Paediatric rheumatology

Are MEFV susceptibility factors in enthesitis-related arthritis patients in the eastern Mediterranean? B. Gülhan1, A. Akkuş1, L. Özçakar2, N. Beşbaş1, S. Özen1

1Department of Paediatric Nephrology ABSTRACT ILAR, with an aim to also cover the and Rheumatology, and Objective. Enthesitis-related arthritis previously suggested terms of juvenile 2 Department of Physical and Rehabilitation, (ERA), is a complex genetic disease. spondyloarthropathy and seronegative Hacettepe University Faculty of Medicine, Although HLA-B27 is well established, enthesitis arthritis syndrome (1, 2). The Ankara, Turkey. it does not explain all the genetic load etiopathogenesis of ERA is not certain Bora Gülhan, MD in ERA. Familial Mediterranean fe- however, it is a complex genetic disease Abdulkadir Akkuş, MD Levent Özçakar, MD ver (FMF), caused by mutations in the like the other subtypes of JIA. Genetic Nesrin Beşbaş, MD MEFV , is a frequent autoinflam- predisposition has a significant impact Seza Özen, MD matory disorder in the eastern Medi- on ERA, where common single nucleo- Please address correspondence to: terranenan. tide polymorphisms in multiple Prof. Seza Özen, Methods. We investigated the clini- are contributory as well, with real but Department of Paediatrics, cal and imaging features of 53 ERA variable environmental components. Faculty of Medicine, patients, as well as the frequency of HLA-B27 accounts for the major Hacettepe University, MEFV gene mutations in those who genetic load and is positive in approxi- 06100 Sıhhiye, Ankara, Turkey. were HLA-B27 negative. E-mail: [email protected] mately 90% of the patients with juve- Results. The mean age of the patients nile ankylosing spondylitis (AS) (3). Received on June 17, 2013; accepted in revised form on November 11, 2013. was 13.3±2.2 years and 49 were boys. Familial Mediterranean fever (FMF) is Peripheral arthritis was present in an autosomal recessive autoinflamma- Clin Exp Rheumatol 2014; 32 (Suppl. 84): S160-S164. all and sacroiletis in 26 patients. Ul- tory disorder characterised by recurrent trasonography showed enthesitis in 6 attacks of fever and polyserositis (4). © Copyright Clinical and Experimental Rheumatology 2014. patients of the tendons, whereas these Mutations in the gene for FMF (MEFV) were assessed to be normal by physical are very common and the carrier rate Key words: enthesitis-related arthritis, examination. Forty patients (75.5%) is very high in the eastern Mediterra- familial Mediterranean fever were positive for HLA-B27. MEFV nean (5). Several case reports and co- analysis was performed for patients horts suggest an association between who were HLA-B27 negative. One pa- mutations in MEFV gene and certain tient refused MEFV analysis. 9 patients inflammatory disorders, such as -rheu carried MEFV mutations: 2 patients matoid arthritis (6), systemic onset JIA were homozygous for M694V (both (7), seronegative spondyloarthropathy patients were subsequently started col- (8), Henoch-Schönlein purpura (9), chicine along with ERA treatment), 5 polyarteritis nodosa (10, 11), Behçet’s patients were heterozygous for M694V disease (BD) (12, 13), and Crohn dis- , 1 patient was heterozygous ease (14). for E148Q, and 1 patient was heterozy- The aim of this study was to analyse the gous for K695R mutation. None of the characteristics of Turkish children with patients had features suggesting FMF ERA in a single referral centre, and to at diagnosis of ERA; 1 patient subse- determine the frequency of MEFV mu- quently developed typical FMF attacks. tations among those who were HLA- Conclusion. Our findings suggest that B27 negative. MEFV mutations may represent a sus- ceptibility factor for ERA in the popu- Patients and methods lations of the eastern Mediterranean. Patients Patients followed in Hacettepe Univer- Introduction sity Paediatric Nephrology and Rheu- Enthesitis-related arthritis (ERA) is a matology Unit were included in this subtype of juvenile idiopathic arthritis study; the ERA population of our cen- (JIA). The term ERA and its pertinent tre consists of children from all around Competing interests: none declared. classification have been introduced by the country, the minority being from

