ORIGINAL ARTICLE

Human leukocyte antigen distribution in Israeli patientswith psoriatic arthritis

Received: 17 October 2002 / Accepted: 20 March 2003 / Published online: 17 June 2003� Springer-Verlag 2003

Abstract Objectives This study was designed to investi-gate the distribution of human leukocyte antigen (HLA)classes I and II in a group of Israeli Jewish patients withpsoriatic arthritis (PsA) and identify HLA markersrelated to disease manifestation in PsA.

Patients and methods Human leukocyte antigensclass I and class II (both serologically and from oligo-typing) were tested in a group of 50 consecutive patientswith PsA, 32 with skin psoriasis (PSO), and 255 healthypersons. Data on age, gender, disease duration, andpattern of rheumatological manifestations—oligoar-thritis, polyarthritis, spinal involvement, involvement ofdistal interphalangeal joints (DIPs), and enthesi-tis—were registered.

Results Human leukocyte antigens A3, B13, and B38alleles were found to be significantly prevalent in PsAcompared with PSO patients and healthy controls.HLA-B27 was found in only two out of 50 patients withPsA. Patients with PSO and PsA had significantlyincreased incidence of HLA-DRB0101 and -DRB0301,while the frequency of HLA-DRB0403 was significantlyhigher among patients with PsA of Ashkenazi origin. Wefound a statistically significant association between DIPinvolvement and the presence of HLA-A26 and -B38,while HLA-DRB0301 was related to spinal involvement.

Conclusions Psoriatic arthritis in Israeli patients seemsto be associatedwith the presence ofHLA-A3, -B13, -B38,

-DRB0101, and -DRB0301. HLA-B27 was not a markerof PsA in this cohort of patients, including patients withpsoriatic spondyloarthropathy.

Keywords Distal interphalangeal joints Æ Humanleukocyte antigen Æ Psoriasis

Introduction

Psoriasis (PSO) is a common disease affecting 1–3% ofthe population [1]. The association between psoriasisand joint disease was first recognized by Alibert [2] andlater defined as an inflammatory arthritis, usually sero-negative for rheumatoid factor, associated with psoriasis[3].

The pathogenesis of psoriatic arthritis (PsA) is notclear, but genetic, environmental, and immunologicalfactors are considered to play a role in development andperpetuation of the disease [4, 5]. The pronounced dif-ference in concordance rates of PsA between monozy-gous and dizygotic twins [6] and the clustering of bothpsoriasis [7] and PsA in families provide strong evidenceof the importance of genetic factors in PsA. Further-more, several studies reported linkage between specificHLAs and PsA [8, 9]. However, divergent distributionsof HLAs were documented among different cohorts,suggesting that HLA distribution may depend uponethnic origin [8, 9].

The Jewish population of Israel includes inhabitantsof heterogeneous ethnic background [10]. The distribu-tion of HLA antigens in different diseases may differfrom that reported in other populations. For example, inrheumatoid arthritis (RA), HLA-DR1 is prominent inIsraeli patients [11], while in other Caucasian popula-tions, HLA-DR4 is clearly related to RA [12].

The aim of this study was to investigate the distri-bution of HLA classes I and II both serologically and byoligotyping in a group of Israeli patients with PsA andto compare it to that of a normal control population and

Rheumatol Int (2004) 24: 93–97DOI 10.1007/s00296-003-0325-0

Ori Elkayam Æ Refael Segal Æ Dan Caspi

O. Elkayam Æ D. CaspiDepartment of Rheumatology, Sourasky Medical Center,Tel Aviv, Israel

R. SegalShmuel Harofeh Geriatric Medical Center,Beer-Yaakov, Israel

O. Elkayam Æ R. Segal Æ D. CaspiSackler Faculty of Medicine,Tel Aviv University, Tel Aviv, Israel

O. Elkayam (&)Department of Rheumatology, Tel Aviv Medical Center,6 Weizman Street, Tel Aviv 64239, IsraelE-mail: oribe14@netvision.net.ilFax: +972-36974577

PSO patients without joint involvement. Secondly, wetried to identify specific HLA markers associated withthe variety of rheumatic manifestations among patientswith PsA.

Material and methods

Patients

Fifty consecutive, unrelated patients with PsA seen at theDepartment of Rheumatology during 1996–1997 were consecu-tively and unselectively recruited and evaluated. All patients ful-filled the currently accepted criteria for psoriasis—defined as thepresence of typical skin lesions confirmed by a dermatologist—andPsA, namely an inflammatory arthritis, usually rheumatoid factor-negative, associated with psoriasis [3]. Other joint diseases such astypical RA, systemic lupus erythematosus, and gout were excluded.

