close

Вход

Забыли?

вход по аккаунту

?

Genetic typing of patients with inflammatory arthritis at presentation can be used to predict outcome.

код для вставкиСкачать
1166
ARTHRITIS & RHEUMATISM Volume 37
Number 8, August 1994, pp 1166-1 170
0 1994, American College of Rheumatology
GENETIC TYPING OF PATIENTS WITH INFLAMMATORY ARTHRITIS
AT PRESENTATION CAN BE USED TO PREDICT OUTCOME
ANDREW GOUGH, JEFF FAINT, MIKE SALMON, ANDREW HASSELL, PAUL WORDSWORTH,
DARRELL PILLING, ANDREW BIRLEY, and PAUL EMERY
Objective. To test the hypothesis that genetic
characterization of patients at the time they present with
inflammatory arthritis can predict subsequent destructive disease.
Methods. We evaluated 177 patients with early
arthritis. Patients were serologically tested for rheumatoid factor (RF) and were DNA oligotyped for the
presence of conserved base sequences in the third hypervariable region (HVR3) of the DRBl gene, previously shown to be associated with severe rheumatoid
arthritis (RA). Homozygosity in the patient’s genotype
was confirmed usiiig restriction fragment length polymorphism analysis. The main outcome measure was
radiologic erosions at 1 year.
Results. At presentation, 120 patients fulfilled the
American College of Rheumatology 1987 criteria for
RA, 64% of whom possessed the conserved base sequences, compared with 45% of 347 healthy controls
(P < 0.001) and with 56% of 57 patients with other
inflammatory arthritis (P not significant). Within the
RA population, the frequency of Dw4/Dw14 compound
heterozygotes was disproportionately increased. The
presence of either HVR3 or RF had a relative risk of
13.49 for erosions, with a sensitivity of 95% (specificity
39%); the presence of both HVR3 and RF had a relative
risk of 8.13, with a specificity of 88% (sensitivity 53%).
Supported by the Arthritis and Rheumatism Council.
Andrew Gough, MRCP: University of Birmingham and
Selly Oak Hospital, Oxford, UK; Jeff Faint, BSc: University of
Birmingham; Mike Salmon, PhD: University of Birmingham; Andrew Hassell, MRCP: University of Birmingham and Selly Oak
Hospital; Paul Wordsworth, MRCP: Institute of Molecular Medicine, Oxford; Darrell Pilling, BSc: University of Birmingham;
Andrew Birley, PhD: University of Birmingham; Paul Emery, MD,
FRCP: University of Birmingham and Selly Oak Hospital.
Address reprint requests t o Paul Emery, MD, FRCP,
Department of Rheumatology, The University of Birmingham,
Edgbaston, Birmingham, B15 2TT, UK.
Submitted for publication November 17, 1993; accepted in
revised form February 27, 1994.
All but 1patient with the Dw4/Dw14 genotype developed
erosions within 1 year.
Conclusion. Knowledge of a patient’s HLA-DR
type and RF status allows clinically useful prediction of
erosive disease; patients possessing Dw4/Dw14 represent
a particularly high-risk subgroup.
The genetic component of rheumatoid arthritis
(RA) has been investigated in many cross-sectional
studies. The initial demonstration was of an increased
prevalence of HLA-Dw4 in patients with RA (l), and
this was confirmed for the closely related serologically
determined HLA-DR4 (2). More recently, it has been
shown that RA is positively associated only with
certain subtypes of DR4, namely Dw4 (0401), Dw14
(0404,0408), and Dw15 (0405) (3,4), and that these
alleles encode a conserved amino acid sequence in the
third hypervariable region of the DRP chain. This
sequence is shared with other alleles associated with
the disease (namely, subtypes of DRl [OIOl, 01021,
DR6,Dw16 [1402], and DRlO [I0011 [5-7]), but not
with those subtypes of DR4 which are not disease
associated (DwlO and Dw13). These observations
have led to the so-called “shared epitope” hypothesis
(8). This epitope has since been demonstrated to be
present in 80-90% of patients with established RA (9).
