occurs in about 13/1000 pregnancies overall. However, the frequency of twinning varies with several maternal characteristics including age, ethnicity, nutrition, and fertility [7,8]. The frequency of twinning is highest in Benin (28/1000) and lowest in Asia (< 9/1000) [8]. Differences across ethnic groups occur chiefly for dizygotic twins. The frequency of monozygotic twinning is almost identical throughout the world, at about 4/1000 pregnancies [7]. Fertility treatments also tend to increase the frequency of twinning.
occupation-related exposures to chemicals, the living environment, and lifestyle factors (e.g., alcohol use, smoking, diet, and birth control methods). Environmental factors may act via epigenetic mechanisms to modify the susceptibility of an individual to the development of a disease. Epigenetic factors are molecular mechanisms that induce reversible changes in gene expression levels, in the absence of alterations in the DNA sequence. They include DNA methylation, histone deacetylation, and the expression of small noncoding RNA fragments known as micro-RNAs [11]. Epigenetic factors allow changes in the expression of certain genes depending on the individual’s environment. The influence of genetic factors can be assessed by computing the concordance rate for the disease between monozygotic twins and comparing this value to that in dizygotic twins.
3. Determining the zygosity of twins Determining whether two same-sex twins are dizygotic or monozygotic relies to a variable degree on an empirical approach. Some studies used questionnaires to collect information from twins about their physical resemblance and instances of one having been mistaken for the other. This similarity-based method is practical in vast studies of several thousand-twin pairs. However, in 5% to 10% of cases, the responses are not assessable or not interpretable. Among assessable data, most are fairly reliable: comparisons with genetic testing for zygosity showed less than 5% of classification errors [9]. Nevertheless, the most reliable method is genetic analysis of polymorphisms or short tandem repeats [10]. 4. Evaluation of the genetic and environmental contributions to a disease The development of a disease is multifactorial. Some diseases are characterized by strong heritability, which indicates a major role for genetic factors. Nevertheless, environmental factors may contribute to the development of a disease to a similar extent as genetic factors. Such environmental factors include
5. Computing concordance rates Two main types of concordance rate are used, the pairwise concordance rate and the probandwise concordance rate. The pairwise concordance rate is the proportion of twin pairs with at least one affected twin in which the other twin is also affected. This statistical parameter is thus the proportion of affected pairs in which both twins are affected (Fig. 2A). The probandwise concordance rate (or rate per affected individuals) is the proportion of affected individuals among co-twins of previously identified affected twins. This parameter measures the risk of having the disease among individuals who have an affected twin (Fig. 2B). A higher concordance rate for a disease among monozygotic twins than among dizygotic twins supports a major role for genetic factors in the development of the disease. Higher concordance among dizygotic twins suggests a strong role for environmental factors. A disease caused solely by genetic factors would consistently affect both members of pairs of monozygotic twins. In reality, concordance rates among monozygotic twins are only about 50%, indicating a strong influence of nongenetic factors (Fig. 2).
Please cite this article in press as: Lambert NC. How twin studies help to understand inflammatory joint disease. Joint Bone Spine (2016), http://dx.doi.org/10.1016/j.jbspin.2016.02.008
G Model
ARTICLE IN PRESS
BONSOI-4340; No. of Pages 7
N.C. Lambert / Joint Bone Spine xxx (2016) xxx–xxx
3
Fig. 2. Methods for computing concordance rates for a disease among monozygotic and dizygotic twins.
Heritability is another parameter used in statistical analyses of the relative roles for genetic and environmental factors in the development of a disease within a predefined population. Several modeling techniques are available. The ACE model, for instance, uses three components: the additive (A) effects of alleles at a given locus, i.e., the role for the genotype; the common environment (C), shared by both twins; and the unique environment (E), i.e., the environmental factors related to the personal experience and lifestyle of each individual.
6. Inflammatory joint diseases with a strong genetic component 6.1. Ankylosing spondylitis (AS) The first large twin study of AS was done in Finland and reported in 1995. Subsequently, twin studies were performed in Denmark and Norway (Table 1). Concordance rates were consistently higher in monozygotic than in dizygotic twin pairs, indicating a strong genetic component. Thus, the HLA-B*27 allele, found in 90% of patients with AS, contributes to the development of the disease. However, other genes may be involved also, since when considering only twin pairs carrying HLA-B*27, concordance rates for AS are 50% among monozygotic twins compared to only 20% among dizygotic twins [12]. Nevertheless, concordance rates among dizygotic twins are higher for AS than for most of the other IJDs (Table 1). This fact supports an influence of environmental factors. A few studies have identified epigenetic variations in patients with AS [20–22]. However, to the best of my knowledge, epigenetic factors have not been studied in monozygotic twins discordant for AS. Data from case reports also provide useful insights. Thus, in a monozygotic twin pair with AS, the onset, sites involved, and course of the disease were closely similar between the two patients [23]. Thus, genetic factors influence not only the susceptibility to AS but also the phenotypic expression of the disease.
6.2. Systemic lupus erythematosus (SLE) Among systemic autoimmune diseases that can involve the joints, SLE has the highest heritability, although the level is lower than for AS [24]. Overall, concordance rates are higher among monozygotic than dizygotic twins, ranging from 11% to 40% [24]. However, these rates leave considerable room for effects of environmental and epigenetic factors. Genetic-epigenetic interactions exert a pivotal influence on the susceptibility to SLE, as shown in two recent review articles [11,25]. Few studies have tested the hypothesis that monozygotic twins discordant for SLE have different epigenetic factors. One study demonstrated differences in methylation profiles between affected individuals and their unaffected twins [18]. Another study, however, found no differences in X chromosome inactivation patterns within monozygotic twin pairs discordant for SLE [26].
6.3. Juvenile idiopathic arthritis (JIA) Juvenile idiopathic arthritis (JIA) is a heterogeneous group of IJDs that include oligoarticular JIA, polyarticular JIA (with or without rheumatoid factors), enthesitis-related arthritis, juvenile psoriatic arthritis, and undifferentiated arthritis. In several studies, non-twin siblings concordant for JIA shared larger numbers of HLA haplotypes than did non-twin siblings discordant for the disease [27,28], suggesting a role for the HLA genes. This evidence for a genetic susceptibility to JIA was confirmed by a study of twins concordant for the disease [29]. Of 14 twin pairs from the publication, 11 had zygosity test results available and were monozygotic; 3 had no zygosity test results, including 1 pair discordant for gender. The twin pairs were identified from a registry of siblings affected with JIA and not from a population-based registry. Consequently, the concordance rate cannot be computed. Nevertheless, that all 11 pairs with zygosity tests were monozygotic is worthy of note. Furthermore, age at disease onset was very similar in affected twins,
Please cite this article in press as: Lambert NC. How twin studies help to understand inflammatory joint disease. Joint Bone Spine (2016), http://dx.doi.org/10.1016/j.jbspin.2016.02.008
G Model BONSOI-4340; No. of Pages 7
ARTICLE IN PRESS N.C. Lambert / Joint Bone Spine xxx (2016) xxx–xxx
4
Table 1 Recorded or estimated concordance rates in twin studies of inflammatory joint diseases strongly related to genetic factors. Disease
Total n of pairs: MZ + DZ
% pairwisea concordance MZ DZ
% probandwise concordance MZ DZ
Source, country; reference
AS
26: 6 MZ, 20 DZ 15 DZB27+ 40: 8 MZ, 32 DZ
67
26
National Twin Registry, Finland; Järvinen, 1995 [12]
86
22
40
8
10HLA : 4 MZ, 6 DZ 247: 151 MZ, 96 DZ
50 50HLA 75 75HLA 25 25HLA 57 71ANA+ 25 40
73 83ANA+ 33 57
0 0ANA+ 0 8
5, all MZ discordant for LED 14; 11 MZ, 1 DZ, 2 ND
Significant differenceb No statistical tests
ND
ND
ND
ND
National Rheumatic Diseases database, UK; Brown et al., 1997 [13] 3 nationwide surveys, Denmark, Norway; Pedersen, et al., 2008 [14] Literature review with 12 added pairs, USA; Block et al., 1975 [15] Australia; Grennan et al., 1997 [16] Review of published case-reports of affected twins, USA; Block, 2006 [17] Clinical center of the National institutes for health (NIH), USA; Javierre, et al., 2010 [18] Registry of 118 pairs with both twins affected
435: 175 MZ, 260 DZ
28[19–100]
44[31–100]
11[0–67]
27: 4 MZ, 23 DZ SLE
Methylation 807 promoters JIA All concordant RhF
s
12 : 7 MZ, 3 DZ
15 20HLA 13 27HLA 4 10HLA 0 0ANA+ 0 4
6[0–50]
Metaanalysis of 6 studies published between 1933 and 1964; Engel et al., 2011 [19]
AS: ankylosing spondylitis; SLE: systemic lupus erythematosus; JIA: juvenile idiopathic arthritis; RhF: rheumatic fever; MZ: monozygotic; DZ: dizygotic; ND: not determined; ANA: antinuclear antibodies. a The concordance percentages are rounded up to the next integer. The superscripts indicate the following: (HLA) : HLA matching of co-twins (e.g., for HLA-B*27 when studying AS); (S) : matching on sex; and numbers (e.g., [0–100] ): variations in concordance rates across studies. The subscripts indicate the factor for which concordance was determined, when this factor was not the disease (e.g., antinuclear antibodies: ANA+ ). b The twins were discordant for the disease and were examined for differences in epigenetic factors such as DNA methylation.
with a mean difference of 5–6 months, compared to nearly 5 years for affected non-twin siblings [29].