S-160 MEFV mutations in enthesitis-related arthritis patients / B. Gülhan et al. PAEDIATRIC RHEUMATOLOGY the local area. The study group of this I692del, M694V, M694I, K695R, methotrexate was also combined for cross-sectional study consisted of 53 V726A, A744S, R761H in exon 10. the peripheral arthritis in 24 patients Caucasian children (92.5% male; M:F (45.3%). Seventeen patients received ratio was 12.2) who met ERA criteria Radiological findings intraarticular steroids. In 11 patients, introduced by ILAR (1). These criteria Sacroileitis was assessed by magnetic short-term steroids were used in ac- are the presence of arthritis and en- resonance imaging (MRI) and/or x-ray. tive arthritis periods. At the end of the thesitis in the same patient or any two Ultrasonographic imaging was avail- follow-up, all patients improved with a of them that accompany arthritis or able in 34 patients and performed mean BASDAI score of 2.0 (IQR: 0.8- enthesitis; sacroiliac joint tenderness, bilaterally for knee and ankle joints. 4.0). Five patients who continued to HLA-B27 positivity, history of arthritis Quadriceps, patellar and Achilles ten- have active arthritis, with high BAS- after the age of 8, acute anterior uveitis, dons and plantar fasciae were scanned DAI scores and acute phase reactants presence of HLA-B27-related disease longitudinally and axially. The same were switched to etanercept. All of history in first degree relatives. All physician (with musculoskeletal ul- those patients were in remission at 6 patients also met European Spondy- trasound experience of more than 10 months of their follow-up. None of the loarthropathy Study Group (ESSG) cri- years) who was blinded to the clinical patients showed side effects related to teria (15). The study was approved by data performed all the evaluations us- the drugs. the ethics committee at the Hacettepe ing a 8–16 MHz linear probe (Diasus University. Dynamic Imaging). Laboratory parameters Mean ESR at diagnosis was 47 (IQR: Clinical assessment Statistical analysis 18-72) mm/hour. Thirty-three patients During routine outpatient visits, com- The results were analysed by using (62.3%) had elevated acute phase re- plete physical and rheumatological SPSS for windows 13.0 and descrip- actants at diagnosis. Forty patients examinations were performed. Ad- tive statistics are presented as per- (75.5%) were positive for HLA-B27. ditionally, spinal mobility was meas- centages, means and SDs if they were MEFV mutation analysis was carried ured with modified Schober test. The equally distributed, and as median out in 12 children with ERA who were Bath Ankylosing Spondylitis Disease (IQR, interquartile ratio) if they were HLA-B27 negative: the study revealed Activity Index (BASDAI) was used unequally distributed. mutations in 9 patients (11 alleles had to assess disease activity (16). Evalu- MEFV mutations among the 24): 2 pa- ations, treatment plans and follow-up Results tients were homozygous for M694V visits of the patients were performed Patients (both patients were put on colchicine by the same paediatric rheumatologist. In the study cohort, 49 patients (92.5%) treatment along with ERA treatment), 5 Tel-Hashomer Criteria and recently were male and 4 patients (7.5%) were patients were heterozygous for M694V proposed set of FMF criteria for chil- female. Clinical features of the sub- mutation, 1 patient was heterozygous dren were used to evaluate all patients jects are summarized in Table I. The for E148Q, and 