These data were derived from patients with
chronic disease, although it had been assumed that this
was a consequence of a major histocompatibility complex (MHC) effect on predisposition or susceptibility
to disease. The observation that the strongest association between MHC and RA is with the most severe
forms of disease (10-12) has led to speculation concerning the timing of the effect of the MHC. In a small
selected group of patients, we recently showed that
the genetic factors were predictive of progressive
destructive disease (13). The present study was instigated to test the hypothesis that genetic typing of
patients has the potential to produce clinically useful
HLA GENOTYPE PREDICTS OUTCOME IN INFLAMMATORY ARTHRITIS
predictions of outcome in the routine setting of an
early arthritis clinic, and to assess the timing of the
major impact of the class 11 genes.
Table 1. HVR3 in inflammatory arthritis patients at presentation*
Group
PATIENTS AND METHODS
Patients. The patients included in this study were
unselected consecutive patients with clinical evidence of
inflammatory joint disease who were referred from primary
care physicians to an early inflammatory arthritis clinic.
Patients were seen prior to the initiation of slow-acting drugs
or steroids. At presentation they were assessed for the 1987
criteria of the American College of Rheumatology (ACR;
formerly, the American Rheumatism Association) for a
diagnosis of RA (14), and underwent a full clinical assessment. Blood was taken for rheumatoid factor (RF) (15) and
HLA analysis, and radiographs of the hands and feet were
performed. Assessments were repeated at 1, 3, 6, and 12
months after presentation, except for radiographs, which
were repeated at 12 months. Healthy subjects from Birmingham and Oxford were used for HLA control data.
HLA-DR typing. Genomic DNA was obtained from
frozen whole blood and HLA-DRB alleles were then determined using polymerase chain reaction (PCRtamplified
DNA and sequence-specific oligonucleotide (SSO) probes,
and DR specificities and subtypes were assigned as described previously (16). Alleles which possess the conserved
sequence QKRAA and QRRAA, spanning residues 7&74 of
the third allelic hypervariable region of the DRPl chain,
were identified. This epitope is subsequently referred to as
HVR3.
Restriction fragment length polymorphism analysis for
genotype. Because SSO probes cannot identify allelic
homozygosity, we used restriction fragment length polymorphism, according to the method of Bidwell and Jarrold (17).
Briefly, genomic DNA was digested with Tuq 1, and the
fragments were resolved by electrophoresis on 0.7% agarose
gels. After denaturing, the DNA was transferred onto
Hybond-C membranes (Amersham Corp., Amersham, UK).
Primers were designed from the sequences of a full-length
DRP complementary DNA to amplify a 485-basepair probe
by PCR. The probe was then purified using a low meltingpoint agarose gel, which was then phenol extracted and
ethanol precipitated. A 400-ng volume was then radiolabeled
and hybridized to the filters. Homozygosity was determined
by comparison with known DR4 cell lines, using the same
method.
Statistical analysis. The incidence of HVR3 in the
various groups was compared using the log-likelihood ratio
test (18). The estimation of relative risk and its statistical
significance (relative incidence ratio) followed the methods
described by Woolf (19) and Svejgaard et a1 (20). The
relationship between gene dose and erosions was determined
using the regression method described by Yates for proportions (21). Deviation of the observed frequency of Dw4/
Dw 14 compound heterozygotes from the predicted frequency (assuming Hardy-Weinberg equilibrium), according
to the individual allele frequency of Dw4Dw14, was tested
by the procedure of Hernandez and Weir (22).
1167
Healthy controls
All patients
RA patients (ACR
criteria)
Other inflammatory
arthritis patients
~~
~
Total
no. of
subjects
No. (%)
with
conserved
HVR3
Loglikelihood
ratio
347
177
120
156 (45)
109 (62)
77 (64)
13.01
13.30
<0.0001
57
32 (56)
3.70
0.054
~
P
<0.001
~
* P values are versus controls. The difference between rheumatoid
arthritis(RA)patients (i.e., those who fulfilled the American College
of Rheumatology [ACR] criteria for RA) and patients with other
inflammatory arthritis did not reach significance. HVR3 = third
hypervariable region conserved epitope.
RESULTS
Characteristics at presentation. Two of the 179
consecutive referral patients with inflammatory arthritis were subsequently excluded (one left the country
and the other died of lymphoma). All but 4 were of
Caucasian origin. One hundred twenty satisfied the
1987 ACR criteria at presentation, at a mean of 7.4
months (range 3-22 months) from onset of symptoms.