7. Inflammatory joint disease (IJDs) with a limited genetic component 7.1. Rheumatoid arthritis (RA) The role for genetic factors in RA has been estimated at 16% to 60%. Over 30 genes are associated with RA. Nevertheless, RA is one of the autoimmune diseases, together with systemic sclerosis [30], for which concordance rates in monozygotic twins are fairly low, from 12% to 21% [24], albeit higher than in dizygotic twins (0% to 4%, Table 2). Twin studies have established that the T-cell receptor repertoire is chiefly under genetic control. For instance, strongly similar repertoires have been demonstrated between monozygotic twins discordant for RA [42]. HLA genes play a strong role in this genetic control [43]. Nevertheless, distinctive variations in the repertoire have been demonstrated in patients with RA, suggesting that the disease itself may affect the repertoire [43]. The main genetic factor in RA is the shared epitope. In an elegant study of 91 monozygotic twin pairs, the concordance rate for RA was higher in twins with than without the shared epitope (18% versus 5%) [44]. Even more striking was the finding of a higher concordance rate among twins homozygous compared to heterozygous for the shared epitope (27% versus 13%). Finally, genotypes conferring susceptibility to RA, i.e., heterozygous HLADRB1 genotypes (e.g., DRB1*0401/*0404) were associated with higher concordance rates compared to homozygous genotypes (e.g., DRB1*0401/*0401) (38% versus 10%) [44]. This genotypedependent variation in the risk of RA was recently evaluated in detail by our group [45]. Genetic factors may play a particularly strong role in RA development in patients who already produce anticitrullinated protein antibodies (ACPAs). However, in a study of 148 twin pairs, heritability was nearly identical (about 66%) for ACPA+ and ACPA− disease, indicating a greater role of genetic factors in ACPA− RA than previously thought [37]. Environmental factors may contribute strongly to the development of auto-antibodies themselves (without
development of the disease) [46] and, according to one study, this contribution may be greater than that of genetic factors [39]. Twin studies can be used to evaluate potential risk factors without confounding by genetic factors (when monozygotic twin pairs are studied) or by a number of environmental factors (gestational age at birth, body mass index of the mother, and socio-economic influences). For instance, a study was conducted to determine whether lower birth weight was associated with a greater risk of developing RA compared to the monozygotic twin with a higher birth weight [47]. The results showed no effect of birth weight. Birth order was the only factor associated with the risk of RA [47]. 7.2. Systemic sclerosis Most of the data from twins with systemic sclerosis come from anecdotal cases of twins concordant for the disease. There is, however, a study of 42 twin pairs, in which concordance among monozygotic twins was low for systemic sclerosis but high for antinuclear antibodies [43]. In contrast, Raynaud’s phenomenon without systemic sclerosis has high heritability, although it is the inaugural symptom in most patients with systemic sclerosis [44]. 7.3. Gout Gout is among the most common IJDs. Hyperuricemia often precedes symptom onset, but only 10% of patients with hyperuricemia experience gouty attacks [48]. A single study, from the US, has evaluated the genetic and environmental factors involved in hyperuricemia and gout [40]. However, it included a large number of twin pairs (n = 514). The results establish that hyperuricemia is somewhat influenced by dietary and other lifestyle factors but depends chiefly on genetic factors, whereas gout is chiefly dependent on environmental factors. This finding implies that many cases of gout could be prevented. 8. Inflammatory joint diseases (IJDs) for which twin-study data are scarce For several IJDs, the number of twin pairs studied to date is too small to provide conclusive evidence about the influence of genetic factors.
Please cite this article in press as: Lambert NC. How twin studies help to understand inflammatory joint disease. Joint Bone Spine (2016), http://dx.doi.org/10.1016/j.jbspin.2016.02.008
G Model
ARTICLE IN PRESS
BONSOI-4340; No. of Pages 7
N.C. Lambert / Joint Bone Spine xxx (2016) xxx–xxx
5
Table 2 Recorded or estimated concordance rates in twin studies of inflammatory joint diseases in which genetic factors play only a small role. Disease
Total n of pairs: MZ + DZ
% pairwisea concordance MZ DZ
% probandwisea concordance MZ DZ
Source, country; reference
RA
285: 97 MZ, 188 DZ
32[0–67]
9[0–29]
48[0–80]
16[0–44]
261: not specified
12 9A/S 21
4 3A/S 0
22
7
25
0
15 17S 12 14S 0
4 4S 4 4S 6
27
7
22
8
0
9
16 17ACPA+ 11ACPA− No differenceb Methylation 807 promoters 6
4 4ACPA+ 0ACPA−
27
7
Metaanalysis of 7 studies published from 1938 to 1968; Lawrence, 1970 [31] Finnish Twin Cohort linked with the health insurance database; Aho et al., 1986 [32] Questionnaire, Australian Twin Registry; Bellamy et al., 1992 [33] Nationwide media campaign, prospective survey of rheumatologists, UK; Silman et al., 1993 [34] 2 nationwide studies, in Finland and the UK, respectively; MacGregoret al., 2000 [35], [34] Questionnaires, nationwide twin population in Denmark; Svendsen et al., 2002 [36] Nationwide twin study, UK; van der Woude et al., 2009 [37]
23: 14 MZ, 9 DZ 203: 91 MZ, 112 DZ 246: 73 MZ, 173 DZ 49: 13 MZ, 36 DZ 148: 64 MZ, 84 DZ
5: 5 MZ discordant for RA 155: 32 MZ, 77 DZS and 46 DZSO 117ACPA+ : 21 MZ, 59 DZS and 37 DZSO 380S : 151 MZ, 229 DZS SSc
42: 24 MZ, 18 DZ
Gout PsA
514: 253 MZ, 261 DZ 36: 10 MZ, 26 DZ
10 4ACPA+ 4ACPA+ RA+ 2ACPA+ RA− 4 90ANA 12 10
Clinical center of the National institutes of health (NIH) Javierre, et al., 2010 [18] 5S 2SO 7S 3SO 3ACPA+ 2ACPA+ RA+ 1ACPA+ RA− 6 40ANA 12 4
8ACPA+ 8ACPA+ RA+ 4ACPA+ RA− 8
6S 2SO 9S 3SO 5ACPA+ 4ACPA+ RA+ 3ACPA+ RA− 11
21 18
21 7
9 14
Questionnaires, nationwide twin population in Denmark; Svendsen et al., 2013 [38]
Swedish Twin Registry; Hensvold et al., 2015 [39]
Invitations in patient-group newsletters and to rheumatologists, USA; Feghali-Bostwick et al., 2003 [30] Veterans Affairs Twin registry, USA; Krishnan et al., 2012 [40] Cohort of Danish twins identified by a nationwide questionnaire; Pedersen et al., 2008 [41]
RA: rheumatoid arthritis; SSc: systemic sclerosis; PsA: psoriatic arthritis; MZ: monozygotic; DZ: dizygotic; ACPA: anticitrullinated peptide antibodies; RA: rheumatoid arthritis; ANA: antinuclear antibodies. a The concordance percentages are rounded up to the next integer. The superscripts indicate the following: (S) : matching on sex; (SO ): co-twins of different sexes; (A/S ): matching on both age and sex; (ACPA ): matching on anticitrullinated protein antibody status; (ANA ): matching on antinuclear antibody status; numbers (e.g., [0–100] ): the variations in concordance rates across studies. The subscripts indicate the factor for which concordance was determined, when this factor was not the disease (e.g., anticitrullinated antibodies, ACPA+ and ACPA− ). b The twins were discordant for the disease and were examined for differences in epigenetic factors such as DNA methylation.
8.1. Psoriatic arthritis The only study in PsA showed little difference in concordance rates between monozygotic and dizygotic twins (1/10 and 1/26 pairs, respectively), indicating a need to search for nongenetic factors implicated in this disease [41]. To the best of my knowledge, epigenetic factors have not been investigated in cohorts of twin pairs having one twin with PsA. The above-mentioned study had a sample size of only 36 twin pairs, which limited its statistical power. Thus, the influence of genetic factors on the development of PsA in patients with psoriasis remains incompletely elucidated. 8.2. Sjögren’s syndrome Although several instances of Sjögren’s disease clustering within families have been reported, the presence of this disease in monozygotic co-twins is rare. In the few studies of monozygotic twins, co-twins usually had similar phenotypes, symptoms, and serological findings [49,50]. Another study, however, showed opposite results, indicating a role for environmental factors [51]. There is a pressing need for studies in large cohorts of twins with Sjögren’s syndrome. 9. Limitations of twin studies 9.1. Monozygotic versus dizygotic: an oversimplified model Monozygotic twins, although considered genetically identical, may have small genetic differences due to rare mutations that occur
immediately after the blastocyst splits into two embryos. These mutations are then found in all the tissues and in the germ cells of only one of the twins. The differences thus produced are used for paternity tests and medicolegal investigations. Dizygotic twins are considered relevant for evaluating the contribution of the environment, as opposed to that of genetic factors. Nevertheless, HLA genes are shared more often by dizygotic twins than by siblings from different pregnancies [52]. Dizygotic twins are therefore different from non-twin siblings who share the same environment. Finally, genetics and the environment cannot be viewed as two completely separate influences, since they are tightly intertwined with each other. For instance, psychological stress, long suspected to be one of the triggers of RA [53], is a response to the environment but manifests differently depending on the genetic background [54]. 9.2. Sources of bias Concordance rates for a given disease vary across studies, for instance from 0% to 32% for RA among monozygotic twins. This considerable variability supports the existence of multiple sources of bias related to analysis methods, recruitment modalities, age of the participants, and other factors. For example, populations will differ depending on whether recruitment is via a questionnaire sent to nationwide twin registries or via office-based rheumatologists (Table 2). Similarly, when inviting twin pairs with a given disease to participate in a study, those in which both twins are affected are more likely to accept. Finally, diseases that manifest at a younger age are more
Please cite this article in press as: Lambert NC. How twin studies help to understand inflammatory joint disease. Joint Bone Spine (2016), http://dx.doi.org/10.1016/j.jbspin.2016.02.008
G Model BONSOI-4340; No. of Pages 7
ARTICLE IN PRESS N.C. Lambert / Joint Bone Spine xxx (2016) xxx–xxx
6
heavily influenced by genetic factors and less dependent on environmental factors. For instance, in a vast non-twin study of patients with spondyloarthritis in Brazil, positive HLA-B27 status was more common among patients who experienced their first symptoms before 16 years of age compared to those with adult-onset disease [55]. These patient selection criteria, particularly in twin studies, may influence the findings, yet are not consistently taken into account. Furthermore, disease onset in two co-twins may be separated by a substantial time interval.