1 patient was heterozy- for FMF (17, 18). All patients were mean age of the patients at time of the gous for K695R mutation. None of the treated with non-steroidal anti-inflam- study was 13.3±2.2. The mean age at patients had features suggesting FMF matory drugs (NSAIDs). For resistant diagnosis was 12.7±2.3 years, ranging at diagnosis of ERA; 1 patient started patients, sulfasalazine or methotrexate from 8.0 years to 16.9 years. All 53 pa- to develop typical FMF attacks during was added as disease-modifying drugs tients had peripheral disease; the mean follow-up (after the mutation analysis (DMARDs). Etanercept was used for number of joints involved was 2.1±1.1 was performed). patients unresponsive to DMARDs. (1-4). Nineteen patients (35.8%) had Mean age of the patients with a MEFV 1 joint, 19 patients had ≥3 joints in- mutation was 13.4±1.8 (11.1-15.8) and Laboratory testing volved. The most commonly affected was comparable to the cohort of our High acute phase reactants was de- joint was the knee (56%), followed patients. The pattern of joint and ten- fined as ESR >20 mm/hour and/or C- by the ankle (32%) and the hip (12%). don involvement among the patients reactive >0.5 mg/dl. HLA-B27 was Mean BASDAI score at diagnosis was who had a MEFV mutation, was again checked in all patients and MEFV mu- 5.7 (IQR: 4.4–6.3). Modified Schober similar to the other patients. All of tation analysis was studied in patients test was used for evaluation of thoraco- these nine patients had elevated acute who were HLA-B27 negative. One pa- lomber spinal mobility and mean value phase reactants at presentation. tient who was HLA-B27 negative re- was 5.3±1.1 (3-8). fused MEFV analysis. MEFV analysis Sixteen patients (30.2%) had HLA- Radiological findings was performed in 12 patients by using B27-related disease history in first or Sacroileitis was diagnosed in 26 pa- restriction fragment length polymor- second degree relatives. 40 patients tients (49.1%). Twenty-three patients phism (RFLP). The following MEFV (75.5%) were HLA B27 positive. were evaluated only with x-ray, 18 pa- mutations were investigated: E148Q Only 8 patients (15%) responded to the tients only with MRI and 12 both with in exon 2, P369S in exon 3, F479L in initial NSAID treatment. Sulfasalazine MRI and x-ray. X-ray showed sacro- exon 5 and M680I (G/C), M680I (G/A), was started in 40 patients (75.5%) and ileitis in 10/35 patients (28.6%). A total

S-161 PAEDIATRIC RHEUMATOLOGY MEFV mutations in enthesitis-related arthritis patients / B. Gülhan et al.

Table I. Clinical features of 53 children by Ansell et al. (27.3%), Häfner et al. concern on the duration of anti-TNF with enthesitis-related arthritis. (14%), and O’Shea (27.7%) (19-21). treatment in these patients who can be Features n (%) Axial involvement is less compared to foreseen to have a long-lasting disease the series reported by by O’Shea et al.; ahead of them. Age at symptom onset (years) 11.1 ± 1.8 49.1% versus 74% (21). The frequency Another new aspect of our study group Age at diagnosis (years) 12.7 ± 12.3 of HLA-B27 is also lower than expect- was the identification of MEFV muta- Disease duration (years) 2.16 ± 1.43 ed (3). Although this is a small series tions in a substantial number of pa- Abnormal Schober test [n (%)] 34 (64.2%) these features are compatible with pre- tients who were HLA-B27 negative. Enthesitis [n (%)] 26 (49.1%) vious unpublished observations of the Genetic predisposition has a significant Sacroileitis [n (%)] 26 (49.1%) centre. Furthermore, being a referral impact on ERA, where common single Uveitis [n (%)] 5 (9.4%) centre the patient population mirrors nucleotide polymorphisms in multiple Morning stiffness [n (%)] 25 (47.2%) the characteristics of the Anatolian genes contribute to risk, with real but population. variable environmental components. of 30 patients were