Controls

The control group was comprised of 32 unrelated patients withchronic plaque psoriasis and 255 healthy subjects from the sameethnic background as the patients.

Human leukocyte antigen typing

Typing for HLA class I (A, B, C) antigens was performed by thestandard two-stage microlymphocytotoxic assay [13] using sera oflocal origin and from other laboratories. Genomic DNA was iso-lated from whole blood by a salting out procedure as described byMiller et al. [14]. The DRB and DQB genes were amplified usingthe 11th International Histocompatibility Workshop primersprobes and recommended amplification profiles [15]. The amplifiedDNA was denatured and blotted on nylon membrane filters. Thefilters were then prehybridized and hybridized with three-tailedsequence-specific oligonucleotide probes 9SSOP with dig-ddUTP[16]. Dots were visualized by means of chemiluminescence detectionafter 5.010-min exposure on Kodak XAR-5 film.

Clinical evaluation

The clinical charts of the 50 patients with PsA were thoroughlyevaluated using specially designed forms recording age, sex, durationof skin and joint disease, family history of psoriasis and PsA, and ageat onset of both skin and joint disease. Patients were dividedaccording to the classification of Wright into five patterns: oligoar-thritis—patients with four or fewer involved joints, polyarthri-tis—patients with five or more affected joints, patients with spinalinvolvement, distal interphalangeal joint (DIP) PsA, and arthritismutilans. Since the numbers of patients in the last two categories wassmall, theywere included in the three first groups according to clinicalinvolvement. Patients were considered to have back involvementbasedon the presence of grade 2or higher sacroiliitis alone, or grade 1sacroiliitis accompanied by syndesmophytes and/or inflammatoryback pain (defined as back pain and stiffness not relieved by rest).

Enthesitis was defined as plantar fasciitis or tendoachilles ten-dinitis, while cervical spondylitis was defined as inflammatorysymptoms attributed to the cervical spine with evidence of ten-derness and/or range limitation on physical examination.

Statistical analysis

Comparison between origin groups and other background vari-ables regarding the prevalence of the various alleles was performedusing the chi-squared or Fisher’s exact test in which the number of

expected observations per cell was less than five. This analysis wascarried out for each allele separately and for the distribution ofalleles in each locus. Gene frequencies were calculated according tothe formula gf=1-sqrt(1-af) as described by Baur and Danilovs.Comparison between observed allele frequencies and expectedfrequencies (Roitberg-Tambur et al.) was done using the multi-nomial test.

Results

The demographic and clinical data of patients with PsAare shown in Table 1 and Table 2. The most commonpattern of joint involvement was oligoarthritis. Evidenceof inflammatory back involvement was found in 18 pa-tients (36%). The differences in distribution of HLAgene frequencies in 50 patients with PsA, compared to 32with psoriasis and 255 healthy unrelated persons, aresummarized in Table 3 and Table 4.

Human leukocyte antigen class I

As can be seen in Table 3, we found a statistically sig-nificantly increased incidence of HLA-A3, -B13, and -B38 in patients with PsA in comparison with those withpsoriasis vulgaris or healthy controls. HLA-A3 and -B13alleles were found, respectively, in 12% and 7% ofAshkenazi patients with PsA, in comparison with 5.7%

Table 1 Demographic characteristics of 50 patients with PsA

Male:female 30:20Mean age (years) 58±15Ashkenazi:Sephardi 34:16Age at onset skin lesions (years SD) 36±16Age at onset joint disease (years SD) 40±15

Table 2 Clinical patterns of psoriatic arthritis. DIP distal inter-phalangeal

Oligoarthritis 38 (76%)Polyarthritis 12 (24%)Spinal involvement 18 (36%)DIP involvement 33 (63%)

Table 3 Different gene frequencies (%) of serologically definedHLA antigens in patients with psoriatic arthritis (PsA), psoriasis(PSO) and healthy controls. S Sephardic, A Ashkenazi

HLA Controls PsA PSO

S A S A S A

A3 12.8 5.7* 20 12* 10 0**B13 7.4 1.6* 6. 7* 0 0**B38 7.8* 16.3 20* 21 0** 35B70 0.4 0.4 0 0 0 5**Cw6 13.3 14 13 19 15 0**

*P<0.05 between PsA and healthy controls**P<0.05 between patients with PsA and PSO

94

and 1.6% in healthy controls, respectively, and 0% inpatients with psoriasis. HLA-B38 was found in 20% ofnon-Ashkenazi patients with PsA, as compared to 7.8%of healthy controls and 0% of psoriatic patients(P<0.05). HLA-B27 was positive in only two patient-s—one of them with spinal involvement—in comparisonwith 3% in the healthy population (Table 3).