9.3. Changes over time in diagnostic criteria The classification criteria for diseases change over time. Consequently, the disease profile in the study population may vary according to the study period, although the name used to designate the disease remains the same. In a study showing a major role for genetic factors even in ACPA− patients with RA [56], some of the included patients would not be classified as having RA based on the 2010 ACR criteria, which include the ACPA status [56]. JIA is a heterogeneous group comprising many clinical patterns (see section 6.3). However, in twin studies, the pattern of JIA in the patients is not consistently specified. Rheumatic fever provides the most striking example of how the diagnosis of a disease can change over time. Rheumatic fever is a severe complication of streptococcal tonsillitis that can result in life-threatening heart valve disease in the medium or long term. In a 2011 metaanalysis with 435 twin pairs from six studies performed between 1933 and 1964, heritability was 60% and concordance rates in monozygotic twins ranged from 31% to 100% [19]. However, these studies were conducted over 45 years ago, in industrialized countries where rheumatic fever no longer occurs and has no clear definition.
9.4. Statistical limitations For a few IJDs, the influence of genetic factors remains unclear. An example is PsA, for which the availability of a single study with only 36 twin pairs results in limited statistical power [41]. Only large cohorts can produce the statistical power needed to generate firm conclusions.
9.5. Missing data and interpretation of data Missing data is a common issue that precludes a full statistical analysis. For instance, for monozygotic twins, information may be lacking on the number of placentas and/or amniotic sacs (Fig. 1). Flawed data interpretation is another issue. For instance, twins of different genders are classified as dizygotic and those who shared a placenta as monozygotic. In a very small number of cases, however, monozygotic twins differ in their karyotype and even their gender if a mutation involves a gender-determining locus [57]. Finally, having separate placentas is often taken as indicating dizygosity, yet is also a feature in about 30% of monozygotic twins (Fig. 1). The interpretation of the data from twin studies also deserves careful scrutiny. Some epigenetic modifications, such as Xchromosome inactivation, may occur long before embryonic development and even before the twinning event [58]. This possibility may explain the similar X-chromosome inactivation patterns in female monozygotic twin pairs discordant for lupus. Therefore, one cannot conclude that epigenetic factors have no influence in the development of lupus, as well demonstrated by numerous studies [11,25]. The marked differences in results between dizygotic twins and non-twin siblings supports a substantial influence of prenatal environmental factors on disease development. Lessons
might therefore be learned from concomitantly examining data on non-twin siblings and on twins. 10. Is being a twin a risk factor per se? Compared to singletons, twins often have lower birth weights and restricted intra-uterine resources. As demonstrated elegantly by two recent reviews, our make-up at conception [59] and our development in utero [60] shape our response patterns to stress and influence our risk of developing diseases in childhood and adulthood. Thus, a legitimate question is whether being a twin influences the risk of developing IJDs. Twin status may be difficult to determine. In vanishing twin syndrome, one twin disintegrates in utero within a first week after conception. This syndrome is more common than previously thought: according to a recent study, it occurs in 2/1000 pregnancies [61]. Our group reported a case in which cells from a female vanished twin persisted in a 40-year-old male with a scleroderma-like disease and clinical features reminiscent of graft-versus-host disease [62]. The long-term persistence of twin microchimerism might be involved in the development of “autoimmune” IJDs. 11. Conclusions and future prospects Twin studies provide insights into the respective roles for genetic and environmental factors in IJDs and other health conditions. They are now increasingly used to investigate epigenetic variations responsible for differences in expression profiles (microarrays and next-generation RNA sequencing [RNA-Seq]). However, thorough familiarity with their limitations and biases is central to a discerning interpretation of twin study results. Genetic and environmental factors are closely intertwined: genetics influence the phenotype within a given environment. Disclosure of interest The author declares that he has no competing interest. Acknowledgments I am grateful to Jean Roudier and Isabelle Auger, at Inserm UMRs1097 in Marseille, France, for making valuable corrections to this manuscript. References [1] Brewerton DA, Hart FD, Nicholls A, et al. Ankylosing spondylitis and HL-A 27. Lancet 1973;1:904–7. [2] Schlosstein L, Terasaki PI, Bluestone R, et al. High association of an HL-A antigen. W27, with ankylosing spondylitis. N Engl J Med 1973;288:704–6. [3] Gregersen PK, Silver J, Winchester RJ. The shared epitope hypothesis. An approach to understanding the molecular genetics of susceptibility to rheumatoid arthritis. Arthritis Rheum 1987;30:1205–13. [4] Stanford SM, Bottini N. PTPN22: the archetypal non-HLA autoimmunity gene. Nat Rev Rheum 2014;10:602–11. [5] Apel M, Uebe S, Bowes J, et al. Variants in RUNX3 contribute to susceptibility to psoriatic arthritis, exhibiting further common ground with ankylosing spondylitis. Arthritis Rheum 2013;65:1224–31. [6] Vineis P, Pearce NE. Genome-wide association studies may be misinterpreted: genes versus heritability. Carcinogenesis 2011;32:1295–8. [7] Bortolus R, Parazzini F, Chatenoud L, et al. The epidemiology of multiple births. Human Reprod Update 1999;5:179–87. [8] Smits J, Monden C. Twinning across the developing world. PLoS One 2011;6:e25239. [9] Heath AC, Nyholt DR, Neuman R, et al. Zygosity diagnosis in the absence of genotypic data: an approach using latent class analysis. Twin Res 2003;6:22–6. [10] Yang MJ, Tzeng CH, Tseng JY, et al. Determination of twin zygosity using a commercially available STR analysis of 15 unlinked loci and the gender-determining marker amelogenin – a preliminary report. Hum Reprod 2006;21:2175–9. [11] Picascia A, Grimaldi V, Pignalosa O, et al. Epigenetic control of autoimmune diseases: from bench to bedside. Clin Immunol 2015;157:1–15.
Please cite this article in press as: Lambert NC. How twin studies help to understand inflammatory joint disease. Joint Bone Spine (2016), http://dx.doi.org/10.1016/j.jbspin.2016.02.008
G Model BONSOI-4340; No. of Pages 7
ARTICLE IN PRESS N.C. Lambert / Joint Bone Spine xxx (2016) xxx–xxx
[12] Jarvinen P. Occurrence of ankylosing spondylitis in a nationwide series of twins. Arthritis Rheum 1995;38:381–3. [13] Brown MA, Kennedy LG, MacGregor AJ, et al. Susceptibility to ankylosing spondylitis in twins: the role of genes. HLA, and the environment. Arthritis Rheum 1997;40:1823–8. [14] Pedersen OB, Svendsen AJ, Ejstrup L, et al. Ankylosing spondylitis in Danish and Norwegian twins: occurrence and the relative importance of genetic vs. environmental effectors in disease causation. Scand J Rheumatol 2008;37: 120–6. [15] Block SR, Winfield JB, Lockshin MD, et al. Studies of twins with systemic lupus erythematosus. A review of the literature and presentation of 12 additional sets. Am J Med 1975;59:533–52. [16] Grennan DM, Parfitt A, Manolios N, et al. Family and twin studies in systemic lupus erythematosus. Dis Markers 1997;13:93–8. [17] Block SR. A brief history of twins. Lupus 2006;15:61–4. [18] Javierre BM, Fernandez AF, Richter J, et al. Changes in the pattern of DNA methylation associate with twin discordance in systemic lupus erythematosus. Genome Res 2010;20:170–9. [19] Engel ME, Stander R, Vogel J, et al. Genetic susceptibility to acute rheumatic fever: a systematic review and meta-analysis of twin studies. PloS One 2011;6:e25326. [20] Lai NS, Chou JL, Chen GC, et al. Association between cytokines and methylation of SOCS-1 in serum of patients with ankylosing spondylitis. Mol Biol Rep 2014;41:3773–80. [21] Toussirot E, Abbas W, Khan KA, et al. Imbalance between HAT and HDAC activities in the PBMCs of patients with ankylosing spondylitis or rheumatoid arthritis and influence of HDAC inhibitors on TNF alpha production. PloS One 2013;8:e70939. [22] Lai NS, Yu HC, Chen HC, et al. Aberrant expression of microRNAs in T cells from patients with ankylosing spondylitis contributes to the immunopathogenesis. Clin Exp Immunol 