Full text of DOI:10.1016/S0016-5085(00)70209-5

GASTROENTEROLOGY 2000;118:274–278
Clinical Phenotype Is Related to HLA Genotype in the
Peripheral Arthropathies of Inflammatory Bowel Disease
‡
§
TIMOTHY R. ORCHARD,* S. THIYAGARAJA,* KENNETH I. WELSH, B. PAUL WORDSWORTH,
࿣
J. S. HILL GASTON, and DEREK P. JEWELL*
*
‡
Gastroenterology Unit, University of Oxford, Radcliffe Infirmary, Oxford; Transplant Immunology Laboratory, Nuffield Department of Surgery,
Churchill Hospital, Oxford; Musculoskeletal Research Unit, Nuffield Department of Medicine, John Radcliffe Hospital, Oxford; and Department
of Rheumatology, University of Cambridge, Addenbrooke’s Hospital, Cambridge, England
§
࿣
Background & Aims: The detection of phenotype-deter-
mining genes as opposed to disease susceptibility
genes requires precise phenotypic characterization of
patients. Peripheral arthropathies in inflammatory bowel
disease (IBD) are well recognized and are classified
with the HLA-B*27-related spondyloarthropathies by
the European Spondyloarthropathy Study Group. How-
ever, previous HLA studies in IBD have only shown this
association with axial disease rather than peripheral
arthropathy. We recently reported a clinical classifica-
tion that describes 2 types of peripheral arthropathy,
distinguished by their natural history and articular
distribution. We now report the results of immunoge-
netic studies in these patients and compare them with
other spondyloarthropathies. Methods: IBD patients
with type 1 (n ؍ 57) and type 2 (n ؍ 45) peripheral
arthropathy were identified by case note review and
questionnaire. Patients and 603 controls from Oxford-
shire were assigned HLA-A, -B, -C, -DR, and -DQ
genotypes by sequence-specific primer polymerase
chain reaction. Patient results were compared with
controls (corrected for multiple comparisons), then
with each other in light of existing hypotheses. The
results were compared with those of a cohort of 30
patients with postenteric reactive arthritis (ReA) and
ease susceptibility (in this case, HLA genes). Type 1
arthropathy is clinically and immunogenetically similar
to the spondyloarthropathies, but different HLA associa-
tions may define phenotypically distinct groups. Type 2
arthropathy has different HLA associations and may
have a different etiology. Further studies are now
required to confirm these associations and to elucidate
the different pathogenetic mechanisms.
lcerative colitis (UC) and Crohn’s disease are charac-
U
terized by intestinal, and sometimes extraintestinal,
inflammation. There is now good evidence that genetic
susceptibility is important in the pathogenesis of these
disorders and that several genes are involved.2 Microsat-
ellite screening of the genome has revealed several loci in
linkage with these diseases. Some of these loci are
common to both UC and Crohn’s disease (presumably
determining overall susceptibility to inflammatory bowel
–4
5
–7
disease [IBD]), but others appear to be disease specific.
However, other genes may influence the clinical course of
the disease without conferring susceptibility, but detec-
tion of these ‘‘phenotype-determining genes’’ depends on
precise clinical characterization of patients. In IBD, this
has been well illustrated by the conflicting data from
previous studies of HLA associations, possibly due to
ethnic differences in the populations studied and the
failure to study a sufficient number of patients to allow
meaningful analysis between phenotype and genotype. A
previous study from Oxford did describe specific pheno-
type-genotype associations in UC, although not in Crohn’s
disease.8
1
6 patients with IBD-associated ankylosing spondylitis
(IBD-AS). Results: Type 1 arthropathy was associated
with HLA-DRB1*0103 (DR103; a rare subtype of DR1)
in 33% (P F 0.0001; relative risk [RR], 12.1), B*35 in
3
0% (P ؍ 0.01; RR, 2.2), and B*27 in 26% (P ؍ 0.001;
RR, 4.0). In contrast, type 2 was associated with
HLA-B*44 in 62% (P ؍ 0.01; RR, 2.1). Similar signifi-
cant associations to type 1 arthropathy were found in
ReA, except that the HLA-B*27 association was signifi-
cantly stronger and an association was found with
DRB1*0101 (DR1) in 43% (P ؍ 0.001; RR, 2.2). IBD-AS
was associated only with HLA-B*27 and DRB1*0101.