evaluated with MRI In our patient group the response to The association with an HLA polymor- and 17 (56.7%) out of them were di- NSAID and sulfasalazine was very phism, B27, with ankylosing spondy- agnosed to have sacroileitis. In patients good. This may be attributed to the litis and ERA is probably one of the who were evaluated with both MRI and dominance of peripheral arthritis pat- most striking associations in rheuma- x-ray (n=12), MRI showed sacroileitis tern in our patient population. Only tology. However, HLA-B27 does not in seven patients although they had about 10% of the patients required account for all the genetic susceptibil- normal x-ray (p<0.05). anti-TNF treatment. These five patients ity. In addition to HLA-B27, non-HLA On ultrasound (n=34), 15 and 21 pa- had an excellent response to etanercept genes IL23R and ERAP1 (also known tients had effusions in the right and left monotherapy. Actually, around 50% as ARTS1) have recently been identi- knee, respectively. All patients with ef- of the patients in our series developed fied as susceptibility factors (23). This fusion on right side also had effusion sacroileitis and thus met the Modi- latter gene product also been shown on the left side. None of the patients fied New York Criteria for Ankylosing to influence shedding of the IL-1 and had effusion on ankles. Pathologic fea- Spondylitis (AS), fulfilling the term IL-6 receptors (24, 25). Being a domi- tures were detected in the periarticular juvenile onset AS (JoAS) (22). In the nant inflammatory molecule, IL-1 may area in 9 patients: six patients had ret- series reported by O’Shea et al. the indeed have an important affect on the rocalcaneal bursitis (2 bilateral, 2 on mean age of the patients with JoAS at pathogenesis of ERA and AS. A num- the right and 2 on the left side). diagnosis was relatively higher (17.2 ber of studies have highlighted the as- On USG evaluation (n=34), 19 patients years), which is higher as compared to sociation of the disease with IL-1 poly- had enthesitis at least in one tendon. that of our series, and could be a cause morphisms (26, 27). Some of these On the other hand, 15 did not have of higher axial involvement (21). We poly-morphisms had a functional effect any enthesitis findings in any joints. cannot predict the final percentage of on the IL1 levels (28). A meta-analysis Enthesitis was detected in 1 tendon our patients who will go on to develop of 2675 AS cases and 2592 healthy in 5 patients, 2 tendons in 10 patients, ankylosing spondylitis. This raises the controls for the IL1A, ILβ, IL1F10 3 tendons in 3 patients, 4 tendons in 1 patients (Fig. 1). In 25 patients the physical examination agreed with the musculoskeletal ultrasound. However, in six patients ultrasound examination detected entesopathy whereas physical examination was judged to be normal. Only the left plantar fascia was found to be thicker in HLA-B27-positive pa- tients when compared with those of HLA-B27-negative patients. The pres- ence of either sacroileitis or uveitis did not have any impact on tendon thick- ness measurements.

Discussion The presented Turkish cohort, pre- sents certain differences from other series. The number of joints involved is fewer and the frequency of uvei- Fig. 1. Distribution of enthesitis in various tendons. tis is less compared to series reported

