Dr. Wordsworth has received consultant fees, speaking fees, and/or honoraria (less than $10,000 each) from Abbott and Pfizer.
Is there a higher genetic load of susceptibility loci in familial ankylosing spondylitis?
Article first published online: 23 APR 2012
Copyright © 2012 by the American College of Rheumatology
Arthritis Care & Research
Volume 64, Issue 5, pages 780–784, May 2012
How to Cite
Joshi, R., Reveille, J. D., Brown, M. A., Weisman, M. H., Ward, M. M., Gensler, L. S., Wordsworth, B. P., Evans, D. M. and Assassi, S. (2012), Is there a higher genetic load of susceptibility loci in familial ankylosing spondylitis?. Arthritis Care Res, 64: 780–784. doi: 10.1002/acr.21601
- Issue published online: 2 APR 2012
- Article first published online: 23 APR 2012
- Accepted manuscript online: 9 JAN 2012 10:59AM EST
- Manuscript Accepted: 4 JAN 2012
- Manuscript Received: 2 AUG 2011
- National Institute of Arthritis and Musculoskeletal and Skin Diseases. Grant Numbers: 1P01-052915-01, AR01-AR-46208, AR-48465, NIH-KL2RR024149-04
- National Center for Research Resources Clinical and Translational Sciences Awards to Cedars-Sinai Medical Center. Grant Numbers: 1U54-RR-23417-01, M01-RR-000425
- Health and Medical Research Council (Australia) Principal Research Fellowship
- Intramural Research Program of the National Institute of Arthritis and Musculoskeletal and Skin Diseases
- Medical Research Council New Investigator Award. Grant Number: G0800582
Several genetic risk variants for ankylosing spondylitis (AS) have been identified in genome-wide association studies. Our objective was to examine whether familial AS cases have a higher genetic load of these susceptibility variants.
Overall, 502 AS patients were examined, consisting of 312 patients who had first-degree relatives (FDRs) with AS (familial) and 190 patients who had no FDRs with AS or spondylarthritis (sporadic). All patients and affected FDRs fulfilled the modified New York criteria for AS. The patients were recruited from 2 US cohorts (the North American Spondylitis Consortium and the Prospective Study of Outcomes in Ankylosing Spondylitis) and from the UK-Oxford cohort. The frequencies of AS susceptibility loci in IL-23R, IL1R2, ANTXR2, ERAP-1, 2 intergenic regions on chromosomes 2p15 and 21q22, and HLA–B27 status as determined by the tag single-nucleotide polymorphism (SNP) rs4349859 were compared between familial and sporadic cases of AS. Association between SNPs and multiplex status was assessed by logistic regression controlling for sibship size.
HLA–B27 was significantly more prevalent in familial than sporadic cases of AS (odds ratio 4.44 [95% confidence interval 2.06, 9.55], P = 0.0001). Furthermore, the AS risk allele at chromosome 21q22 intergenic region showed a trend toward higher frequency in the multiplex cases (P = 0.08). The frequency of the other AS risk variants did not differ significantly between familial and sporadic cases, either individually or combined.
HLA–B27 is more prevalent in familial than sporadic cases of AS, demonstrating higher familial aggregation of AS in patients with HLA–B27 positivity. The frequency of the recently described non–major histocompatibility complex susceptibility loci is not markedly different between the sporadic and familial cases of AS.
Ankylosing spondylitis (AS) is a chronic inflammatory disorder primarily affecting the spine, peripheral joints, and entheses. Extraarticular features can affect the eye, heart, and gastrointestinal tract. AS can also occur in the setting of other spondylarthritides (SpA), including psoriatic arthritis, reactive arthritis, and enteropathic arthritis. Genetic factors provide over 90% of the overall susceptibility to AS. AS occurs in up to 75% of monozygotic twins compared to 27% in HLA–B27–positive dizygotic twins, indicating that there is a substantial non–major histocompatibility complex (MHC) genetic component underlying the disease (1). Up to half of the genetic contribution has been attributed to HLA–B27 and possibly other MHC genes. Several non-MHC genes, specifically in ERAP-1, IL-23R, IL1R2, and ANTXR2, as well as 2 intergenic regions at the chromosomes 2 and 21, have been reported as AS susceptibility loci in genome-wide association studies (2, 3).