Of these 120 patients, 79 (65%) were seropositive, and
38 (31%) had erosive disease at presentation. The
remaining 57 patients had an assortment of inflammatory arthritides (19 were postviral, 17 had undifferentiated connective tissue disease, 5 had sarcoidosis, 5
had other connective tissue diseases, 4 had psoriatic
arthritis, 3 had spondylarthropathy, and 4 had other
inflammatory arthritis). These 57 patients had a mean
disease duration of 4.2 months (range 2-15 months).
None of the patients were positive for DRlO
(1001) or for the Dw15 subtype of DR4 (0405). The
prevalence of HVR3 in the RA patients (64%) was
significantly increased compared with healthy controls
(45%) (P < 0.001), but was not significantly different
from the prevalence in the patients with miscellaneous
inflammatory arthritis (56%) (Table 1).
The allele frequencies are shown in Table 2.
DRl alone did not reach significance in any comparison. For RA patients, there were significant differences in Dw4 and Dw14 frequencies compared with
controls. The other inflammatory arthritis patients
were significantly different from controls for Dw4
alone (P < 0.05). Although there were no significant
differences between the two arthritis groups for Dw4,
the difference did reach significance for Dw14 (P <
0.05).
GOUGH ET AL
1168
Table 2. Allele frequencies with Z statistic comparison*
Dw4
DR 1
Controls versus all patients
Controls versus other inflammatory
arthritis patients
Controls versus rheumatoid arthritis
patients
Other inflammatory arthritis
patients versus rheumatoid
arthritis patients
v
Dw14
z
P
z
P
z
P
1.32
0.62
NS
NS
4.46
2.44
<0.001
1.93
0.36
<0.05
C0.05
0.88
NS
4.38
<0.001
2.63
<0.01
0.01
NS
0.63
NS
2.06
c0.05
NS
v
* Z statistic = (1 arc sin p, - arc p2 /)/28.648 d l h , + 1h2,where n, and n2 are the sample sizes
of populations p, and pz, respectively. NS = not significant.
The HLA-DR genotype was also examined
with a correction for the enrichment of DR4 within the
RA population (Table 3). We found a selective association of Dw4/Dw14 heterozygosity with RA (P <
0.01).
Findings at 1 year followup. Of the 177 patients
initially recruited, 117 (66%) had evidence of persistent disease (clinical or laboratory evidence of active
disease, or required therapy with slow-acting drugs).
The ACR criteria provided the greatest sensitivity
(88%) for persistent disease, with a specificity of 73%
(Table 4). The presence of HVR3 alone had a relatively low sensitivity, which did not reach significance.
In contrast, among the factors governing erosive disease, the highest relative risk (i.e., 13) was the possession of either R F or HVR3, with a high sensitivity of
95% (Table 5). However, this was not very specific
(39%). Possession of both RF and HVR3 was highly
specific for erosive disease 88% (relative risk of 8.13),
but had a sensitivity of only 53%.
Effect of gene dose. There was a significant
increase in the observed frequency of Dw4/Dw14
Table 3. Percentage HVR3 homozygotes and heterozygotes in the
120 RA patients and 347 healthy controls*
RA patients
0bser ved
Heterozygotes
DRUDw4
DRUDwl4
Dw4iDw14
Homozygotes
DR4/Dw4/Dw4
Dw 14/Dw14
Expected
Control
subjects
4.2
2.5
9.27
4.6
1.8
4.6T
2.7
1.1
1.1
5.8
0.8
5.8
2.1
1.4
0.2
* Rheumatoid arthritis (RA) patients fulfilled the American College
of Rheumatology criteria. HVR3 = third hypervariable region
conserved epitope.
7 Test of Hardy-Weinberg deviation: 2, = 6.78; P = 0.01.
heterozygotes within the RA population over that
predicted by the Hardy-Weinberg equation from the
individual allele frequencies (Table 3). Of the 11 patients with Dw4Dw14, all but l developed destructive
disease within 1 year. There was evidence for a linear
[I] =
gene-dose effect on the incidence of erosions
6.78; P = 0.01). Similar effects of gene dose and the
development of erosions were found for the presence
of any of the Dw4 and Dw14 alleles (1 copy or 2 copies
in heterozygous or homozygous state), and independently for each of the 2 alleles.