S-162 MEFV mutations in enthesitis-related arthritis patients / B. Gülhan et al. PAEDIATRIC RHEUMATOLOGY and IL1RN polymorphisms revealed a the initial response to the environmen- in other populations. However, in a strong association of three single nucle- tal trigger (34, 37). Thus it has been British study the E148Q mutation / otide polymorphisms in the IL1A gene suggested that for the aforementioned polymorphism of the MEFV gene was (rs2856836, rs17561, rs18943999) clinical associations, mutations/poly- surprisingly found in a group of Brit- (29). Since mutations in the MEFV morphisms in the MEFV gene may be ish and Indian chronic arthritis patients gene are also associated with high IL1 serving as susceptibility factors for the who developed secondary levels may serve as a possible expla- disease or a more severe course (34). (40). MEFV mutations may indeed be nation of the association that we have The role of MEFV mutation in one of the susceptible genetic factors in shown in our population. In the adult diseases is not completely understood. ERA/AS in this part of the world. literature, there are studies that investi- Current medical evidence suggest that gate the prevalence of MEFV gene var- MEFV mutations are gain-of-function References iants in AS patients with no history of mutations which lead to increased re- 1. PETTY RE, SOUTHWOOD TR, BAUM J et al.: FMF. Cosan et al. genotyped 193 AS sponsiveness to bacterial products (38). Revision of proposed classification crite- patients and 103 healthy controls for Mutations of the MEFV gene have been ria for juvenile idiopathic arthritis: Durban, 1997. J Rheumatol 1998; 25: 1991-4. the most common mutations (M694V, previously suggested to be associated 2. PETTY RE, SOUTHWOOD TR, MANNERS P et M680I, V726A, E148Q) by RFLP and with certain rheumatic diseases. Re- al.: International League of Associations for they also showed that exon 10 variants cent paper published by Kirino et al., Rheumatology. International League of As- of the MEFV gene were more frequent investigated the role of the MEFV gene sociations for Rheumatology classification of juvenile idiopathic arthritis: second revi- among HLA B27 (-) patients with AS mutations in the development of BD sion, Edmonton, 2001. J Rheumatol 2004; (30). They further analysed M694V in Japanese and Turkish populations 31: 390-2. and MEFV mutations in their whole (39). In Turkish population, M694V 3. CASSIDY JT, PETTY RE: Juvenile ankylosing AS cohort: MEFV mutations were mutation (both homozygous and het- spondylitis. In: CASSIDY JT, PETTY RE, LAX- ER LM, LINDSLEY (Eds.). Texbook of Pedi- more frequent among the AS patients erozygous) conferred strong risk for atric Rheumatology. Philadelphia: Elsevier compared to healthy controls however, BD. On the other hand, this association Saunders 2005: 304-323. exon 10 mutations were significantly of M694V with BD was not found in 4. SHOHAT M, HALPERN GJ: Familial Medi- more frequent among the HLA B27(-) Japanese population. In our study, we terranean fever – a review. Genet Med 2011; 13: 487-98. patients as compared to HLA 27+ ones saw this association with ERA patients. 5. YILMAZ E, OZEN S, BALCI B et al.: Mutation (30). Two other uncontrolled studies Thus it is tempting to speculate that frequency of familial Mediterranean fever conducted in central Anatolia estimat- certain MEFV mutations may be acting and evidence for a high carrier rate in the ed a high allelic frequency of M694V as susceptility factors for ERA/JAS in Turkish population. Eur J Hum Genet 200l; 9: 553-5. among AS patients at 6.3% and 12.3% the eastern Mediterranean. 6. RABINOVICH E, LIVNEH A, LANGEVITZ P et in the general population of the same On the other hand, the presence of two al.: Severe disease in patients with rheuma- region (31, 32). The previously report- patients with homozygous mutations toid arthritis carrying a mutation in the ed figure in the general population of also suggest that ERA should be con- Mediterranean fever gene. Ann Rheum Dis 2005; 64: 1009-14. the same region was about 1.1–3% (30, sidered in FMF patients with specific 7. AYAZ NA, OZEN S, BILGINER Y et al.: MEFV 33). This data suggests that MEFV mu- features. In fact spondyloarthropathies mutations in systemic onset juvenile idi- tations may be contributing to the over- have been previously recognised to be opathic arthritis. Rheumatology (Oxford) all pathogenesis of AS as well. more frequent among FMF patients with 2009; 48: 23-5. 