Human leukocyte antigen class II

A significantly higher incidence of HLA-DRB0101 wasfound in both PsA and PSO than in healthy subjects.This was shown in both Ashkenazi and non-Ashkenazipatients (6.4% of Sephardic and in 10% of Ashkenazipatients with PsA, compared with 1.9% and 0.5% inhealthy controls). We could also show a significant in-crease in HLA-DRB0301 among non-Ashkenazi pa-tients with psoriasis and PsA in comparison with healthycontrols. On the other hand, HLA-DRB0403 was sig-nificantly increased in Ashkenazi patients with PsA,while the rate in PSO patients was similar to that ofhealthy controls (Table 4).

Correlation between human leukocyte antigensand clinical variables

Human leukocyte antigen class I

We found a statistically significant association betweenDIP involvement and the presence of HLA-A26 and -B38. HLA-A26 was found in ten patients, nine of them(90%) with DIP involvement, while 21 out of 38 patients(55%) who were HLA-A26-negative showed DIPinvolvement (P=0.04). Likewise, 15 of 18 patients(83%) with HLA-B38 had DIP involvement, vs 50% inHLA-B38-negative patients (P=0.02). A significant

association was found between HLA-B35 and axialinflammatory involvement (P<0.05). HLA-B27 wasextremely rare in our patients, and no association wasfound between its presence and clinical or radiologicalsacroiliac or spinal involvement. Only one patient withinflammatory and radiological spinal involvement wasHLA-B27-positive, while 17 were negative.

Human leukocyte antigen class II

HLA-DRB0301 was found to be associated with spinalinvolvement. Five out of six patients with HLA-DRB0301 (84%) presented inflammatory back pain, ascompared to 25% of those who were HLA-DRB0301-negative (P =0.04). HLA-DRB0301 was also found tobe increased in patients with inflammatory spinalinvolvement (67% vs 8% in patients with other clinicalpatterns of joint involvement, P=0.01). Although HLA-DRB0101 was clearly related to PsA, we could not findany specific clinical characteristics of this antigen.Likewise, HLA distribution was similar within groups ofpatients with oligoarthritis or polyarthritis.

Discussion

In our study on 50 patients with PsA, we have shown anassociation between PsA and HLA-A3, -B13, -B38,-DRB0101, -DRB0301, and -DRB0403. The proportionof patients with HLA-B27 was low and similar to that ofthe general population. We could show a significantassociation between HLA-A26, HLA-B38, and DIPinvolvement as well as between HLA-B35, HLA-DRB0301, and inflammatory spinal involvement.

Our results are concordant with other reports ofincreased frequency of HLA-B13 in PsA [17, 18] as wellas an association between HLA-B38 and the polyar-thritis pattern of PsA [9, 19, 20]. However, in severalways our results differ from most of the known asso-ciations between HLA antigens and PsA, supportingthe assumption of this study that Israeli patients withPsA may present different HLA associations, which ispossibly related to the varied ethnic origin of thispopulation.

One of the most striking findings of this study was thelow frequency of HLA-B27 in our cohort of patients. Innone of the rheumatic diseases has the genetic contri-bution to pathogenesis been so well characterized as inseronegative spondyloarthropathies, in which HLA-B27has a central role [21]. Although the frequency of HLA-B27 in PsA is lower than in ankylosing spondylitis orReiter’s syndrome, it still is considered a predisposingfactor to PsA, especially with axial involvement [22,23, 24, 25]. However, the clear relationship betweenspondyloarthropathy and HLA-B27 is based on Cau-casian populations, in which the background frequencyof this gene is 8–11%. In our cohort, only two out of 50patients (4%) were HLA-B27-positive, although 18

Table 4 Different gene frequencies (%) of HLA-DR and -DQ allelevariants in patients with psoriatic arthritis and psoriasis and heal-thy controls. S Sephardic, A Ashkenazi, NT not tested