Conclusions: These data suggest that the clinical
classification into type 1 and type 2 arthropathies
describes immunogenetically distinct entities and es-
tablish that in polygenic disorders, genes may deter-
mine clinical phenotype without conferring overall dis-
The most striking of these associations was between
HLA-DRB1*0103 (a rare subtype of DR1) and extraint-
estinal manifestations (arthropathy, erythema nodosum,
uveitis).8 For patients with arthropathy, 27.2% pos-
,9
Abbreviations used in this paper: PeA, peripheral arthropathy;
ReA, reactive arthritis.
2
000 by the American Gastroenterological Association
௠
0016-5085/00/$10.00

February 2000
ARTHROPATHY IN IBD 275
sessed HLA-DRB1*0103 (DR103) compared with 8.6%
for UC as a whole and 3.3% for healthy controls.9
30 patients with type 2 PeA. Briefly, sequence-specific primers
for the HLA alleles were used in polymerase chain reaction with
patient DNA and Taq polymerase (Bioline, London, England).
One hundred forty-four separate reactions per patient allow the
determination of HLA class I and II typing as described by
The peripheral arthropathies (PeAs) of IBD are often
classified with the HLA-B*27–related spondyloarthropa-
1
thies. However, previous association studies have failed
1
5
Bunce et al. The products were then run on 1% agarose gels
with ethidium bromide and viewed under UV light. HLA
types were assigned according to the primers that had ampli-
fied. This method has been previously validated and is in
routine use for histocompatibility testing in the United
to find any significant associations with HLA-B*27.10
This may be a result of the small numbers of patients
studied and the failure to characterize the different forms
of arthropathy accurately. We recently reported a clinical
classification of PeAs in IBD that describes 2 types,
distinguished by their articular distribution, natural
15
Kingdom.
The results were analyzed by the Fisher exact test against a
panel of 603 normal controls from Oxfordshire, using 2 ϫ 2
contingency tables for alleles of interest (HLA-B*27, -B*35,
and -DRB1*0103), with the Bonferroni correction for multiple
comparisons for other alleles (correction factor of 33 for the
HLA-B locus). Further hypotheses were generated and tested
against 92 IBD disease controls (52 UC patients and 40 Crohn’s
disease patients) and each other. The disease controls were
patients randomly selected from the Oxfordshire IBD popula-
tion who had not suffered from IBD-associated arthropathy.
The remaining patients with arthropathy (27 type 1 and 15
type 2) were typed for HLA-B*27, -B*35, -B*44, -DRB1*0101
1
1
history, and disease associations (Table 1). In this
classification, type 1 PeA is clinically similar to posten-
teric reactive arthritis (ReA), which is associated with
HLA-B*27 in up to 80% of cases, depending on the
initiating organism and severity of the arthropathy,12–14
and with HLA-B*35 in some studies. This classification
may facilitate the demonstration of HLA associations
previously masked by allowing the definition of more
homogeneous clinical groups. This study was performed
to identify the HLA associations of type 1 and type 2 PeA
of IBD, with the a priori hypothesis that type 1 is
associated with HLA-B*27, -B*35, and -DRB1*0103.
In addition, the study aimed to determine whether HLA
status influences the clinical course of the arthropathies
and to elucidate their relationship to the other spondylo-
arthropathies.
(
DR1), and -DRB1*0103 (DR103) by the same methods and
were analyzed as a cohort to replicate the findings from the first
cohort. The results were then combined with the first cohort to
undertake analysis by disease type (UC and Crohn’s disease).
A further cohort of 30 patients with established postenteric
ReA and the 16 patients with IBD-AS also underwent HLA
genotyping for HLA-B*27, -B*35, -DRB1*0101, and
Materials and Methods
-DRB1*0103, and the results were compared with the results
for IBD arthropathy and healthy controls.