8. LANGEVITZ P, LIVNEH A, ZEMER D, SHEM- Thus MEFV mutations may well be the existing case reports (8). It is very ER J, PRAS M: Seronegative spondyloar- one of the susceptibility factors show- interesting that two of the presented pa- thropathy in familial Mediterranean fever. ing a significant association with ERA tients lacked other clinical features of Semin Arthritis Rheum 1997; 27: 67-72. and AS in at least in certain areas of the FMF at diagnosis. They were neverthe- 9. GERSHONI-BARUCH R, BROZA Y, BRIK R: Prevalence and significance of mutations world. In fact we know that the car- less started on concomitant colchicine. in the familial Mediterranean fever gene in rier rate for the MEFV mutations are New technology improves the detection Henoch-Schönlein purpura. J Pediatr 2003; increased among patients with various of musculoskeletal features in these pa- 143: 658-61. rheumatic diseases and complaints in tients. MRI showed sacroileitis in 7 sev- 10. OZEN S, BEN-CHETRIT E, BAKKALOGLU A et al.: Polyarteritis nodosa in patients with the eastern Mediterranean including en patients that would be missed with Familial Mediterranean Fever (FMF): a con- polyarteritis nodosa, Henoch Schonlein conventional radiography. Ultrasono- comitant disease or a feature of FMF? Semin purpura and some forms of juvenile graphy was more sensitive for detecting Arthritis Rheum 2001; 30: 281-7. idiopathic arthritis (7, 9, 11, 12, 34). A enthesitis, suggesting the need to use 11. YALÇINKAYA F, OZÇAKAR ZB, KASAP- ÇOPUR O et al.: Prevalence of the MEFV simple explanation of this association these tools in our paediatric practice. gene mutations in childhood polyarteritis no- may be the non-specific effect of- in Mutations and polymorphisms may rep- dosa. J Pediatr 2007; 151: 675-8. creased inflammatory state introduced resent a further group of SNPs to assess 12. TOUITOU I, MAGNE X, MOLINARI N et al.: by the presence of MEFV mutations in the etiopathogenesis of ERA. Muta- MEFV mutations in Behçet’s disease. Hum Mutat 2000;16: 271-2. (34-37). MEFV mutations may be ef- tions and polymorphisms of the MEFV 13. LIVNEH A, AKSENTIJEVICH I, LANGEVITZ fective through the upregulation of the gene are common to eastern Mediter- P et al.: A single mutated MEFV allele in innate immune system which serves as ranean (5). They are very uncommon Israeli patients suffering from familial Medi-

S-163 PAEDIATRIC RHEUMATOLOGY MEFV mutations in enthesitis-related arthritis patients / B. Gülhan et al.

terranean fever and Behçet’s disease. Eur J autoimmunity variants. Nat Genet 2007; 39: on disease severity in ankyloing spondylitis Hum Genet 2001; 9: 191-6. 1329-37. EULAR Meeting Abstracts. 2007; 66: 400. 14. FIDDER HH, CHOWERS Y, LIDAR M et al.: 24. CUI X, ROUHANI FN, HAWARI F, LEVINE SJ: 33. YILMAZ E, OZEN S, BALCI B et al.: Mutation Crohn disease in patients with familial Medi- Shedding of the type II IL-1 decoy receptor frequency of Familial Mediterranean Fever terranean fever. Medicine (Baltimore) 2002; requires a multifunctional aminopeptidase, and evidence for a high carrier rate in the 81: 411-6. aminopeptidase regulator of TNF receptor Turkish population. Eur J Hum Genet 2001; 15. DOUGADOS M, van der LINDEN S, JUHLIN type 1 shedding. J Immunol 2003; 171: 6814- 9: 553-5. R et al.: The European Spondylarthropathy 9. 34. OZEN S: Mutations/polymorphisms in a Study Group preliminary criteria for the 25. CUI X, ROUHANI FN, HAWARI F, LEVINE SJ: monogenetic autoinflammatory disease may classification of spondylarthropathy. Arthri- An aminopeptidase, ARTS-1, is required for be susceptibility markers for certain rheu- tis Rheum 1991; 34: 1218-27. interleukin-6 receptor shedding. J Biol Chem matic diseases: lessons from the bedside for 16. GARRETT S, JENKINSON T, KENNEDY LG et 2003; 278: 28677-85. the benchside. Clin Exp Rheumatol 2009; 27 al.: A new approach to defining disease sta- 26. MAKSYMOWYCH WP, RAHMAN P, REEVE (Suppl. 