There are several reports of familial aggregation in AS and SpA. The risk of developing AS is as high as 21% in HLA–B27–positive first-degree relatives (FDRs) of AS cases, while this risk decreases to less than 1% in HLA–B27–negative FDRs (4). Furthermore, only 6% of HLA–B27–positive subjects in the general population develop AS (5), underscoring the importance of other genetic loci beyond HLA–B27 in contributing to the risk of AS. The frequencies of the above mentioned novel AS susceptibility loci in familial cases in comparison to sporadic cases of AS have not been previously reported.
In the current study, we hypothesized that genetic susceptibility variants will be enriched in familial as compared to sporadic cases of AS. Therefore, we examined the genotype frequency of AS susceptibility loci in IL-23R, IL1R2, ANTXR2, ERAP, and 2 intergenic regions at the chromosomes 2 and 21, in addition to a recently described HLA–B27 tag single-nucleotide polymorphism (SNP) in familial and sporadic cases of AS.
Significance & Innovations
This is the first genetic study to compare the frequency of the recently described susceptibility variants between the familial and sporadic ankylosing spondylitis (AS) cases.
There is a higher frequency of HLA–B27 positivity in familial than sporadic cases of AS.
The frequency of the other risk variants implicated in AS was not significantly increased in familial cases, either individually or combined.
Materials and methods
Familial or multiplex cases of AS were defined as having >1 family member (parent, sister, or brother only, offspring were excluded in this study) with a confirmed diagnosis of AS. Singleton or sporadic cases of AS were defined as having only 1 family member with a confirmed diagnosis of AS and no other reported FDRs with AS or SpA. All index cases and FDRs with AS fulfilled the modified New York criteria for AS (6). The investigated cohorts included the North American Spondylitis Consortium (NASC), the Prospective Study of Outcomes in Ankylosing Spondylitis (PSOAS), and the UK-Oxford cohorts.
The NASC is a collaboration of 8 sites in US cities (Houston, Texas; Los Angeles, California; Portland, Oregon; Cleveland, Ohio; Minneapolis, Minnesota; Philadelphia, Pennsylvania; Salt Lake City, Utah; and Tampa, Florida), the Spondylitis Association of America, and 2 sites in Canada (Toronto, Ontario and Edmonton, Alberta) examining multiplex families with AS (7). The patients in the NASC included families with 2 or more FDRs meeting the modified 1984 New York criteria for AS and for whom radiographs were available to confirm the diagnosis. The diagnoses were reevaluated by rheumatologists with experience in assessing patients with AS and the radiographs were assessed by the study investigators (JDR, MHW).
The PSOAS is an ongoing longitudinal study of AS patients from 4 sites in the US (Cedars-Sinai Medical Center, Los Angeles, California; the University of Texas at Houston; the University of California at San Francisco; and the National Institutes of Health, Bethesda, Maryland) (8). Multiplex and singleton cases of AS were enrolled in this study. The presence of sacroiliitis on pelvic radiographs was confirmed by a musculoskeletal radiologist.
The UK-Oxford study is an ongoing cross-sectional genetic study of multicase and sporadic AS cases where, in all cases, the diagnosis of AS had been confirmed by a qualified rheumatologist. To further confirm the diagnosis, patients in all cases were interviewed by telephone by the study investigators (MAB, BPW). In patients with an atypical medical history or for whom radiographs had not been performed previously, pelvic and lumbosacral spine radiographs were obtained and attending physicians were contacted to confirm the diagnosis (9). In this study, all affected siblings per self-report were assessed according to the above mentioned methodology.