From our population of 976 UC patients and 483
Crohn’s disease patients, 59 patients with type 1 PeA were
identified, 46 with type 2 PeA, and 16 patients with IBD-
associated ankylosing spondylitis (IBD-AS). Blood was ob-
tained from 57 type 1 patients, 45 type 2 patients, and 16
patients with IBD-AS (for which ethical approval had been
given). Patients were white subjects from Northern Europe,
with less than 5% having Jewish ancestry. DNA was separated
from peripheral blood leukocytes by a salting-out technique
and was dissolved in sterile water with concentrations of
Results
In the first cohort, when compared with controls
(with correction for multiple comparisons where appropri-
ate), type 1 arthropathy was significantly associated with
possession of HLA-B*27 (27% vs. 7%; P ϭ 0.001;
relative risk [RR], 4.0), HLA-B*35 (33% vs. 15%;
P ϭ 0.01; RR, 2.2), and HLA-DRB1*0103 (40% vs.
3%; P Ͻ 0.0001; RR, 12.1). Type 2 arthropathy was
associated with possession of HLA-B*44 (63% vs. 31%,
10–100 ng/µL. HLA-A, -B, -C, -DR, and -DQ genotyping was
performed using the phototyping technique described by
15
Bunce et al. on initial cohorts of 30 patients with type 1 and
P ϭ 0.0005, P ϭ 0.017; RR, 2.1). These differences
c
were also significant when comparisons were made
against the 92 disease controls (B*27, P ϭ 0.004; B*35,
P ϭ 0.02; DRB1*0103, P Ͻ 0.0001; B*44, P ϭ 0.0001)
and when the 2 PeA groups were compared (B*27,
P ϭ 0.03; B*35, P ϭ 0.02; DRB1*0103, P ϭ 0.0001;
and B*44, P ϭ 0.0001). These associations were all
confirmed in the smaller second cohorts compared with
IBD and healthy controls (type 1 arthropathy: B*27,
26% vs. 7%, P ϭ 0.02 [RR, 3.9]; B*35, 33% vs. 15%,
P ϭ 0.025 [RR, 2.2]; DRB1*0103, 30% vs. 3%,
P Ͻ 0.0001 [RR, 8.9]; type 2 arthropathy: B*44, 60%
Table 1. Classification of Peripheral Arthropathy in IBD
Type 1 (pauciarticular)
An acute, self-limiting arthropathy lasting a median of 5 wk; affects
less than 5 joints, normally including a large joint; associated
with relapses of IBD and both erythema nodosum and uveitis
Type 2 (polyarticular)
A symmetrical seronegative polyarthropathy that runs a course
independent of IBD; tends to cause persistent symptoms with a
median duration of 3 yr; is associated with uveitis but not ery-
thema nodosum.
NOTE. Neither form is seropositive, erosive, or deforming.
1
1
Modified and reprinted with permission.

276 ORCHARD ET AL.
GASTROENTEROLOGY Vol. 118, No. 2
vs. 31%, P ϭ 0.02 [RR ϭ 1.9]). The results of the entire
population with IBD-associated PeA are shown in Table
Table 3. HLA Associations of Type 1 IBD Associated
Peripheral Arthropathy and Other
Spondyloarthropathies
2. When the groups were split and analyzed separately for
IBD
type 1PeA
n ϭ 57)
Reactive
arthritis
(n ϭ 30)
UC and Crohn’s disease, all the associations remained
statistically significant, and there were no significant
differences between UC and Crohn’s disease. Further
analysis of type 1 arthropathy was undertaken by clinical
phenotype; 17 patients had recurrent (2 or more) episodes
of arthropathy, and 40 had a single episode. The preva-
lence of DRB1*0103 in the former was 11 of 17 (65%)
and in the latter was 9 of 40 (23%) (P ϭ 0.003). No other
positive HLA associations were noted, but the prevalence
of HLA-DR15 was decreased in type 1 compared with
type 2 (7% vs. 30%; P ϭ 0.04).
IBD-AS
(n ϭ 16) (n ϭ 603)
Controls
(
HLA type
HLA-B*27
HLA-B*35
HLA-DRB1*0101
HLA-DRB1*0103
No.
%
No.
17 57
30
13 43
%
No.