2013;173:47–57. [23] Weber U, Pfirrmann CW, Kissling RO, et al. Early spondyloarthritis in an HLA-B27-positive monozygotic twin pair: a highly concordant onset, sites of involvement, and disease course. J Rheumatol 2008;35:1464–6. [24] Bogdanos DP, Smyk DS, Rigopoulou EI, et al. Twin studies in autoimmune disease: genetics, gender and environment. J Autoimmun 2012;38:J156–69. [25] Meroni PL, Penatti AE. Epigenetics and systemic lupus erythematosus: unmet needs. Clin Rev Allergy Immunol 2015, http://dx.doi.org/10.1007/ s12016-015-8497-4. [26] Huang DJ, Zimmermann BG, Holzgreve W, et al. Improvement of methods for the isolation of cell-free fetal DNA from maternal plasma: comparison of a manual and an automated method. Ann N Y Acad Sci 2006;1075:308–12. [27] Clemens LE, Albert E, Ansell BM. Sibling pairs affected by chronic arthritis of childhood: evidence for a genetic predisposition. J Rheumatol 1985;12: 108–13. [28] Prahalad S, Ryan MH, Shear ES, et al. Juvenile rheumatoid arthritis: linkage to HLA demonstrated by allele sharing in affected sibpairs. Arthritis Rheum 2000;43:2335–8. [29] Prahalad S, Ryan MH, Shear ES, et al. Twins concordant for juvenile rheumatoid arthritis. Arthritis Rheum 2000;43:2611–2. [30] Feghali-Bostwick C, Medsger Jr TA, Wright TM. Analysis of systemic sclerosis in twins reveals low concordance for disease and high concordance for the presence of antinuclear antibodies. Arthritis Rheum 2003;48:1956–63. [31] Lawrence JS. Heberden oration, 1969. Rheumatoid arthritis – nature or nurture? Ann Rheum Dis 1970;29:357–79. [32] Aho K, Koskenvuo M, Tuominen J, et al. Occurrence of rheumatoid arthritis in a nationwide series of twins. J Rheumatol 1986;13:899–902. [33] Bellamy N, Duffy D, Martin N, et al. Rheumatoid arthritis in twins: a study of aetiopathogenesis based on the Australian Twin Registry. Ann Rheum Dis 1992;51:588–93. [34] Silman AJ, MacGregor AJ, Thomson W, et al. Twin concordance rates for rheumatoid arthritis: results from a nationwide study. Br J Rheumatol 1993;32:903–7. [35] MacGregor AJ, Snieder H, Rigby AS, et al. Characterizing the quantitative genetic contribution to rheumatoid arthritis using data from twins. Arthritis Rheum 2000;43:30–7. [36] Svendsen AJ, Holm NV, Kyvik K, et al. Relative importance of genetic effects in rheumatoid arthritis: historical cohort study of Danish nationwide twin population. BMJ 2002;324:264–6. [37] van der Woude D, Houwing-Duistermaat JJ, Toes RE, et al. Quantitative heritability of anticitrullinated protein antibody-positive and
[38]
[39]
[40] [41]
[42]
[43]
[44]
[45]
[46] [47]
[48]
[49] [50] [51] [52]
[53]
[54]
[55]
[56]
[57] [58]
[59] [60] [61]
[62]
7
anti-citrullinated protein antibody-negative rheumatoid arthritis. Arthritis Rheum 2009;60:916–23. Svendsen AJ, Kyvik KO, Houen G, et al. On the origin of rheumatoid arthritis: the impact of environment and genes – a population-based twin study. PLoS One 2013;8:e57304. Hensvold AH, Magnusson PK, Joshua V, et al. Environmental and genetic factors in the development of anticitrullinated protein antibodies (ACPAs) and ACPApositive rheumatoid arthritis: an epidemiological investigation in twins. Ann Rheum Dis 2015;74:375–80. Krishnan E, Lessov-Schlaggar CN, Krasnow RE, et al. Nature versus nurture in gout: a twin study. Am J Med 2012;125:499–504. Pedersen OB, Svendsen AJ, Ejstrup L, et al. On the heritability of psoriatic arthritis. Disease concordance among monozygotic and dizygotic twins. Ann Rheum Dis 2008;67:1417–21. Nanki T, Kohsaka H, Mizushima N, et al. Genetic control of T cell receptor BJ gene expression in peripheral lymphocytes of normal and rheumatoid arthritis monozygotic twins. J Clin Invest 1996;98:1594–601. Mizushima N, Kohsaka H, Nanki T, et al. HLA-dependent peripheral T cell receptor (TCR) repertoire formation and its modification by rheumatoid arthritis (RA). Clin Exp Immunol 1997;110:428–33. Jawaheer D, Thomson W, MacGregor AJ, et al. “Homozygosity” for the HLA-DR shared epitope contributes the highest risk for rheumatoid arthritis concordance in identical twins. Arthritis Rheum 1994;37:681–6. Balandraud N, Picard C, Reviron D, et al. HLA-DRB1 genotypes and the risk of developing anti citrullinated protein antibody (ACPA) positive rheumatoid arthritis. PLoS One 2013;8:e64108. Svendsen AJ, Hjelmborg JV, Wiik A, et al. Autoantibodies in twins discordant for rheumatoid arthritis. Ann Rheum Dis 2011;70:708–9. Svendsen AJ, Kyvik KO, Houen G, et al. Newborn infant characteristics and risk of future rheumatoid arthritis: a twin-control study. Rheumatol Int 2014;34:523–8. Vitart V, Rudan I, Hayward C, et al. SLC2A9 is a newly identified urate transporter influencing serum urate concentration, urate excretion and gout. Nat Genet 2008;40:437–42. Scofield RH, Kurien BT, Reichlin M. Immunologically restricted and inhibitory anti-Ro/SSA in monozygotic twins. Lupus 1997;6:395–8. Bolstad AI, Haga HJ, Wassmuth R, et al. Monozygotic twins with primary Sjogren’s syndrome. J Rheumatol 2000;27:2264–6. Lee WS, Yoo WH. Primary Sjogren’s syndrome in monozygotic twins. Int J Rheum Dis 2014;17:578–9. Jawaheer D, MacGregor AJ, Gregersen PK, et al. Unexpected HLA haplotype sharing in dizygotic twin pairs discordant for rheumatoid arthritis. J Med Genet 1996;33:1015–8. Kasl SV, Cobb S. The intrafamilial transmission of rheumatoid arthritis. V. Differences between rheumatoid arthritics and controls on selected personality variables. J Chronic Dis 1969;22:239–58. Sprangers MA, Bartels M, Veenhoven R, et al. Which patient will feel down, which will be happy? The need to study the genetic disposition of emotional states. Qual Life Res 2010;19:1429–37. Duarte AP, Marques CD, Bortoluzzo AB, et al. [Epidemiologic profile of juvenileonset compared to adult-onset spondyloarthritis in a large Brazilian cohort]. Rev Bras Reumatol 2014;54:424–30. Aletaha D, Neogi T, Silman AJ, et al. 2010 rheumatoid arthritis classification criteria: an American College of Rheumatology/European League Against Rheumatism collaborative initiative. Arthritis Rheum 2010;62:2569–81. Cutler TL, Murphy K, Hopper JL, et al. Why accurate knowledge of zygosity is important to twins. Twin Res Hum Genet 2015;18:298–305. Trejo V, Derom C, Vlietinck R, et al. X chromosome inactivation patterns correlate with fetal-placental anatomy in monozygotic twin pairs: implications for immune relatedness and concordance for autoimmunity. Mol Med 1994;1:62–70. Lane M, Robker RL, Robertson SA. Parenting from before conception. Science 2014;345:756–60. Arck PC, Hecher K. Fetomaternal immune cross-talk and its consequences for maternal and offspring’s health. Nat Med 2013;19:548–56. Curnow KJ, Wilkins-Haug L, Ryan A, et al. Detection of triploid, molar, and vanishing twin pregnancies by a single-nucleotide polymorphism-based non invasive prenatal test. Am J Obstet Gynecol 2014. de Bellefon LM, Heiman P, Kanaan SB, et al. Cells from a vanished twin as a source of microchimerism 40 years later. Chimerism 2010;1:56–60.
Please cite this article in press as: Lambert NC. How twin studies help to understand inflammatory joint disease. Joint Bone Spine (2016), http://dx.doi.org/10.1016/j.jbspin.2016.02.008
ARTICLE IN PRESS
BONSOI-4340; No. of Pages 7
Joint Bone Spine xxx (2016) xxx–xxx
Available online at
ScienceDirect www.sciencedirect.com
Review
How twin studies help to understand inflammatory joint disease Nathalie C. Lambert Inserm UMRs1097, parc scientifique de Luminy, 163, avenue de Luminy, case 939, Bâtiment TPR2 Inserm, Entrée A, 1er étage, 13288 Marseille cedex 09, France
a r t i c l e
i n f o
Article history: Accepted 6 January 2016 Available online xxx Keywords: Inflammatory joint disease Twins Monozygotic Dizygotic Heritability
a b s t r a c t Inflammatory joint disease (IJD) is a group of conditions that target the joints and periarticular structures. The contribution of genetic factors to these conditions is often less than 50%, suggesting a major role for environmental influences. Twin studies are the best means of assessing the role for genetic factors in IJD. Conclusive evidence has been provided by a few studies in vast samples of monozygotic and dizygotic twins, with only one twin in each pair having IJD. These studies have been most successful in ankylosing spondylitis and psoriatic arthritis. The other IJDs have proven more difficult to evaluate. This review demonstrates that genetic and environmental factors are inextricably linked and that ascribing IJDs to one or the other is misguided. Awareness of the limitations and possible sources of bias in twin studies is important when seeking to understand the development of these complex diseases. © 2016 Published by Elsevier Masson SAS on behalf of Société française de rhumatologie.