HLA Controls PsA PSO

S A S A S A

DRB10101 0.5 1.9 6.4* 10* 6.4 7DRB10301 2.7 7.1 9.8* 4.9 6.6 15DRB10402 2.7 1.9 0 0 0 21**DRB10403 2.7 1.9 3.1 6.6* 3.3 0**DRB10405 2.2 0.4 0 0 10** 14**DRB10406 3.2 0 0 0 20** 14**DRB11401 4.9 2.7 9.8 1.6 0 15**DRB10402 2.2 1.1 0 0 0 21**DQB10503 NT NT 2.5 18** 0 5DQB10201 23.5 18.4 0 0 3 7**

*Statistically significant (P<0.05) between patients psoriatic ar-thritis and healthy controls**Statistically significant (P<0.05) between psoriatic arthritis andpsoriasis vulgaris

95

(36%) had evidence of axial involvement. This findingmay reflect the low representation of HLA-B27 in theIsraeli population (0.9–3.2%) and in our healthy con-trols [26]. Following the same trend, the frequency ofHLA-B27 among Israeli patients with ankylosingspondylitis is only 79%, in comparison with 95% inother Caucasian populations [24]. Although it is signif-icantly higher than in the Israeli population, it is muchlower than is found in other populations of ankylosingspondylitis [27].

With regard to the expression of class II HLAs, wecould clearly show a statistically significant increase inthe distribution of HLA-DRBO1O1 and -DRB0403,which were found in 10% and 6.6% of Ashkenazi pa-tients, respectively, while their prevalence in the Israelipopulation is less than 2%. HLA-DRB0101 is one of thealleles previously characterized as HLA-DR1 in theformer classification of HLA class II. Interestingly,HLA-DR1 is also frequent in Israeli patients with RA,although the predominant HLA-DR1 subtype in thesepatients is DRB1*0102 [28]. We could find no increasedincidence of HLA-DR4, as reported by others [1].Gladman et al. reported that HLA-DRB0401 and-DRB0402 are the most frequent alleles found in PsA.Sequencing studies in Caucasians have shown thatHLA-DR4 and -DR1 antigen molecules share a com-mon amino sequence in the third hypervariable region ofthe DR molecule, suggesting that this sequence is pri-marily associated with RA [28]. The same argument maybe true in PsA.

Some reports have suggested an association betweencertain HLA and the degree of severity of PsA. Erosivedisease has been associated with HLA-DR4 and -B27 inthe presence of HLA-DR7, while HLA-B22 seems toprovide protection through all states [29]. Our study wasnot designed to investigate longitudinally the prognosticvalue of HLAs. However, if polyarthritis is consideredthe most severe form of PsA, we could not demonstratea clear correlation between any HLA and the degree ofseverity of PsA in our cohort.

In conclusion, in 50 patients with PsA, we showedincreases in the distribution of HLA-B13, -B38,-DRB0101, and -DRB0301, while HLA-B27 and -DR4were distributed in a way similar to that of the healthyIsraeli population. These results suggest that, as inrheumatoid arthritis, Israeli patients with PsA present adifferent HLA distribution than reported until now. Thisstudy included a relatively small number of patients, andfurther studies based on larger populations of Israelipatients with PsA may be warranted.

References

1. Gladman D, Farewell VT, Nadeau C (1995) Clinical indicatorsof progression in psoriatic arthritis—multivariate relative riskmodel. J Rheumatol 22:675–679

2. Alibert JL (1918) Precis theorique sur les maladies de la peau.Ravier, Paris

3. Gladman DD, Shuckett R, Russel ML, Thorne JC, ShachterRH (1987) Psoriatic arthritis—an analysis of 220 patients. Q JMed 62:127–141

4. Espinoza LR (1985) Psoriatic arthritis: further epidemiologicand genetic consideration. In: Gerber LH, Espinoza LR (eds)Psoriatic arthritis. Grune and Stratton, Orlando, pp 9–32

5. Gladmam DD (1985) The role of immunological factors in thepathogenesis and etiology of psoriatic arthritis. In: Gerber LH,Espinoza LR (eds) Psoriatic arthritis. Grune and Stratton,Orlando, pp 33–44

6. Farber EM, Nall ML (1971) Genetics of psoriasis—twin study.In: Farber EM, Cox AJ (eds) Psoriasis. Proceeding of theInternational Symposium, Stanford University. Stanford Uni-versity Press, Stanford, pp 7–13

7. Moll JMH, Wright V (1973) Familial occurrence of psoriaticarthritis. Ann Rheum Dis 32:181–201

8. Kammer GM, Soter NA, Gibson DJ et al (1979) Psoriaticarthritis: a clinical, immunologic and HLA study of 100 pa-tients. Semin Arthritis Rheum 9:75–97