%
%
a
e
f
56b
19
7c,d
15
19
10
20
26
33
18
35
9
3
9
15
19
3
g
h,i
l
10 63
1
j
6k
4
13
a
b
HLA-B*27: IBD PeA vs. ReA, P ϭ 0.009; IBD PeA vs. IBD-AS, P ϭ
0
c
d
.03; ReA vs. controls, P Ͻ 0.0001; IBD-AS vs. controls, P Ͻ 0.0001.
e
HLA-B*35: ReA vs. controls, P ϭ 0.03; RR, 2.0.
f
g
HLA-DRB1*0101: IBD PeA vs. ReA, P ϭ 0.02; IBD PeA vs. IBD-AS,
P ϭ 0.002; ReA vs. controls, P ϭ 0.001; IBD-AS vs. controls, P Ͻ
.0001.
HLA-DRB1*0103: IBD PeA vs. ReA, P ϭ 0.04; IBD PeA vs. IBD-AS,
P ϭ 0.03; ReA vs. controls, P ϭ 0.02.
h
i
The results of HLA typing in the cohort of postenteric
patients with ReA are shown in Table 3. There was a
significant increase in prevalence of HLA-B*27, -B*35,
0
j
k
l
-DRB1*0101, and -DRB1*0103 in this group compared
with controls. There is no significant difference between
HLA-DRB1*0103 or HLA-B*35 compared with type 1
IBD PeA, but there was a significant increase in HLA-
B*27 and -DRB1*0101 over and above that seen in
patients with type 1 IBD PeA. An increase in HLA-B*27
was also found in the small group of IBD-AS patients, in
which there was also a significant association with
HLA-DRB1*0101 but not with HLA-DRB1*0103 or
DRB1*09, DRB1*10, DRB1*11, DRB1*12, DRB1*13,
and DRB1*14.
No positive or negative associations were found with
any of the above.
Discussion
This study reports significant HLA associations
with the arthropathies of IBD that influence the pattern
of peripheral joint involvement and that may indicate
their relationship to other spondyloarthropathies. The
clinical phenotypes of the articular problems in UC and
Crohn’s disease are the same, and the HLA associations
are identical and are not present in IBD patients without
joint complications. This implies that genes of the HLA
region have an important role in determining the clinical
course of the articular disease in IBD, although not in
susceptibility to IBD itself. Thus, phenotype-determin-
ing genes may have a role in other polygenic disorders,
affecting the clinical outcome without conferring disease
susceptibility.
-B*35.
Other HLA-B alleles tested were B*7, B*8, B*13,
B*18, B*39, B*41, B*45, B*46, B*47, B*48, B*49,
B*50, B*51, B*52, B*53, B*54, B*55, B*56, B*57,
B*58, B*60, B*61, B*62, B*63, B*70, B*71, B*72,
B*75, B*76, B*77, B*78, and B*82.
Other HLA-DR alleles tested were DRB1*0101,
DRB1*15, DRB1*03, DRB1*04, DRB1*07, DRB1*08,
Table 2. HLA Associations of Peripheral Arthropathy in IBD
Type 1
arthropathy arthropathy
n ϭ 57) (n ϭ 45)
Type 2
IBD
controls
(n ϭ 92)
Controls
(n ϭ 603)
(
Association studies of this sort allow the detection of
associations with specific genes that may be pathogeneti-
cally important. However, they may simply reflect
linkage disequilibrium with the actual pathogenic gene
located close by. Subdividing the IBD population may
create inadvertant stratification of the population that
may lead to errors, and adequate statistical allowance for
multiple comparisons is also essential to minimize the
chances of detecting spurious associations. This study
shows that the 2 types of IBD arthritis have distinct
immunogenetic characteristics, irrespective of whether
the genes identified are pathogenetically relevant. How-
ever, the consistency of the findings and the relationship
HLA type Number
%
Number
%
Number
40
90 15
186 31d
%
Number
%
26i
33
12
7a
b
e
HLA-B*27
HLA-B*35
HLA-B*44
HLA-DRB1*
15
19
7
2
3
28
4
7
62
5
13
10
5
14f
11
j
k
h
0103
20
35l
0
0
20
3c
4
4g
a
b
c
Type 1 vs. controls: P Ͻ 0.0001; RR, 4.0; P ϭ 0.01; RR, 2.2; P Ͻ
.0001; RR, 12.1.