1. Introduction Inflammatory joint diseases (IJD) are conditions that target the joints and periarticular structures (bone, muscles, and tendons). The most common IJD is spondyloarthritis, i.e., ankylosing spondylitis (AS) and psoriatic arthritis (PsA). Features shared by the various forms of spondyloarthritis include inflammation of the spine and pelvis, peripheral arthritis and, to variable degrees, involvement of other organs. The second in frequency is rheumatoid arthritis (RA), which is the most common autoimmune joint disease. Polymyalgia rheumatica (PMR) is less common and chiefly affects individuals older than 60 years of age. The IJD family also includes crystal-deposition and autoinflammatory arthritides such as gout and articular chondrocalcinosis, which are relatively common. Finally, some arthritides related to microorganisms lie in the gray zone between infection and inflammation: a viral or bacterial agent can trigger an inflammatory response, as occurs in reactive arthritis. The immune response is often disproportionate, similar to that seen in rheumatic fever. Finally, the joints are sometimes involved in various systemic autoimmune disorders such as scleroderma, dermatomyositis, and systemic lupus erythematosus (SLE). Many genetic factors are involved in these diseases. The human leukocyte antigen (HLA) genes have long been identified as the main risk factors for most IJDs. For instance, HLA-B*27 is associated with AS [1,2] and some of the HLA-DRB1 alleles (e.g., *0101, *0401,
E-mail address: [email protected]
*0404, *0405, and *1001) that carry an amino acid motif known as the shared epitope are associated with RA [3]. Genome-wide association studies indicate that numerous other genes are risk factors; however, they convey far lower levels of risk, and some polymorphisms are associated with more than one IJD. Thus, PTPN22, which encodes the protein tyrosine phosphatase non-receptor type 22, is associated with many autoimmune diseases [4]. Recent research indicates an association linking PsA to variants in RUNX3, a previously identified AS-susceptibility gene [5]. In most IJDs, concordance between monozygotic twins is lower than 50%. This fact supports a strong role for environmental factors. Twin studies are the best means of assessing the contribution of genetic factors to the development of a disease. 2. Twins: identical versus fraternal Dizygotic or fraternal twins are produced when two ova are independently fertilized by two sperm cells. Each twin develops in its own amniotic sac and has its own placenta: the pregnancy is dichorionic and diamniotic (Fig. 1). Very rarely, dizygotic twins are produced during different ovulations and therefore have different gestational ages. About three-quarters of all twin pairs are dizygotic. Monozygotic twins are formed from a single ovum fertilized by a single sperm cell and account for about one-quarter of all twins. Depending on the stage at which the zygote divides into two embryos, there may be two placentas and amniotic sacs (dichorionic diamniotic, 29%), a single placenta but two amniotic sacs (monochorionic diamniotic, 70%), or a single placenta and a single
http://dx.doi.org/10.1016/j.jbspin.2016.02.008 1297-319X/© 2016 Published by Elsevier Masson SAS on behalf of Société française de rhumatologie.
Please cite this article in press as: Lambert NC. How twin studies help to understand inflammatory joint disease. Joint Bone Spine (2016), http://dx.doi.org/10.1016/j.jbspin.2016.02.008
G Model BONSOI-4340; No. of Pages 7
ARTICLE IN PRESS N.C. Lambert / Joint Bone Spine xxx (2016) xxx–xxx
2
Fig. 1. Development of monozygotic and dizygotic twins.
amniotic sac (monochorionic monoamniotic, 1%) (Fig. 1). Late separation of the zygote is an exceedingly rare event that results in conjoined twins. Twins may differ in their development, with one weighing more than the other. In vanishing twin syndrome, one twin fails to develop and disintegrates [6]. Twinning occurs in about 13/1000 pregnancies overall. However, the frequency of twinning varies with several maternal characteristics including age, ethnicity, nutrition, and fertility [7,8]. The frequency of twinning is highest in Benin (28/1000) and lowest in Asia (< 9/1000) [8]. Differences across ethnic groups occur chiefly for dizygotic twins. The frequency of monozygotic twinning is almost identical throughout the world, at about 4/1000 pregnancies [7]. Fertility treatments also tend to increase the frequency of twinning.
occupation-related exposures to chemicals, the living environment, and lifestyle factors (e.g., alcohol use, smoking, diet, and birth control methods). Environmental factors may act via epigenetic mechanisms to modify the susceptibility of an individual to the development of a disease. Epigenetic factors are molecular mechanisms that induce reversible changes in gene expression levels, in the absence of alterations in the DNA sequence. They include DNA methylation, histone deacetylation, and the expression of small noncoding RNA fragments known as micro-RNAs [11]. Epigenetic factors allow changes in the expression of certain genes depending on the individual’s environment. The influence of genetic factors can be assessed by computing the concordance rate for the disease between monozygotic twins and comparing this value to that in dizygotic twins.
3. Determining the zygosity of twins Determining whether two same-sex twins are dizygotic or monozygotic relies to a variable degree on an empirical approach. Some studies used questionnaires to collect information from twins about their physical resemblance and instances of one having been mistaken for the other. This similarity-based method is practical in vast studies of several thousand-twin pairs. However, in 5% to 10% of cases, the responses are not assessable or not interpretable. Among assessable data, most are fairly reliable: comparisons with genetic testing for zygosity showed less than 5% of classification errors [9]. Nevertheless, the most reliable method is genetic analysis of polymorphisms or short tandem repeats [10]. 4. Evaluation of the genetic and environmental contributions to a disease The development of a disease is multifactorial. Some diseases are characterized by strong heritability, which indicates a major role for genetic factors. Nevertheless, environmental factors may contribute to the development of a disease to a similar extent as genetic factors. Such environmental factors include
5. Computing concordance rates Two main types of concordance rate are used, the pairwise concordance rate and the probandwise concordance rate. The pairwise concordance rate is the proportion of twin pairs with at least one affected twin in which the other twin is also affected. This statistical parameter is thus the proportion of affected pairs in which both twins are affected (Fig. 2A). The probandwise concordance rate (or rate per affected individuals) is the proportion of affected individuals among co-twins of previously identified affected twins. This parameter measures the risk of having the disease among individuals who have an affected twin (Fig. 2B). A higher concordance rate for a disease among monozygotic twins than among dizygotic twins supports a major role for genetic factors in the development of the disease. Higher concordance among dizygotic twins suggests a strong role for environmental factors. A disease caused solely by genetic factors would consistently affect both members of pairs of monozygotic twins. In reality, concordance rates among monozygotic twins are only about 50%, indicating a strong influence of nongenetic factors (Fig. 2).
Please cite this article in press as: Lambert NC. How twin studies help to understand inflammatory joint disease. Joint Bone Spine (2016), http://dx.doi.org/10.1016/j.jbspin.2016.02.008
G Model
ARTICLE IN PRESS
BONSOI-4340; No. of Pages 7
N.C. Lambert / Joint Bone Spine xxx (2016) xxx–xxx
3
Fig. 2. Methods for computing concordance rates for a disease among monozygotic and dizygotic twins.
Heritability is another parameter used in statistical analyses of the relative roles for genetic and environmental factors in the development of a disease within a predefined population. Several modeling techniques are available. The ACE model, for instance, uses three components: the additive (A) effects of alleles at a given locus, i.e., the role for the genotype; the common environment (C), shared by both twins; and the unique environment (E), i.e., the environmental factors related to the personal experience and lifestyle of each individual.
6. Inflammatory joint diseases with a strong genetic component 6.1. Ankylosing spondylitis (AS) The first large twin study of AS was done in Finland and reported in 1995. Subsequently, twin studies were performed in Denmark and Norway (Table 1). Concordance rates were consistently higher in monozygotic than in dizygotic twin pairs, indicating a strong genetic component. Thus, the HLA-B*27 allele, found in 90% of patients with AS, contributes to the development of the disease. However, other genes may be involved also, since when considering only twin pairs carrying HLA-B*27, concordance rates for AS are 50% among monozygotic twins compared to only 20% among dizygotic twins [12]. Nevertheless, concordance rates among dizygotic twins are higher for AS than for most of the other IJDs (Table 1). This fact supports an influence of environmental factors. A few studies have identified epigenetic variations in patients with AS [20–22]. However, to the best of my knowledge, epigenetic factors have not been studied in monozygotic twins discordant for AS. Data from case reports also provide useful insights. Thus, in a monozygotic twin pair with AS, the onset, sites involved, and course of the disease were closely similar between the two patients [23]. Thus, genetic factors influence not only the susceptibility to AS but also the phenotypic expression of the disease.
6.2. Systemic lupus erythematosus (SLE) Among systemic autoimmune diseases that can involve the joints, SLE has the highest heritability, although the level is lower than for AS [24]. Overall, concordance rates are higher among monozygotic than dizygotic twins, ranging from 11% to 40% [24]. However, these rates leave considerable room for effects of environmental and epigenetic factors. Genetic-epigenetic interactions exert a pivotal influence on the susceptibility to SLE, as shown in two recent review articles [11,25]. Few studies have tested the hypothesis that monozygotic twins discordant for SLE have different epigenetic factors. One study demonstrated differences in methylation profiles between affected individuals and their unaffected twins [18]. Another study, however, found no differences in X chromosome inactivation patterns within monozygotic twin pairs discordant for SLE [26].