All of the investigated patients in the current study were of self-reported white European descent. Only multiplex AS cases with confirmed affected FDRs (parents and siblings only) and singleton AS cases were investigated. Only 1 patient with AS per family was included in the analysis.
The following non-MHC susceptibility variants for AS were investigated in the current study: IL-23R, rs11209026; IL1R2, rs2310173; ANTXR2, rs4333130; ERAP-1, rs27434; and 2 gene desert loci, rs10865331 and rs2242944 (2, 3). In addition, HLA–B27 was studied using the findings of a recently described HLA–B27 tag SNP, rs4349859 (10). The genotype information was obtained from a previously described genome-wide association study conducted in patients enrolled in the NASC, PSOAS, and UK-Oxford cohorts (2).
Additionally, the clinical features of singleton AS cases in the PSOAS cohort were compared to the multiplex cases in the UK-Oxford study cohort. The information on the time of disease onset was not available in the NASC cohort; therefore, patients enrolled in the NASC cohort were not included in the comparison of clinical features (Table 1).
|Sporadic||Familial||Effect size†||95% CI||P|
|Sex, % male||76.4||60||0.46||0.71, 0.3||0.0004|
|Age at disease onset, mean years||23.7||22.9||−0.84||−2.54, 0.86||0.331|
|Disease duration at enrollment, mean years||24.2||24.9||0.68||−2.03, 3.39||0.62|
|Uveitis, %‡||45.2||39.4||0.75||0.49, 1.13||0.16|
The association between SNPs and multiplex status was assessed by logistic regression assuming an additive effect of alleles on log odds of being from a multiplex family. Individuals from multiplex families were coded as cases, and individuals from simplex families were coded as controls. The best guess genotypes at rs4349859 were imputed using MaCH software (11) as described by Evans et al (10). Quality control metrics indicated that this SNP could be imputed with a high degree of certainty (Rsquare >0.98). Imputed genotypes at rs4349859 were only available for 292 multiplex and 177 sporadic cases of AS. In order to minimize the effect of possibly misclassifying multiplex cases from smaller families as sporadic cases, we examined the effect of adding sibship size as a covariate in the logistic regression model.
In order to ensure that our results were robust to the effect of population stratification, we also examined the effect of including EigenStrat-derived principal components from a previous genome-wide association study of AS in our analysis (2) and restricting our analysis to US cases only (128 multiplex cases and 176 sporadic cases).
The relationship between the number of affected siblings per family and the number of HLA–B27 alleles in the index case was examined in the UK-Oxford cohort by linear regression.
In the current study, 312 unrelated patients with multiplex AS and 190 patients with singleton AS were investigated. In detail, 160 multiplex AS cases were recruited from the UK-Oxford cohort, for whom the required genotype data were available. These patients had a mean ± SD age of 47.9 ± 10.2 years at enrollment and 60% were men. The mean ± SD disease duration at enrollment was 24.9 ± 11 years. We also recruited 152 multiplex AS cases from the NASC study. In this group, the mean ± SD age at enrollment was 44.4 ± 10.4 years and 58% of the patients were men. Furthermore, 190 singleton cases were recruited from the PSOAS. This patient group had a mean ± SD age of 47.1 ± 16.1 years, 76% were men, and they had a mean ± SD disease duration of 22.8 ± 15.7 years at enrollment. Table 1 shows the comparison of clinical features between the singleton and multiplex AS cases. The familial cases had a lower percentage of male patients (P = 0.0004). The other clinical features did not differ significantly between the singleton and multiplex cases.