(2,
DISCUSSION
This study examined the role of the conserved
sequence of amino acids expressed in the HVR3 of
DRPl in a large group of consecutive patients presenting with early inflammatory arthritis. By combining
DR typing for this conserved sequence with rheumatoid factor status, clinically important predictions
could be made. Interestingly, the prevalence of this
Table 4. Predictive values for persistent disease at 12 months in
the 177 study patients, according to markers present at disease
onset*
Sensitivity Specificity Relative
(%I
(%I
risk
ACR criteria for RA
Erosions noted on
radiographs
Rheumatoid factor
HVR3
(HVR3)’
88
30
13
95
19.96
7.87
65
66
23
86
46
91
11.35
1.66
3.25
P
<0.001
<0.001
NS
<0.001
<0.05
* ACR = American College of Rheumatology; RA = rheumatoid
arthritis; HVR3 = third hypervariable region conserved epitope;
(HVR3)’ = heterozygote or homozygote for conserved epitope; NS
= not significant.
HLA GENOTYPE PREDICTS OUTCOME IN INFLAMMATORY ARTHRITIS
Table 5. Predictive values for the development of erosions at 12
months in the 120 RA patients, according to markers present at
disease onset*
RF
HVR3
(HVR3)*
Either RF or HVR3
Both RF and HVR3
Sensitivity
Specificity
(%I
(%I
75
74
26
95
53
67
61
88
39
88
Relative
risk
5.91
4.28
2.61
13.49
8.13
P
<0.001
<O.Ool
NS
<o.ool
<0.001
* Rheumatoid arthritis (RA) patients fulfilled the American College
of Rheumatology criteria for RA. Erosions were defined radiologically. RF = rheumatoid factor; HVR3 = third hypervariable region
conserved epitope; (HVR3)’ = heterozygote or homozygote for
conserved epitope; NS = not significant.
conserved sequence was similar in patients who fulfilled the ACR 1987 criteria for RA and those who did
not. Although the prevalence of HVR3 in the RA
patients was significantly higher than that in healthy
controls, its major effect was on subsequent clinical
manifestations such as erosions.
The 1987 ACR criteria for RA were originally
devised and validated using a patient population with
established disease (14). They have not previously
been assessed in any large-scale study of early disease.
Here, they predicted outcome in patients presenting
with early synovitis, with a sensitivity of 88% and
specificity of 73%. This is clinically important since it
may allow the identification of those patients who will
require followup. However, it is important to determine which patients will develop destructive disease,
so that the use of potentially toxic drug regimens can
be restricted to those patients who actually need them.
The combination of HVR3 with RF identified these
patients, showing an 88% likelihood (specificity) of
destructive disease by 1 year. This ability to predict a
likely outcome on the basis of fixed features, rather
than clinical variables, is important. Patients are now
frequently seen earlier in the disease process, before
the appearance of the standard indicators of poor
prognosis. Thus, stable genetic and serologic markers
could be crucial in the distinction between benign and
serious disease. The presence of either RF or HVR3
had a high sensitivity for predicting erosions, but the
low specificity of 39% limits its usefulness, except for
non-toxic therapy.
The patients who fulfilled the criteria for RA at
presentation had a lower prevalence of the conserved
HVR3 sequence compared with previous studies of
hospital populations with established disease (9,l I).
However, with followup it was clear that patients who
1169
developed a persistent erosive disease showed a much
stronger association. This suggests that the milder
forms of RA are not as strongly associated with the
HVR3, although the disease is not clinically distinguishable from more severe RA at this early stage. It is
of interest that the non-RA patients also had a weak
association with Dw4. The relatively low prevalence
of the conserved sequence in patients with mild RA is
consistent with the results of previous cross-sectional
studies (10-12,23,24). Two possible explanations for
this finding are that the action of the locus occurs after
the initiation of RA or that patients with mild or
self-limiting RA have an entirely separate disease that
is not associated with the MHC. It is not currently
possible to distinguish which of these possibilities is
correct, but the most important disease markers, to
patient and clinician alike, are those associated with
disability.