6.3. Juvenile idiopathic arthritis (JIA) Juvenile idiopathic arthritis (JIA) is a heterogeneous group of IJDs that include oligoarticular JIA, polyarticular JIA (with or without rheumatoid factors), enthesitis-related arthritis, juvenile psoriatic arthritis, and undifferentiated arthritis. In several studies, non-twin siblings concordant for JIA shared larger numbers of HLA haplotypes than did non-twin siblings discordant for the disease [27,28], suggesting a role for the HLA genes. This evidence for a genetic susceptibility to JIA was confirmed by a study of twins concordant for the disease [29]. Of 14 twin pairs from the publication, 11 had zygosity test results available and were monozygotic; 3 had no zygosity test results, including 1 pair discordant for gender. The twin pairs were identified from a registry of siblings affected with JIA and not from a population-based registry. Consequently, the concordance rate cannot be computed. Nevertheless, that all 11 pairs with zygosity tests were monozygotic is worthy of note. Furthermore, age at disease onset was very similar in affected twins,
Please cite this article in press as: Lambert NC. How twin studies help to understand inflammatory joint disease. Joint Bone Spine (2016), http://dx.doi.org/10.1016/j.jbspin.2016.02.008
G Model BONSOI-4340; No. of Pages 7
ARTICLE IN PRESS N.C. Lambert / Joint Bone Spine xxx (2016) xxx–xxx
4
Table 1 Recorded or estimated concordance rates in twin studies of inflammatory joint diseases strongly related to genetic factors. Disease
Total n of pairs: MZ + DZ
% pairwisea concordance MZ DZ
% probandwise concordance MZ DZ
Source, country; reference
AS
26: 6 MZ, 20 DZ 15 DZB27+ 40: 8 MZ, 32 DZ
67
26
National Twin Registry, Finland; Järvinen, 1995 [12]
86
22
40
8
10HLA : 4 MZ, 6 DZ 247: 151 MZ, 96 DZ
50 50HLA 75 75HLA 25 25HLA 57 71ANA+ 25 40
73 83ANA+ 33 57
0 0ANA+ 0 8
5, all MZ discordant for LED 14; 11 MZ, 1 DZ, 2 ND
Significant differenceb No statistical tests
ND
ND
ND
ND
National Rheumatic Diseases database, UK; Brown et al., 1997 [13] 3 nationwide surveys, Denmark, Norway; Pedersen, et al., 2008 [14] Literature review with 12 added pairs, USA; Block et al., 1975 [15] Australia; Grennan et al., 1997 [16] Review of published case-reports of affected twins, USA; Block, 2006 [17] Clinical center of the National institutes for health (NIH), USA; Javierre, et al., 2010 [18] Registry of 118 pairs with both twins affected
435: 175 MZ, 260 DZ
28[19–100]
44[31–100]
11[0–67]
27: 4 MZ, 23 DZ SLE
Methylation 807 promoters JIA All concordant RhF
s
12 : 7 MZ, 3 DZ
15 20HLA 13 27HLA 4 10HLA 0 0ANA+ 0 4
6[0–50]
Metaanalysis of 6 studies published between 1933 and 1964; Engel et al., 2011 [19]
AS: ankylosing spondylitis; SLE: systemic lupus erythematosus; JIA: juvenile idiopathic arthritis; RhF: rheumatic fever; MZ: monozygotic; DZ: dizygotic; ND: not determined; ANA: antinuclear antibodies. a The concordance percentages are rounded up to the next integer. The superscripts indicate the following: (HLA) : HLA matching of co-twins (e.g., for HLA-B*27 when studying AS); (S) : matching on sex; and numbers (e.g., [0–100] ): variations in concordance rates across studies. The subscripts indicate the factor for which concordance was determined, when this factor was not the disease (e.g., antinuclear antibodies: ANA+ ). b The twins were discordant for the disease and were examined for differences in epigenetic factors such as DNA methylation.
with a mean difference of 5–6 months, compared to nearly 5 years for affected non-twin siblings [29].
7. Inflammatory joint disease (IJDs) with a limited genetic component 7.1. Rheumatoid arthritis (RA) The role for genetic factors in RA has been estimated at 16% to 60%. Over 30 genes are associated with RA. Nevertheless, RA is one of the autoimmune diseases, together with systemic sclerosis [30], for which concordance rates in monozygotic twins are fairly low, from 12% to 21% [24], albeit higher than in dizygotic twins (0% to 4%, Table 2). Twin studies have established that the T-cell receptor repertoire is chiefly under genetic control. For instance, strongly similar repertoires have been demonstrated between monozygotic twins discordant for RA [42]. HLA genes play a strong role in this genetic control [43]. Nevertheless, distinctive variations in the repertoire have been demonstrated in patients with RA, suggesting that the disease itself may affect the repertoire [43]. The main genetic factor in RA is the shared epitope. In an elegant study of 91 monozygotic twin pairs, the concordance rate for RA was higher in twins with than without the shared epitope (18% versus 5%) [44]. Even more striking was the finding of a higher concordance rate among twins homozygous compared to heterozygous for the shared epitope (27% versus 13%). Finally, genotypes conferring susceptibility to RA, i.e., heterozygous HLADRB1 genotypes (e.g., DRB1*0401/*0404) were associated with higher concordance rates compared to homozygous genotypes (e.g., DRB1*0401/*0401) (38% versus 10%) [44]. This genotypedependent variation in the risk of RA was recently evaluated in detail by our group [45]. Genetic factors may play a particularly strong role in RA development in patients who already produce anticitrullinated protein antibodies (ACPAs). However, in a study of 148 twin pairs, heritability was nearly identical (about 66%) for ACPA+ and ACPA− disease, indicating a greater role of genetic factors in ACPA− RA than previously thought [37]. Environmental factors may contribute strongly to the development of auto-antibodies themselves (without
development of the disease) [46] and, according to one study, this contribution may be greater than that of genetic factors [39]. Twin studies can be used to evaluate potential risk factors without confounding by genetic factors (when monozygotic twin pairs are studied) or by a number of environmental factors (gestational age at birth, body mass index of the mother, and socio-economic influences). For instance, a study was conducted to determine whether lower birth weight was associated with a greater risk of developing RA compared to the monozygotic twin with a higher birth weight [47]. The results showed no effect of birth weight. Birth order was the only factor associated with the risk of RA [47]. 7.2. Systemic sclerosis Most of the data from twins with systemic sclerosis come from anecdotal cases of twins concordant for the disease. There is, however, a study of 42 twin pairs, in which concordance among monozygotic twins was low for systemic sclerosis but high for antinuclear antibodies [43]. In contrast, Raynaud’s phenomenon without systemic sclerosis has high heritability, although it is the inaugural symptom in most patients with systemic sclerosis [44]. 7.3. Gout Gout is among the most common IJDs. Hyperuricemia often precedes symptom onset, but only 10% of patients with hyperuricemia experience gouty attacks [48]. A single study, from the US, has evaluated the genetic and environmental factors involved in hyperuricemia and gout [40]. However, it included a large number of twin pairs (n = 514). The results establish that hyperuricemia is somewhat influenced by dietary and other lifestyle factors but depends chiefly on genetic factors, whereas gout is chiefly dependent on environmental factors. This finding implies that many cases of gout could be prevented. 8. Inflammatory joint diseases (IJDs) for which twin-study data are scarce For several IJDs, the number of twin pairs studied to date is too small to provide conclusive evidence about the influence of genetic factors.
Please cite this article in press as: Lambert NC. How twin studies help to understand inflammatory joint disease. Joint Bone Spine (2016), http://dx.doi.org/10.1016/j.jbspin.2016.02.008
G Model
ARTICLE IN PRESS
BONSOI-4340; No. of Pages 7
N.C. Lambert / Joint Bone Spine xxx (2016) xxx–xxx
5
Table 2 Recorded or estimated concordance rates in twin studies of inflammatory joint diseases in which genetic factors play only a small role. Disease
Total n of pairs: MZ + DZ
% pairwisea concordance MZ DZ
% probandwisea concordance MZ DZ
Source, country; reference
RA
285: 97 MZ, 188 DZ
32[0–67]
9[0–29]
48[0–80]
16[0–44]
261: not specified
12 9A/S 21
4 3A/S 0
22
7
25
0
15 17S 12 14S 0
4 4S 4 4S 6
27
7
22
8
0
9
16 17ACPA+ 11ACPA− No differenceb Methylation 807 promoters 6
4 4ACPA+ 0ACPA−
27
7
Metaanalysis of 7 studies published from 1938 to 1968; Lawrence, 1970 [31] Finnish Twin Cohort linked with the health insurance database; Aho et al., 1986 [32] Questionnaire, Australian Twin Registry; Bellamy et al., 1992 [33] Nationwide media campaign, prospective survey of rheumatologists, UK; Silman et al., 1993 [34] 2 nationwide studies, in Finland and the UK, respectively; MacGregoret al., 2000 [35], [34] Questionnaires, nationwide twin population in Denmark; Svendsen et al., 2002 [36] Nationwide twin study, UK; van der Woude et al., 2009 [37]
23: 14 MZ, 9 DZ 203: 91 MZ, 112 DZ 246: 73 MZ, 173 DZ 49: 13 MZ, 36 DZ 148: 64 MZ, 84 DZ
5: 5 MZ discordant for RA 155: 32 MZ, 77 DZS and 46 DZSO 117ACPA+ : 21 MZ, 59 DZS and 37 DZSO 380S : 151 MZ, 229 DZS SSc
42: 24 MZ, 18 DZ
Gout PsA
514: 253 MZ, 261 DZ 36: 10 MZ, 26 DZ
10 4ACPA+ 4ACPA+ RA+ 2ACPA+ RA− 4 90ANA 12 10
Clinical center of the National institutes of health (NIH) Javierre, et al., 2010 [18] 5S 2SO 7S 3SO 3ACPA+ 2ACPA+ RA+ 1ACPA+ RA− 6 40ANA 12 4
8ACPA+ 8ACPA+ RA+ 4ACPA+ RA− 8
6S 2SO 9S 3SO 5ACPA+ 4ACPA+ RA+ 3ACPA+ RA− 11
21 18
21 7
9 14
Questionnaires, nationwide twin population in Denmark; Svendsen et al., 2013 [38]
Swedish Twin Registry; Hensvold et al., 2015 [39]
Invitations in patient-group newsletters and to rheumatologists, USA; Feghali-Bostwick et al., 2003 [30] Veterans Affairs Twin registry, USA; Krishnan et al., 2012 [40] Cohort of Danish twins identified by a nationwide questionnaire; Pedersen et al., 2008 [41]
RA: rheumatoid arthritis; SSc: systemic sclerosis; PsA: psoriatic arthritis; MZ: monozygotic; DZ: dizygotic; ACPA: anticitrullinated peptide antibodies; RA: rheumatoid arthritis; ANA: antinuclear antibodies. a The concordance percentages are rounded up to the next integer. The superscripts indicate the following: (S) : matching on sex; (SO ): co-twins of different sexes; (A/S ): matching on both age and sex; (ACPA ): matching on anticitrullinated protein antibody status; (ANA ): matching on antinuclear antibody status; numbers (e.g., [0–100] ): the variations in concordance rates across studies. The subscripts indicate the factor for which concordance was determined, when this factor was not the disease (e.g., anticitrullinated antibodies, ACPA+ and ACPA− ). b The twins were discordant for the disease and were examined for differences in epigenetic factors such as DNA methylation.