53): S29-31. tus in ankylosing spondylitis: the Bath An- JP et al.: Association of the IL1 gene cluster 35. DUZOVA A, BAKKALOGLU A, BESBAS N et kylosing Spondylitis Disease Activity Index. with susceptibility to ankylosing spondylitis: al.: Role of A-SAA in monitoring subclini- J Rheumatol 1994; 21: 2286-91. an analysis of three Canadian populations. cal and in colchicine dosage in 17. LIVNEH A, LANGEVITZ P, ZEMER D et al.: Arthritis Rheum 2006; 54: 974-85. familial Mediterranean fever. Clin Exp Rheu- Criteria for the diagnosis of familial Medi- 27. SIMS AM, TIMMS AE, BRUGES-ARMAS J et matol 2003; 21: 509-14. terranean fever. Arthritis Rheum 1997; 40: al.: International Genetics of Ankylosing 36. LACHMANN HJ, SENGÜL B, YAVUZŞEN TU 1879-85. Spondylitis. Prospective meta-analysis of et al.: Clinical and subclinical inflammation 18. YALÇINKAYA F, OZEN S, OZÇAKAR ZB et interleukin 1 gene complex polymorphisms in patients with familial Mediterranean fever al.: A new set of criteria for the diagnosis of confirms associations with ankylosing spon- and in heterozygous carriers of MEFV muta- familial Mediterranean fever in childhood. dylitis. Ann Rheum Dis 2008; 67: 1305-9. tions. Rheumatology (Oxford) 2006; 45: 746- Rheumatology (Oxford) 2009; 48: 395-8. 28. HALL SK, PERREGAUX DG, GABEL CA et 50. 19. ANSELL BM: Juvenile spondylitis and related al.: Correlation of polymorphic variation in 37. BEN-CHETRIT E, TOUITOU I: The impact of disorders. In: MOLL JMH (Ed). Ankylosing the promoter region of the interleukin-1 beta MEFV gene identification on FMF: an- ap Spondylitis. Edinburgh, Churchill Lining- gene with secretion of interleukin-1 beta pro- praisal after 15 years. Clin Exp Rheumatol stone, 1980, p.120. tein. Arthritis Rheum 2004; 50: 1976-83. 2012; 30 (Suppl. 72): S3-6. 20. HÄFNER R: Juvenile spondarthritis. Retro- 29. SIMS AM, TIMMS AE, BRUGES-ARMAS J et 38. CHAE JJ, CHO YH, LEE GS et al.: Gain-of- spective study of 71 patients. Monatsschr al.: Prospective meta-analysis of interleukin function Pyrin mutations induce NLRP3 Kinderheilkd 1987; 135: 41-6. 1 gene complex polymorphisms confirms as- -independent interleukin-1β activa- 21. O’SHEA FD, BOYLE E, RIARH R et al.: sociations with ankylosing spondylitis. Ann tion and severe autoinflammation in mice. Comparison of clinical and radiographic Rheum Dis 2008; 67: 1305-9. Immunity 2011; 34: 755-68. severity of juvenile-onset versus adult-on- 30. COSAN F, USTEK D, OKU B et al.: Associa- 39. KIRINO Y, ZHOU Q, ISHIGATSUBO Y et al.: set ankylosing spondylitis. Ann Rheum Dis tion of familial Mediterranean fever-related Targeted resequencing implicates the famil- 2009; 68: 1407-12. MEFV variations with ankylosing spondyli- ial Mediterranean fever gene MEFV and the 22. van der LINDEN S, VALKENBURG HA, CATS tis. Arthritis Rheum 2010; 62: 3232-6. toll-like receptor 4 gene TLR4 in Behçet A: Evaluation of diagnostic criteria for an- 31. CINAR M, DINC A, SIMSEK I et al.: The rate disease. Proc Natl Acad Sci USA 2013; 110: kylosing spondylitis. A proposal for modi- and significance of Mediterranean fever gene 8134-9. fication of the New York criteria. Arthritis mutations in patients with ankylosing spon- 40. BOOTH DR, LACHMANN HJ, GILLMORE JD, Rheum 1984; 27: 361-8. dylitis: a three-month, longitudinal clinical BOOTH SE, HAWKINS PN: Prevalence and 23. BURTON PR, CLAYTON DG, CARDON LR study. Rheumatol Int 2008; 29: 37-42. significance of the familial Mediterranean et al.: Association scan of 14,500 nonsyn- 32. MARAS Y, CALGUNERI M, AKDOGAN A et fever gene mutation encoding pyrin Q148. onymous SNPs in four diseases identifies al.: MEFV mutations frequency and effect QJM 2001; 94: 527-31.

S-164