Table 2 shows the frequency of susceptibility loci in the multiplex and singleton cases. HLA–B27 was significantly more prevalent in multiplex than in singleton AS cases (odds ratio [OR] 5.41 [95% confidence interval (95% CI) 2.92, 10.03], P = 8.4 × 10−8). The frequency of the other AS susceptibility loci in the IL-23R, IL1R2, ANTXR2, and ERAP-1 genes and the chromosome 2p15 and 21q22 intergenic regions did not differ significantly between the singleton and multiplex cases. Furthermore, sex did not modify the significant relationship between HLA–B27 and occurrence of familial AS (P for the interaction term between sex and HLA–B27 = 0.763).
|SNP||Candidate gene||Risk allele||OR (95% CI)||P|
|rs11209026||IL-23R||G||1.23 (0.64, 2.36)||0.54|
|rs10865331||2p15†||A||0.87 (0.67, 1.12)||0.26|
|rs2310173||IL1R2||T||1.10 (0.86, 1.42)||0.44|
|rs4333130||ANTXR2||T||1.15 (0.87, 1.52)||0.32|
|rs27434||ERAP-1||A||0.85 (0.63, 1.13)||0.26|
|rs2242944||21q22†||G||1.20 (0.91, 1.59)||0.19|
|rs4349859||HLA–B27 tag SNP||A||5.41 (2.92, 10.03)||8.4 × 10−8‡|
Similar results were observed after adjustment for sibship size (Table 3). HLA–B27 was more prevalent among multiplex cases of AS (P = 0.0001). Furthermore, the risk allele in the intergenic region on 21q22 showed a trend toward higher frequency in the multiplex cases (OR 1.37 [95% CI 0.96, 1.95], P = 0.08). The other susceptibility loci did not show a significant association with familial occurrence of AS. As shown in Table 3, these results did not change when the analysis was restricted to patients recruited from the 2 US cohorts, and covariates were added to reflect population substructure.
|SNP||Candidate gene||Overall cohort||US sample|
|OR (95% CI)||P||OR (95% CI)||P|
|rs11209026||IL-23R||1.44 (0.63, 3.27)||0.39||2.33 (0.76, 7.14)||0.14|
|rs10865331||2p15†||0.85 (0.62, 1.16)||0.31||0.99 (0.69, 1.42)||0.95|
|rs2310173||IL1R2||1.26 (0.9, 1.73)||0.16||1.31 (0.91, 1.89)||0.15|
|rs4333130||ANTXR2||1.24 (0.81, 1.64)||0.42||1.15 (0.76, 1.72)||0.51|
|rs27434||ERAP-1||0.81 (0.55, 1.17)||0.26||0.89 (0.57, 1.38)||0.6|
|rs2242944||21q22†||1.37 (0.96, 1.95)||0.08||1.45 (0.95, 2.17)||0.08|
|rs4349859||HLA–B27 tag SNP||4.44 (2.06, 9.55)||0.0001||4.08 (1.7, 9.83)||0.002|
The numbers of risk alleles an individual carries at non-HLA loci were added together and then divided by the total number of successfully genotyped SNPs at the susceptibility loci to obtain an allelic risk score. This score was regressed against the log odds of being from a multiplex family in order to assess the cumulative effect of the AS non-MHC risk alleles on the multiplex status. This allelic score was not associated with a risk of being from a multiplex family (P = 0.64).
There was no significant relationship between the number of affected siblings and the number of HLA–B27 alleles (OR 1.2 [95% CI 0.89, 1.62], P = 0.226).
The current report is the first genetic study to investigate the frequency of recently described susceptibility variants in familial and sporadic AS patients. We report that the relative prevalence of HLA–B27 was higher in familial AS, while the frequency of other risk variants implicated in AS was not significantly increased. This implies a higher genetic load of HLA–B27 in familial AS patients.
Significant advances have been made in understanding the genetic basis of AS. The most prominent genetic susceptibility locus remains HLA–B27. Definite non-MHC associations with AS have been demonstrated with polymorphisms in the ERAP-1, IL-23R, IL1R2, and ANTXR2 genes, in addition to 2 intergenic regions at chromosomes 2p15 and 21q22 (2, 3). ERAP-1 is the second strongest genetic association after HLA–B27 and the strongest non-MHC susceptibility locus. IL-23R has also been implicated in AS cases of populations of white European ancestry, although not in populations from Eastern Asia. Other genes associated with AS include IL-1R and ANTXR2. Chromosomes 2p15 and 21q22 are 2 intergenic regions or gene deserts associated with AS. Although these non-MHC susceptibility loci were confirmed in large independent cohorts (2, 3), the effect size of these risk variants is markedly lower than HLA–B27.