8.1. Psoriatic arthritis The only study in PsA showed little difference in concordance rates between monozygotic and dizygotic twins (1/10 and 1/26 pairs, respectively), indicating a need to search for nongenetic factors implicated in this disease [41]. To the best of my knowledge, epigenetic factors have not been investigated in cohorts of twin pairs having one twin with PsA. The above-mentioned study had a sample size of only 36 twin pairs, which limited its statistical power. Thus, the influence of genetic factors on the development of PsA in patients with psoriasis remains incompletely elucidated. 8.2. Sjögren’s syndrome Although several instances of Sjögren’s disease clustering within families have been reported, the presence of this disease in monozygotic co-twins is rare. In the few studies of monozygotic twins, co-twins usually had similar phenotypes, symptoms, and serological findings [49,50]. Another study, however, showed opposite results, indicating a role for environmental factors [51]. There is a pressing need for studies in large cohorts of twins with Sjögren’s syndrome. 9. Limitations of twin studies 9.1. Monozygotic versus dizygotic: an oversimplified model Monozygotic twins, although considered genetically identical, may have small genetic differences due to rare mutations that occur
immediately after the blastocyst splits into two embryos. These mutations are then found in all the tissues and in the germ cells of only one of the twins. The differences thus produced are used for paternity tests and medicolegal investigations. Dizygotic twins are considered relevant for evaluating the contribution of the environment, as opposed to that of genetic factors. Nevertheless, HLA genes are shared more often by dizygotic twins than by siblings from different pregnancies [52]. Dizygotic twins are therefore different from non-twin siblings who share the same environment. Finally, genetics and the environment cannot be viewed as two completely separate influences, since they are tightly intertwined with each other. For instance, psychological stress, long suspected to be one of the triggers of RA [53], is a response to the environment but manifests differently depending on the genetic background [54]. 9.2. Sources of bias Concordance rates for a given disease vary across studies, for instance from 0% to 32% for RA among monozygotic twins. This considerable variability supports the existence of multiple sources of bias related to analysis methods, recruitment modalities, age of the participants, and other factors. For example, populations will differ depending on whether recruitment is via a questionnaire sent to nationwide twin registries or via office-based rheumatologists (Table 2). Similarly, when inviting twin pairs with a given disease to participate in a study, those in which both twins are affected are more likely to accept. Finally, diseases that manifest at a younger age are more
Please cite this article in press as: Lambert NC. How twin studies help to understand inflammatory joint disease. Joint Bone Spine (2016), http://dx.doi.org/10.1016/j.jbspin.2016.02.008
G Model BONSOI-4340; No. of Pages 7
ARTICLE IN PRESS N.C. Lambert / Joint Bone Spine xxx (2016) xxx–xxx
6
heavily influenced by genetic factors and less dependent on environmental factors. For instance, in a vast non-twin study of patients with spondyloarthritis in Brazil, positive HLA-B27 status was more common among patients who experienced their first symptoms before 16 years of age compared to those with adult-onset disease [55]. These patient selection criteria, particularly in twin studies, may influence the findings, yet are not consistently taken into account. Furthermore, disease onset in two co-twins may be separated by a substantial time interval.
9.3. Changes over time in diagnostic criteria The classification criteria for diseases change over time. Consequently, the disease profile in the study population may vary according to the study period, although the name used to designate the disease remains the same. In a study showing a major role for genetic factors even in ACPA− patients with RA [56], some of the included patients would not be classified as having RA based on the 2010 ACR criteria, which include the ACPA status [56]. JIA is a heterogeneous group comprising many clinical patterns (see section 6.3). However, in twin studies, the pattern of JIA in the patients is not consistently specified. Rheumatic fever provides the most striking example of how the diagnosis of a disease can change over time. Rheumatic fever is a severe complication of streptococcal tonsillitis that can result in life-threatening heart valve disease in the medium or long term. In a 2011 metaanalysis with 435 twin pairs from six studies performed between 1933 and 1964, heritability was 60% and concordance rates in monozygotic twins ranged from 31% to 100% [19]. However, these studies were conducted over 45 years ago, in industrialized countries where rheumatic fever no longer occurs and has no clear definition.
9.4. Statistical limitations For a few IJDs, the influence of genetic factors remains unclear. An example is PsA, for which the availability of a single study with only 36 twin pairs results in limited statistical power [41]. Only large cohorts can produce the statistical power needed to generate firm conclusions.
9.5. Missing data and interpretation of data Missing data is a common issue that precludes a full statistical analysis. For instance, for monozygotic twins, information may be lacking on the number of placentas and/or amniotic sacs (Fig. 1). Flawed data interpretation is another issue. For instance, twins of different genders are classified as dizygotic and those who shared a placenta as monozygotic. In a very small number of cases, however, monozygotic twins differ in their karyotype and even their gender if a mutation involves a gender-determining locus [57]. Finally, having separate placentas is often taken as indicating dizygosity, yet is also a feature in about 30% of monozygotic twins (Fig. 1). The interpretation of the data from twin studies also deserves careful scrutiny. Some epigenetic modifications, such as Xchromosome inactivation, may occur long before embryonic development and even before the twinning event [58]. This possibility may explain the similar X-chromosome inactivation patterns in female monozygotic twin pairs discordant for lupus. Therefore, one cannot conclude that epigenetic factors have no influence in the development of lupus, as well demonstrated by numerous studies [11,25]. The marked differences in results between dizygotic twins and non-twin siblings supports a substantial influence of prenatal environmental factors on disease development. Lessons
might therefore be learned from concomitantly examining data on non-twin siblings and on twins. 10. Is being a twin a risk factor per se? Compared to singletons, twins often have lower birth weights and restricted intra-uterine resources. As demonstrated elegantly by two recent reviews, our make-up at conception [59] and our development in utero [60] shape our response patterns to stress and influence our risk of developing diseases in childhood and adulthood. Thus, a legitimate question is whether being a twin influences the risk of developing IJDs. Twin status may be difficult to determine. In vanishing twin syndrome, one twin disintegrates in utero within a first week after conception. This syndrome is more common than previously thought: according to a recent study, it occurs in 2/1000 pregnancies [61]. Our group reported a case in which cells from a female vanished twin persisted in a 40-year-old male with a scleroderma-like disease and clinical features reminiscent of graft-versus-host disease [62]. The long-term persistence of twin microchimerism might be involved in the development of “autoimmune” IJDs. 11. Conclusions and future prospects Twin studies provide insights into the respective roles for genetic and environmental factors in IJDs and other health conditions. They are now increasingly used to investigate epigenetic variations responsible for differences in expression profiles (microarrays and next-generation RNA sequencing [RNA-Seq]). However, thorough familiarity with their limitations and biases is central to a discerning interpretation of twin study results. Genetic and environmental factors are closely intertwined: genetics influence the phenotype within a given environment. Disclosure of interest The author declares that he has no competing interest. Acknowledgments I am grateful to Jean Roudier and Isabelle Auger, at Inserm UMRs1097 in Marseille, France, for making valuable corrections to this manuscript. References [1] Brewerton DA, Hart FD, Nicholls A, et al. Ankylosing spondylitis and HL-A 27. Lancet 1973;1:904–7. [2] Schlosstein L, Terasaki PI, Bluestone R, et al. High association of an HL-A antigen. W27, with ankylosing spondylitis. N Engl J Med 1973;288:704–6. [3] Gregersen PK, Silver J, Winchester RJ. The shared epitope hypothesis. An approach to understanding the molecular genetics of susceptibility to rheumatoid arthritis. Arthritis Rheum 1987;30:1205–13. [4] Stanford SM, Bottini N. PTPN22: the archetypal non-HLA autoimmunity gene. Nat Rev Rheum 2014;10:602–11. [5] Apel M, Uebe S, Bowes J, et al. Variants in RUNX3 contribute to susceptibility to psoriatic arthritis, exhibiting further common ground with ankylosing spondylitis. Arthritis Rheum 2013;65:1224–31. [6] Vineis P, Pearce NE. Genome-wide association studies may be misinterpreted: genes versus heritability. Carcinogenesis 2011;32:1295–8. [7] Bortolus R, Parazzini F, Chatenoud L, et al. The epidemiology of multiple births. Human Reprod Update 1999;5:179–87. [8] Smits J, Monden C. Twinning across the developing world. PLoS One 2011;6:e25239. [9] Heath AC, Nyholt DR, Neuman R, et al. Zygosity diagnosis in the absence of genotypic data: an approach using latent class analysis. Twin Res 2003;6:22–6. [10] Yang MJ, Tzeng CH, Tseng JY, et al. Determination of twin zygosity using a commercially available STR analysis of 15 unlinked loci and the gender-determining marker amelogenin – a preliminary report. Hum Reprod 2006;21:2175–9. [11] Picascia A, Grimaldi V, Pignalosa O, et al. Epigenetic control of autoimmune diseases: from bench to bedside. Clin Immunol 2015;157:1–15.