0
d
Type 2 vs. controls: P ϭ 0.01; RR, 2.1.
e
f
Type 1 vs. IBD controls: P ϭ 0.0004; RR, 4.8; P ϭ 0.008; RR, 2.4;
P Ͻ 0.0001; RR, 8.1.
g
h
Type 2 vs. IBD controls: P Ͻ 0.0001; RR, 5.7.
i
j
k
Type 1 vs. type 2: P ϭ 0.003; RR, 5.9; P ϭ 0.001; RR, 5.0; P Ͻ
l
0
.0001; RR, 5.1; P Ͻ 0.0001; RR, incalculable.

February 2000
ARTHROPATHY IN IBD 277
to ReA suggest that the identified associations may be of
pathogenic importance.
The class I associations of type 1 IBD arthropathy and
ReA are very similar. Previous studies of ReA have also
triggering bacteria in the synovium of ReA patients are
HLA-DR restricted,22 and also with the recurrence of
joint disease that may occur in IBD when there is an acute
relapse of gut disease.11
1
2–14
shown associations with HLA-B*27
and the B*35
It has previously been shown that up to 40% of
patients with ReA develop chronic problems, including
1
6
cross-reactive group of antigens ; therefore, classification
of type 1 PeA with the spondyloarthropathies seems
justified.
23
axial disease. Nearly all of these patients are HLA-B*27
positive. Longitudinal studies of type 1 IBD PeA are now
required to show whether the B*27-positive patients
show a predisposition to the development of axial disease.
Type 2 arthropathy is not associated with HLA-B*27
or class II alleles, and its immunogenetic characteristics
suggest an etiology distinct from the other seronegative
spondyloarthropathies. It has been suggested that HLA-
B*44, which is associated with type 2 arthropathy, may
form part of an extended haplotype with DR4 in
This study also shows strong associations between
HLA class II and type 1 IBD PeA, especially with HLA
DRB1*0103, an uncommon subtype of DR1.
DRB1*0103 is also increased in patients with postenteric
ReA, but DRB1*0101 was also increased in these
patients and in the IBD-AS group. This association was
not seen in type 1 IBD PeA.
Idiopathic ankylosing spondylitis has been associated
with DR1 in a previous study, but this was largely
attributed to linkage disequilibrium with HLA-B*27.17
Because of the linkage disequilibrium between HLA-
B*27 and DRB1*0101, it is difficult to assess the relative
importance of the class I and class II associations in the
current study. However, in peripheral disease (type 1 IBD
PeA and ReA), the DRB1*0103 association was clearly
independent of HLA-B*27. This may have important
implications for the interpretation of the HLA associa-
tions in acute peripheral joint disease of the spondyloar-
thropathies: HLA-B*27 may primarily be a determinant
of axial disease, in which it is the most potent genetic
trigger of axial inflammation. However, other class I
molecules may be able to initiate disease in the presence
of other susceptibility factors such as intestinal inflamma-
tion or increased permeability, allowing increased antigen
presentation to the intestinal immune system. This
would be consistent with the findings of Mielants et al.18
that pure axial ankylosing spondylitis is not associated
with intestinal inflammation, whereas in the spondyloar-
thropathies, generally 68% of patients were found to have
rheumatoid arthritis (particularly in association with
24,25
Felty’s syndrome).
However, in our patients, the
prevalence of DR4 in type 2 arthropathy was only 28%,
which is similar to the normal population (35%). This, in
conjunction with the nonerosive and seronegative nature
of the condition, would suggest that the primary associa-
tion is with HLA-B*44 and that the pathogenesis is
distinct from the seropositive rheumatoid diseases as well
as the seronegative spondyloarthropathies.
This study strongly supports the concept of 2 immuno-
genetically distinct forms of PeA in IBD and that HLA
genes are important determinants of these phenotypes.
Furthermore, it shows for the first time strong HLA class
II associations in both type 1 IBD PeA and other
spondyloarthropathies. Further studies are now required
to extend these results to other patient populations. For
type 1 PeA and ReA, long-term prospective studies will
be required to elucidate the role of DR1 subtypes in
determining clinical features such as axial involvement
and intestinal inflammation.