The MHC association with RA does not appear
to follow a simple model of dominant susceptibility
(25). Patients with severe RA are frequently homozygous for DR4 (26-28). Since Dw4 is the commonest
subtype of DR4, it is a little surprising that several
recent studies have shown an excess of DwWDw14 or
Dw4/Dw15 compound heterozygotes rather than true
Dw4 homozygotes (26,27), although an excess of true
Dw4 homozygotes was found in a third group of
patients (28). In the present study, a disproportionate
representation of Dw4/Dw14 heterozygotes in patients
presenting with RA was observed. All but 1 of the
patients who possessed this pair of alleles developed
erosive RA by 12 months. These patients accounted
for 12% of the group with erosions and are likely to
represent an even larger percentage of the group with
severe disease over the long term; this compares with
0.1% of the general population of the UK (26). Rapid
identification of this group of patients at presentation
with arthritis may be of value, since aggressive therapy is needed at a very early stage in order to attempt
to modulate disease progression. In addition, as a
group of early patients with a high probability of
developing persistent and erosive disease, they merit
detailed biologic study from a pathogenic point of
view. However, the highly specific association between the compound heterozygote genotype Dw4/
Dw14 and RA is not likely to have a major screening
role in routine clinical practice because of its low
sensitivity.
In summary, the findings suggest that the predictive value of oligotyping in combination with determining rheumatoid factor status is of sufficient accu-
1170
GOUGH ET AL
racy to be of clinical use, especially for targeting
patients for early aggressive intervention. This study
also suggests that the major effect of the MHC in RA
occurs after initiation of disease, determining the
persistence and severity of the disease process, which
may have implications for other MHC-related
diseases.
REFERENCES
1. Stastny P Mixed lymphocyte culture typing cells from patients
with rheumatoid arthritis. Tissue Antigens 4571-579, 1974
2. Stastny P: Associations of the B-cell alloantigen DRw4 with
rheumatoid arthritis. N Engl J Med 298:869-871, 1978
3. Wordsworth BP, Lanchbury JSS, Sakkas L1, Welsh KI, Panayi
GS, Bell JI: HLA DR4 subtype frequency in rheumatoid arthritis indicates that DRpl is the major susceptibility locus within
the human leukocyte antigen class two region. Proc Natl Acad
Sci U S A 86:10049-10053, 1989
4. Pile K, Gibson K, Jessop J, Camilleri J, Emberton R, Wordsworth P: HLA-DR4 subtyping by single-strand polymorphism
analysis. Lancet 342:300-301, 1993
5 . Nepom GT, Seyfried C, Holbech S: HLA DR4 associated
disease: oligonucleotide probes identify specific class I1 susceptibility genes in type 1 diabetes and rheumatoid arthritis, Immunobiology of HLA. Volume 2. Edited by B Dupont. New York,
Springer Verlag, 1988
6. Willkens RF, Nepom GT, Marks CR, Nettles JW, Nepom BS:
Association of HLA-Dwl6 with rheumatoid arthritis in Yakima
Indians: further evidence for the “shared epitope” hypothesis.
Arthritis Rheum 34:43-47, 1991
7. Sanchez B, Moreno I, Magarino R, Garzon M, Gonzalez MF,
Garcia A, Nunezroldan A: HLA DRwlO confers the highest
susceptibility to rheumatoid arthritis in a Spanish population.
Tissue Antigens 36:174-176, 1990
8. Gregersen PK, Silver J, Winchester RJ: The shared epitope
hypothesis: an approach to understanding the molecular genetics of susceptibility to rheumatoid arthritis. Arthritis Rheum
30: 1205-1213, 1987
9. Wordsworth BP, Stedeford J, Rosenberg WMC, Bell JI: Limited heterogeneity of the HLA class I1 contribution to susceptibility to rheumatoid arthritis is suggested by positive associations with HLA DR4, DR1 and DRwlO. Br J Rheumatol
30:17&180, 1991
10. De Jongh BM, Van Romunde LKJ, Valkenburg HA, De Lange
GG, Van Rood JJ: Epidemiological study of HLA and GM in
rheumatoid arthritis and related symptoms in an open Dutch
population. Ann Rheum Dis 43:613-619, 1984
11. Van Zeben D, Hazes JMW, Zwinderman AH, Cats A,
Schreuder GMT, D’Amaro J, Breedveld FC: Association of
HLA-DR4 with a more progressive disease course in patients
with rheumatoid arthritis: results of a followup study. Arthritis
Rheum 34:822-830, 1991
12. Salmon M, Wordsworth P, Emery P, Tunn E, Bacon PA, Bell
JI: The association of HLA DRB alleles with self-limiting and
persistent forms of early symmetrical polyarthritis. Br J Rheumatol 32:628-630, 1993
13. Emery P, Salmon M, Bradley H, Wordsworth P, Tunn E, Bacon
PA, Waring R: Genetically determined factors as predictors of
radiological change in patients with early symmetrical arthritis.