Please cite this article in press as: Lambert NC. How twin studies help to understand inflammatory joint disease. Joint Bone Spine (2016), http://dx.doi.org/10.1016/j.jbspin.2016.02.008
G Model BONSOI-4340; No. of Pages 7
ARTICLE IN PRESS N.C. Lambert / Joint Bone Spine xxx (2016) xxx–xxx
[12] Jarvinen P. Occurrence of ankylosing spondylitis in a nationwide series of twins. Arthritis Rheum 1995;38:381–3. [13] Brown MA, Kennedy LG, MacGregor AJ, et al. Susceptibility to ankylosing spondylitis in twins: the role of genes. HLA, and the environment. Arthritis Rheum 1997;40:1823–8. [14] Pedersen OB, Svendsen AJ, Ejstrup L, et al. Ankylosing spondylitis in Danish and Norwegian twins: occurrence and the relative importance of genetic vs. environmental effectors in disease causation. Scand J Rheumatol 2008;37: 120–6. [15] Block SR, Winfield JB, Lockshin MD, et al. Studies of twins with systemic lupus erythematosus. A review of the literature and presentation of 12 additional sets. Am J Med 1975;59:533–52. [16] Grennan DM, Parfitt A, Manolios N, et al. Family and twin studies in systemic lupus erythematosus. Dis Markers 1997;13:93–8. [17] Block SR. A brief history of twins. Lupus 2006;15:61–4. [18] Javierre BM, Fernandez AF, Richter J, et al. Changes in the pattern of DNA methylation associate with twin discordance in systemic lupus erythematosus. Genome Res 2010;20:170–9. [19] Engel ME, Stander R, Vogel J, et al. Genetic susceptibility to acute rheumatic fever: a systematic review and meta-analysis of twin studies. PloS One 2011;6:e25326. [20] Lai NS, Chou JL, Chen GC, et al. Association between cytokines and methylation of SOCS-1 in serum of patients with ankylosing spondylitis. Mol Biol Rep 2014;41:3773–80. [21] Toussirot E, Abbas W, Khan KA, et al. Imbalance between HAT and HDAC activities in the PBMCs of patients with ankylosing spondylitis or rheumatoid arthritis and influence of HDAC inhibitors on TNF alpha production. PloS One 2013;8:e70939. [22] Lai NS, Yu HC, Chen HC, et al. Aberrant expression of microRNAs in T cells from patients with ankylosing spondylitis contributes to the immunopathogenesis. Clin Exp Immunol 2013;173:47–57. [23] Weber U, Pfirrmann CW, Kissling RO, et al. Early spondyloarthritis in an HLA-B27-positive monozygotic twin pair: a highly concordant onset, sites of involvement, and disease course. J Rheumatol 2008;35:1464–6. [24] Bogdanos DP, Smyk DS, Rigopoulou EI, et al. Twin studies in autoimmune disease: genetics, gender and environment. J Autoimmun 2012;38:J156–69. [25] Meroni PL, Penatti AE. Epigenetics and systemic lupus erythematosus: unmet needs. Clin Rev Allergy Immunol 2015, http://dx.doi.org/10.1007/ s12016-015-8497-4. [26] Huang DJ, Zimmermann BG, Holzgreve W, et al. Improvement of methods for the isolation of cell-free fetal DNA from maternal plasma: comparison of a manual and an automated method. Ann N Y Acad Sci 2006;1075:308–12. [27] Clemens LE, Albert E, Ansell BM. Sibling pairs affected by chronic arthritis of childhood: evidence for a genetic predisposition. J Rheumatol 1985;12: 108–13. [28] Prahalad S, Ryan MH, Shear ES, et al. Juvenile rheumatoid arthritis: linkage to HLA demonstrated by allele sharing in affected sibpairs. Arthritis Rheum 2000;43:2335–8. [29] Prahalad S, Ryan MH, Shear ES, et al. Twins concordant for juvenile rheumatoid arthritis. Arthritis Rheum 2000;43:2611–2. [30] Feghali-Bostwick C, Medsger Jr TA, Wright TM. Analysis of systemic sclerosis in twins reveals low concordance for disease and high concordance for the presence of antinuclear antibodies. Arthritis Rheum 2003;48:1956–63. [31] Lawrence JS. Heberden oration, 1969. Rheumatoid arthritis – nature or nurture? Ann Rheum Dis 1970;29:357–79. [32] Aho K, Koskenvuo M, Tuominen J, et al. Occurrence of rheumatoid arthritis in a nationwide series of twins. J Rheumatol 1986;13:899–902. [33] Bellamy N, Duffy D, Martin N, et al. Rheumatoid arthritis in twins: a study of aetiopathogenesis based on the Australian Twin Registry. Ann Rheum Dis 1992;51:588–93. [34] Silman AJ, MacGregor AJ, Thomson W, et al. Twin concordance rates for rheumatoid arthritis: results from a nationwide study. Br J Rheumatol 1993;32:903–7. [35] MacGregor AJ, Snieder H, Rigby AS, et al. Characterizing the quantitative genetic contribution to rheumatoid arthritis using data from twins. Arthritis Rheum 2000;43:30–7. [36] Svendsen AJ, Holm NV, Kyvik K, et al. Relative importance of genetic effects in rheumatoid arthritis: historical cohort study of Danish nationwide twin population. BMJ 2002;324:264–6. [37] van der Woude D, Houwing-Duistermaat JJ, Toes RE, et al. Quantitative heritability of anticitrullinated protein antibody-positive and
[38]
[39]
[40] [41]
[42]
[43]
[44]
[45]
[46] [47]
[48]
[49] [50] [51] [52]
[53]
[54]
[55]
[56]
[57] [58]
[59] [60] [61]
[62]
7
anti-citrullinated protein antibody-negative rheumatoid arthritis. Arthritis Rheum 2009;60:916–23. Svendsen AJ, Kyvik KO, Houen G, et al. On the origin of rheumatoid arthritis: the impact of environment and genes – a population-based twin study. PLoS One 2013;8:e57304. Hensvold AH, Magnusson PK, Joshua V, et al. Environmental and genetic factors in the development of anticitrullinated protein antibodies (ACPAs) and ACPApositive rheumatoid arthritis: an epidemiological investigation in twins. Ann Rheum Dis 2015;74:375–80. Krishnan E, Lessov-Schlaggar CN, Krasnow RE, et al. Nature versus nurture in gout: a twin study. Am J Med 2012;125:499–504. Pedersen OB, Svendsen AJ, Ejstrup L, et al. On the heritability of psoriatic arthritis. Disease concordance among monozygotic and dizygotic twins. Ann Rheum Dis 2008;67:1417–21. Nanki T, Kohsaka H, Mizushima N, et al. Genetic control of T cell receptor BJ gene expression in peripheral lymphocytes of normal and rheumatoid arthritis monozygotic twins. J Clin Invest 1996;98:1594–601. Mizushima N, Kohsaka H, Nanki T, et al. HLA-dependent peripheral T cell receptor (TCR) repertoire formation and its modification by rheumatoid arthritis (RA). Clin Exp Immunol 1997;110:428–33. Jawaheer D, Thomson W, MacGregor AJ, et al. “Homozygosity” for the HLA-DR shared epitope contributes the highest risk for rheumatoid arthritis concordance in identical twins. Arthritis Rheum 1994;37:681–6. Balandraud N, Picard C, Reviron D, et al. HLA-DRB1 genotypes and the risk of developing anti citrullinated protein antibody (ACPA) positive rheumatoid arthritis. PLoS One 2013;8:e64108. Svendsen AJ, Hjelmborg JV, Wiik A, et al. Autoantibodies in twins discordant for rheumatoid arthritis. Ann Rheum Dis 2011;70:708–9. Svendsen AJ, Kyvik KO, Houen G, et al. Newborn infant characteristics and risk of future rheumatoid arthritis: a twin-control study. Rheumatol Int 2014;34:523–8. Vitart V, Rudan I, Hayward C, et al. SLC2A9 is a newly identified urate transporter influencing serum urate concentration, urate excretion and gout. Nat Genet 2008;40:437–42. Scofield RH, Kurien BT, Reichlin M. Immunologically restricted and inhibitory anti-Ro/SSA in monozygotic twins. Lupus 1997;6:395–8. Bolstad AI, Haga HJ, Wassmuth R, et al. Monozygotic twins with primary Sjogren’s syndrome. J Rheumatol 2000;27:2264–6. Lee WS, Yoo WH. Primary Sjogren’s syndrome in monozygotic twins. Int J Rheum Dis 2014;17:578–9. Jawaheer D, MacGregor AJ, Gregersen PK, et al. Unexpected HLA haplotype sharing in dizygotic twin pairs discordant for rheumatoid arthritis. J Med Genet 1996;33:1015–8. Kasl SV, Cobb S. The intrafamilial transmission of rheumatoid arthritis. V. Differences between rheumatoid arthritics and controls on selected personality variables. J Chronic Dis 1969;22:239–58. Sprangers MA, Bartels M, Veenhoven R, et al. Which patient will feel down, which will be happy? The need to study the genetic disposition of emotional states. Qual Life Res 2010;19:1429–37. Duarte AP, Marques CD, Bortoluzzo AB, et al. [Epidemiologic profile of juvenileonset compared to adult-onset spondyloarthritis in a large Brazilian cohort]. Rev Bras Reumatol 2014;54:424–30. Aletaha D, Neogi T, Silman AJ, et al. 2010 rheumatoid arthritis classification criteria: an American College of Rheumatology/European League Against Rheumatism collaborative initiative. Arthritis Rheum 2010;62:2569–81. Cutler TL, Murphy K, Hopper JL, et al. Why accurate knowledge of zygosity is important to twins. Twin Res Hum Genet 2015;18:298–305. Trejo V, Derom C, Vlietinck R, et al. X chromosome inactivation patterns correlate with fetal-placental anatomy in monozygotic twin pairs: implications for immune relatedness and concordance for autoimmunity. Mol Med 1994;1:62–70. Lane M, Robker RL, Robertson SA. Parenting from before conception. Science 2014;345:756–60. Arck PC, Hecher K. Fetomaternal immune cross-talk and its consequences for maternal and offspring’s health. Nat Med 2013;19:548–56. Curnow KJ, Wilkins-Haug L, Ryan A, et al. Detection of triploid, molar, and vanishing twin pregnancies by a single-nucleotide polymorphism-based non invasive prenatal test. Am J Obstet Gynecol 2014. de Bellefon LM, Heiman P, Kanaan SB, et al. Cells from a vanished twin as a source of microchimerism 40 years later. Chimerism 2010;1:56–60.
Please cite this article in press as: Lambert NC. How twin studies help to understand inflammatory joint disease. Joint Bone Spine (2016), http://dx.doi.org/10.1016/j.jbspin.2016.02.008