1
9
References
evidence of intestinal inflammation. It would also be
consistent with the general observation that the HLA-
B*27 association in ankylosing spondylitis is weaker in
patients with coexisting IBD than in idiopathic ankylos-
ing spondylitis.
1. Dougados M, van der Linden S, Juhlin R, Huitfeldt B, Amor B, Calin
A, Cato A, Dijkmaas B, Olivieri I, Paseo G. The European
Spondyloarthropathy Study Group preliminary criteria for the
classification of spondyloarthropathy. Arthritis Rheum 1991;34:
1218–1227.
2
. Tysk C, Lindberg E, J a¨ rnerot G, Flod e´ rus-Myrhed B. Ulcerative
colitis and Crohn’s disease in an unselected population of
monozygotic and dizygotic twins. A study of heritability and the
influence of smoking. Gut 1988;29:990–996.
In contrast with axial disease, the acute peripheral
arthritis seen in type 1 IBD PeA and ReA may primarily
be determined by class II HLA alleles, and the structur-
ally similar HLA-DRB1*0103 and -DRB1*0101 may be
the most effective initiators of disease. For the disease to
progress, there has to be active intestinal inflammation
that is associated with the induction of HLA-DR in the
3
. Thompson NP, Driscoll R, Pounder RE, Wakefield AJ. Genetics
versus environment in inflammatory bowel disease: results of a
British twin study. BMJ 1996;312:95–96.
4. Satsangi J, Grootscholten C, Holt H, Jewell DP. Clinical patterns of
familial inflammatory bowel disease. Gut 1996;38:738–741.
2
0
5. Hugot JP, Laurent-Puig P, Gower-Rousseau C, Olson JM, Lee JC,
Beaugerie L, Naom I, Dupas JL, Van Gossum A, Orholim M,
Bariti-Pellic C, Weissenbach J, Mathew CG, Lennard-Jones JE,
Cortot A, Colombol JF, Thomas G. Mapping of a susceptibility
colon and up-regulation of HLA-DR expression in the
2
1
ileum. This would be consistent with the finding that
the majority of proliferative lymphocyte responses to

278 ORCHARD ET AL.
GASTROENTEROLOGY Vol. 118, No. 2
locus for Crohn’s disease on chromosome 16. Nature 1996;379:
17. Brown MA, Kennedy LG, Darke C, Gibson K, Pile KD, Shatford JL,
Taylor A, Calin A, Wordsworth BP. The effect of HLA-DR genes on
susceptibility to and severity of ankylosing spondylitis. Arthritis
Rheum 1998;41:460–465.
821–823.
6
. Ohmen JD, Yang H, Yamomoto KK, Zhao HY, Ma Y, Bentley LG,
Huang Z, Gerwehr S, Pressman S, McEltree C, Targon S, Rotter JI,
Fischel-Galston N. Susceptibility locus for IBD on chromosome 16
has a role in Crohn’s disease but not in ulcerative colitis. Hum Mol
Genet 1996;5:1679–1683.
. Satsangi J, Parkes M, Louis E, Hashimoto L, Kato N, Welsh KL,
Bell JI, Jewell DP. Two-stage genome-wide search in inflammatory
bowel disease: evidence for susceptibility loci on chromosomes
1
8. Mielants H, Veys E, Cuvelier C, DeVos M, Gecmaer S, DeClereq L,
Schatterman L, Gyselbreche L, Elevant D. The evolution of
spondyloarthropathies in relation to gut histology. III. Relation
between gut and joint. J Rheumatol 1995;22:2279–2284.
9. Mielants H, Veys E, Cuvelier C, DeVos M, Gecmaer S, DeClereq L,
Schatterman L, Gyselbreche L, Elevant D. The evolution of
spondyloarthropathies in relation to gut histology. II. Histological
aspects. J Rheumatol 1995;22:2273–2278.
7
8
9
1
3, 7 and 12. Nat Genet 1996;14:199–202.
. Satsangi J, Welsh KI, Bunce M, Julier C, Farrant JM, Bell JI, Jewell
DP. Contribution of genes of the major histocompatibility complex
to susceptibility and disease phenotype in inflammatory bowel
disease. Lancet 1996;347:1212–1217.