9. Gladman DD, Anhorn KAB, Schachter RK, Mervart H(1986) HLA antigens in psoriatic arthritis. J Rheumatol13:586–592

10. Roiterg-Tambur CS, Friedman AW, Safirman C, Sherman L,Battat S, Nelken D, Brautbar C (1995) Comparative analysisof HLA polymorphism at the serologic and molecular levelin Moroccan and Ashkenazi Jews. Tissue Antigens 46:104–110

11. Schiff B, Mizrahi Y, Orgad S, Yaron M, Gazit E (1982)Association of HLA-Aw31 and HLA-DR1 in patients withadult rheumatoid arthritis. Ann Rheum Dis 4:403–404

12. Roudier J (2000) Association of MHC and Rheumatoidarthritis with HLA-DR4: the role of repertoire selection.Arthritis Res 2:217–220

13. Terasaki PI, Bernoco D, Park MS, Ozturk G, Iwaki Y (1978)Microdoplet testing for HLA-A, B, C and D antigens. Isr J ClinPathol 69:103–120

14. Miller SA, Dykes DD, Polesky HF (1988) A sampling saltingout procedure for extracting DNA from human nucleated cells.Nucleic Acids Res 16:1215

15. Kimura A, Sasazuki T (1991) Eleventh International Histo-compatibility Workshop reference protocol for the HLA-DNAtyping technique. In: HLA 1991, Proceedings of the EleventhInternational Histocompatibility Workshop Conference, vol 1.Oxford University Press, New York, pp 397–419, 1991

16. Eleventh International Histocompatibility Workshop Confer-ence (1991) HLA 1991, vol 1. Oxford University Press, NewYork

17. Roux H, Mercier P, Maestracci D et al (1977) Psoriatic arthritisand HLA antigens. J Rheumatology 3 [Suppl 1]:64–65

18. O’Donnel, O’Loughlin S, Codd MB, Powell FC (1993) HLAtyping in Irish psoriatics. Ir Med J 86:65–68

19. Espinoza LR, Vasey FB, Wilkinson OR, Osterland CH (1978)Association between HLA-B38 and peripheral psoriaticarthritis. Arthritis Rheum 21:72–75

20. Murray C, Mann DL, Gerber LN, Barth W, Perlmann S,Decker JL, Nigra TP (1980) Histocompatibility alloantigens inpsoriasis and psoriatic arthritis. Evidence for the influence ofmultiple genes in the major histocompatibility complex. J ClinInvest 66:670–675

21. Braun J, Bollow M, Reminger G, Eggens U, Rudwaleit M,Distler A, Sieper J (1998) Prevalence of spondyloarthropathiesin HLA-B27 positive and negative blood donors. ArthritisRheum 41:58–67

22. Lambert JR, Wright S, Rajah SM, Moll MH (1976) Histo-compatibility antigens in psoriatic arthritis. Ann Rheum Dis35:526–530

23. Eastmond CJ, Woodrow JC (1977) The HLA system and thearthropathies associated with psoriasis. Ann Rheum Dis36:112–120

24. Amstrong RD, Panayi GS, Welsh KI (1983) Histocompatibilityantigens in psoriasis, psoriatic arthritis and ankylosing spon-dylitis. Ann Rheum Dis 42:142–146

96

25. Marsal S, Armandans-Gil L, Martinez M, Gallardo D, RiberaA, Lience E (1999) Clinical, radiographic and HLA associa-tions as markers for different patterns of psoriatic arthritis.Rheumatology (Oxford) 38:332–337

26. Brautbar C, Porat S, Nerken D, Gabriel KR, Cohen T (1977)HLA B27 and ankylosing spondylitis in the Israeli population.J Rheumatol Suppl 3:24–32

27. Schlosstein L, Terasaki PI, Bluestone R et al (1973) Highassociation of an HLA antigen w27 with ankylosing spondyli-tis. N Engl J Med 288:704–706

28. De Vries N, Ronningen KS, Tilanus MG, Bouwns-RomboutsA, Segal R, Egeland T, Thorsby E, van de Putte LB, BrautbarC (1993) HLA-DR1 and rheumatoid arthritis in Israeli Jews:sequencing reveals that DRB1*0102 is the predominant HLA-DR1 subtype. Tissue Antigens 41:26–30

29. Gladman D, Farewell VT, Kopciuk KA, Cook RJ (1998) HLAmarkers and progression in psoriatic arthritis. J Rheumatol25:730–733

97