Br Med J 305:1387-1389, 1992
4. Arnett FC, Edworthy SM, Bloch DA, McShane DJ, Fries JF,
Cooper NS, Healey LA, Kaplan SR, Liang MH, Luthra HS,
Medsger TA Jr, Mitchell DM, Neustadt DH, Pinals RS, Schaller
JG, Sharp JR, Wilder RL, Hunder GG: The American Rheumatism Association 1987 revised criteria for the classification of
rheumatoid arthritis. Arthritis Rheum 31:315-324, 1988
5 . Winkles JW, Lunec J, Gray L: Automated enhanced latex
agglutination assay for rheumatoid factors in serum. Clin Chem
35:303-307, 1989
16. Wordsworth BP, Allsopp CEM, Young RP, Bell JI: HLA DR
typing using DNA amplification by the polymerase chain reaction and sequential hybridization to sequence specific oligonucleotide probes. Immunogenetics 32:413418, 1990
17. Bidwell JL, Jarrold EA: HLA DR allogenotyping using exon
specific cDNA probes and amplification of rapid minigel methods. Mol Immunol23: I 1 11-1 116, 1986
18. Kullback S: Information Theory and Statistics. New York,
Dover Publications, 1968
19. Woolf B: On estimating the relation between blood group and
disease. Ann Hum Genet 19:251-253, 1955
20. Svejgaard A, Jersild C, Staub Nielsen L, Bodman WF: HL-A
antigens and disease: statistical and genetical considerations.
Tissue Antigens 4:95-105, 1974
21. Yates F: The analysis of contingency tables with groupings
based on quantitative characters. Biomedika 25: 176-181, 1948
22. Hernandez JL, Weir BS: A disequilibrium coefficient approach
to Hardy-Weinberg testing. Biometrics 4553-70, 1989
23. Fisher LR, Kirk A, Awad J, Festenstein H, Alonso A, Perry JD,
Shipley M: HLA antigens in palindromic rheumatism and palindromic onset rheumatoid arthritis. Br J Rheumatol 23:345348, 1986
24. Puttick AJ, Briggs DC, Welsh KI, Vanghan R, Williamson EA,
Boyce M, Jacoby RK, Jones VE: Genes associated with rheumatoid arthritis and mild inflammatory arthritis. I. Major histocompatibility complex class I, I1 and I11 allotypes. Ann Rheum
Dis 49:219-224, 1990
25. Rigby AS, Silman AJ, Voelm L, Gregory JC, Ollier WER, Khan
MA, Nepom GT, Thomson G: Investigating the HLA complement in rheumatoid arthritis: an additional (dominant) mode of
inheritance is rejected, a recessive mode is preferred. Genet
Epidemiol8: 153-175, 1991
26. Wordsworth P, Pile KD, Buckley JD, Lanchbury JSS, Ollier B,
Lathrop M, Bell JI: HLA heterozygosity contributes to susceptibility to rheumatoid arthritis. Am J Hum Genet 51:585-591,
1992
27. Lanchbury JSS, Jaeger EEM, Sansom DM, Hall MA, Wordsworth P, Stedeford J, Bell JI, Panayi GS: Strong primary
selection for the Dw4 subtype of DR4 accounts for the HLADQw7 association with Felty’s syndrome. Hum Immunol 32:
56-64, 1991
28. Weyand CM, Xie CP, Goronzy JJ: Homozygosity for the
HLA-DRB 1 allele selects for extraarticular manifestations in
rheumatoid arthritis. J Clin Invest 89:2033-2039, 1992
Документ
Категория
Без категории
Просмотров
2
Размер файла
522 Кб
Теги
outcomes, can, patients, presentation, arthritis, inflammatory, predict, used, genetics, typing
1/--страниц
Пожаловаться на содержимое документа