. Roussomoustakaki M, Satsangi J, Welsh KI, Louis E, Fanning G,
Targan S, Landers C, Jewell DP. Genetic markers may predict
disease behaviour in patients with ulcerative colitis. Gastroenter-
ology 1997;112:1645–1653.
0. Dekker-Saeys B, Meuwissen S, Van den Berg-Loonen E, Haas WD,
Agenant D, Tytgat G. Prevalence of peripheral arthritis, sacroiliitis
and ankylosing spondylitis in patients suffering from inflamma-
tory bowel disease. Ann Rheum Dis 1978;37:33–35.
1. Orchard T, Wordsworth B, Jewell D. The peripheral arthropathies
of inflammatory bowel disease: their articular distribution and
natural history. Gut 1998;42:387–391.
2
0. Selby W, Janossy G, Mason D, Jewell D. Expression of HLA-DR
antigens by colonic epithelium in inflammatory bowel disease.
Clin Exp Immunol 1983;53:614.
2
1. Cuvelier C, Mielants H, de Vos M, Veys E, Roels H. Major
histocompatibility complex class II antigen (HLA-DR) expression
by ileal epithelial cells in patients with seronegative spondyloar-
thropathy. Gut 1990;31:545–549.
1
2
2. Hermann E, Fleischer B, Mayet WJ, Poralla T, Meyer zum Buschen-
felde KH. Response of synovial fluid T cell clones to Yersinia
enterocolitica antigens in patients with reactive Yersinia arthritis.
Clin Exp Immunol 1989;75:365–370.
1
1
1
1
1
2
3. Calin A, Fries JF. An ‘‘experimental’’ epidemic of Reiter’s syn-
drome revisited. Follow-up evidence on genetic and environmen-
tal factors. Ann Intern Med 1976;84:564–566.
2. Thomson G, Chiu B, deRubeis D, Falk J, Inman RD. Immunoepide-
miology of post-Salmonella reactive arthritis in a cohort of
women. Clin Immunol Immunopathol 1992;64:227–232.
3. Sieper J, Braun J, Wu P, Hauer R, Laitko S. The possible role of
Shigella in sporadic enteric reactive arthritis. Br J Rheumatol
24. Hillarby M, Davies E, Donn R, Grennan D, Ollier W. TAP2D is
associated with HLA-B44 and DR4 and may contribute to rheuma-
toid arthritis and Felty’s syndrome susceptibility. Clin Exp Rheuma-
tol 1996;14:67–70.
25. Fraser PA, Stern S, Larson MG, Marcus-Bagley D, Awdeh Z, Glass
DN, Alper CA. HLA extended haplotypes in childhood and adult
onset HLA-DR4 associated arthropathies. Tissue Antigens 1990;
35:56–59.
1993;32:582–585.
4. Herrlinger J, Asmussen J. Longterm prognosis in Yersinia arthri-
tis: clinical and serological findings. Ann Rheum Dis 1992;51:
1332–1334.
5. Bunce M, O’Neill CM, Barnardo MC, Krauser P, Browning MJ,
Morris PJ, Welsh KI. Phototyping: comprehensive DNA typing for
HLA-A, B, C, DRB1, DRB3, DRB4, DRB5 & DQB1 by PCR with 144
primer mixes utilizing sequence-specific primers (PCR-SSP). Tis-
sue Antigens 1995;46:355–367.
6. Mielants H, Veys E, Joos R, Noens L, Cuvelier C, DeVos M. HLA
antigens in seronegative spondyloarthropathies. Reactive arthri-
tis and arthritis in ankylosing spondylitis: relation to gut inflamma-
tion. J Rheumatol 1987;14:466–471.
Received July 2, 1999. Accepted October 27, 1999.
Address requests for reprints to: T. R. Orchard, M.D., University of
Oxford, Radcliffe Infirmary, Gastroenterology Unit, Woodstock Road,
Oxford OX2 6HE, England. Fax: (44) 1865-224829.
Supported by a grant from the National Association for Crohn’s
and Colitis and the Arthritis Research Campaign.
1