The current study compared the genetic data between confirmed familial and sporadic cases of AS. In addition to the abovementioned non-MHC susceptibility loci, the current study investigated an MHC susceptibility locus. Evans et al identified a single SNP rs4349859 near the gene MICA that tagged HLA–B27 with a sensitivity of 98% and a specificity of 99% in AS cases of those of white European descent. Therefore, it can be used as an alternative to HLA–B27 genotyping in this ethnicity (10). We report that the HLA–B27 tag SNP was more prevalent in familial AS, while we did not observe a significantly higher prevalence of non-MHC susceptibility loci among familial cases. This implies higher genetic load of HLA–B27 in familial AS patients. While FDRs of AS patients who have the HLA–B27 allele are at higher risk of developing AS than their HLA-negative counterparts (4), there are no previous reports of higher frequency of HLA–B27 positivity in familial versus sporadic cases of AS (4). The current study is the first that demonstrates higher familial aggregation of AS in patients with HLA–B27 positivity. This finding can be explained by the large contribution of HLA–B27 to the susceptibility of AS.
A study of 55 familial and 110 sex- and age-matched sporadic cases of AS found no difference in disease onset, age at diagnosis, or prevalence of peripheral arthritis or uveitis (12). Another study of 160 familial and 160 age- and sex-matched sporadic cases of AS indicated that familial disease was significantly milder in terms of spinal mobility score, pain score, psychological functioning, and social functioning (13). Differences in the frequency of severity genes were postulated as an explanation for this finding; however, a selection bias toward milder disease among familial cases is another possible explanation. In the current study, the familial AS cohorts included a higher percentage of female patients. This might be secondary to lower threshold of assessing and diagnosing a female patient with AS if another family member has AS (verification bias). Similar to the above mentioned study (12), the other clinical features did not differ between the comparison groups.
The strengths of the current study are the relatively large sample size, the strict confirmation of diagnosis according to the modified New York criteria (6) in index cases as well as in affected FDRs, and the examination of genotype frequencies after correction for sibship size. Although our study is the largest to date comparing familial AS with sporadic AS, it is still underpowered to detect more subtle differences in frequency of non-MHC susceptibility genes between the 2 investigated groups. Furthermore, we cannot exclude that potential multiplex cases were misclassified as singleton families because of a low number of siblings. We tried to account for this possible misclassification by adjusting for sibship size. There are several other potential sources of heritability among familial cases that were not examined in the current study. There might be further relevant SNPs (either common or rare variants) or insertion/deletion polymorphisms not tagged by the investigated SNPs. Moreover, gene–environment interaction might affect familial occurrence of AS.
HLA tag SNP rs4349859, which tracks closely with HLA–B27, is more prevalent in familial than sporadic cases of AS. This finding demonstrates higher familial aggregation of AS in patients with HLA–B27 positivity. The frequency of the recently described non-MHC susceptibility loci is not markedly different between the sporadic and familial cases of AS, although we cannot exclude more subtle differences in frequency of these susceptibility variants between the 2 investigated groups.
All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Assassi had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study conception and design. Joshi, Reveille, Brown, Weisman, Gensler, Assassi.
Acquisition of data. Joshi, Reveille, Brown, Weisman, Ward, Gensler, Wordsworth, Assassi.
Analysis and interpretation of data. Reveille, Brown, Weisman, Gensler, Evans, Assassi.
- 12Dutch patients with familial and sporadic ankylosing spondylitis do not differ in disease phenotype. J Rheumatol 2002; 29: 2583–4., , , .