Rheumatol Int (1999) 19:47±49
Ó Springer-Verlag 1999
ORIGINAL ARTICLE
A. Bharadwaj á A. Aggarwal á R. Misra
Clinical relevance of IgA rheumatoid factor (RF) in children
with juvenile rheumatoid arthritis
Received: 29 January 1999 / Accepted: 10 August 1999
Abstract This study proposed to investigate the prevalence and clinical relevance of serum immunoglobulin A
(IgA) rheumatoid factor (RF) in juvenile rheumatoid
arthritis (JRA) as published reports vary in their conclusion. Sera of 82 children with JRA and 25-age and
sex-matched healthy children were measured for IgA RF
by an enzyme linked immunoassay using human IgG as
the antigen. Forty-three percent of the disease population were positive and the prevalence in pauciarticular,
polyarticular and systemic onset was 9/18 (50%), 21/47
(44.7%) and 5/17 (27.7%) respectively when mean +
2SD of normal was taken as the cut-o€ value. By de®ning the upper limit of normal as mean + 6SD, 16/47
(34%) were positive in the polyarticular as compared to
2/18 (11.1%) in pauciarticular and 1/17 (5.8%) of systemic onset disease groups. The prevalence in the polyarticular subset with the upper cut-o€ limit was signi®cantly higher than the pauciarticular and the systemic
onset group (P < 0.05). Furthermore, the mean level of
IgA RF was signi®cantly higher in the polyarticular
group compared to the mean level in the systemic onset
group (P < 0.05). The mean level of IgA RF was also
signi®cantly higher (P < 0.05) in 61 children with active
diseases.
Key words Childhood arthritis á Autoantibodies á
Anti IgG
Introduction
The majority of children with juvenile rheumatoid arthritis (JRA) are negative for rheumatoid factor (RF) by
latex agglutination whereas with the use of a sensitive
A. Bharadwaj á A. Aggarwal á R. Misra (&)
Department of Immunology,
Sanjay Gandhi Postgraduate Institute of Medical Sciences,
Lucknow 226014, India
Fax: 91-522-440 017
assay, such as enzyme-linked immunosorbent assay
(ELISA), 50±70% of these patients are positive for IgM
RF [1]. Moreover, by using this technique it is possible to
study IgG RF and IgA RF. This is relevant since IgA RF
has been used as a marker of disease severity in adult
patients with rheumatoid arthritis (RA) [2, 3, 4]. While
the prevalence of IgA RF has been reported to vary from
22 to 58% in patients with JRA [5, 6, 7], the clinical relevance of this ®nding has not been consistent with one
study [6] reporting IgA RF to be speci®cally present in
active polyarticular disease and another [5] in the three
subsets, namely polyarticular onset, pauciarticular onset
and systemic onset.
We have previously reported a signi®cant association
of deforming joint disease with the presence of IgM RF
[8]. Subsequently, we found an association of early onset
pauciarticular disease with IgM rheumatoid factor that
is complexed with IgG or hidden RF in our cohort of
children with JRA [9]. In this study, we have tried to ®nd
the prevalence and clinical relevance of IgA RF in JRA.
Patients and methods
Eighty-two consecutive children ful®lling the American College of
Rheumatology (ACR) criteria [10] of JRA and seen between 1989
and 1997 in the Clinical Immunology Department of old tertiary
care referral hospital were included in the study. Clinical details of
disease duration, morning sti€ness, age, sex, subtypes, fever, rash,
lymphadenopathy, number of active and swollen joints, and limitation of movement were recorded. The disease was categorised
as active if there was presence of systemic symptoms (fever,
rash, lymphadenopathy, hepatosplenomegaly) and/or tender swollen joint with raised erythrocyte sedimentation rate (ESR; >30 mm
fall in 1st hour) and C-reactive protein (CRP; >0.6 mg/dl.). Sera
samples were collected and stored at )40 °C until analysis.
ELISA for IgA RF
Necessary modi®cations were made in the Faith's protocol for
detection of IgM RF (11). Ninety±six well ¯at-bottom microtitre
plates (Nunc) were coated with 100 ll of 10 lg/ml of human IgG
(Sigma, St. Louis, Mo., USA) in 0.05 M sodium carbonate bu€er
48
(pH 9.6) for 2 h at 37 °C followed by overnight incubation at 4 °C.
The plates were washed with phosphate-bu€ered saline (PBS; 0.
15 M, pH 7.2) and blocked with 150 ll of PBS containing 3%
bovine serum albumin (PBS-BSA) for 3 h at 37 °C. After washing
three times with PBS containing 0.05% Tween 20 (PBS-T), 100 ll
of 1:250 diluted serum in PBS±BSA was added to each well and
incubated for 2 h at 37 °C. Following further three washings,
100 ll/well of 1:4000 diluted antihuman IgA HRP conjugate
(DAKO, Denmark) was added and the plate was incubated for 2 h
at 37 °C. The plate was then extensively washed with PBS-T and
developed with 50 ll/well of orthophenyenediamine solution in
citrate phosphate bu€er containing 30% hydrogen peroxide. The
reaction was stopped after 30 min by adding 25 ll/well of 4 N
sulphuric acid. The absorbance was read at 492 nm using an
ELISA reader.
Serum sample from a patient with RA containing IgA RF was
used as standard in each plate. Doubling dilution of this serum
yielded a sigmoid shaped curve. The lowest level of detection, i.e.
where the curve ¯attened out was assigned as 1 arbitrary unit (au/
ml). All the test and control sera were read against this curve. Mean
(1.8 au/ml) + 2SD (1.18), i.e. 4.16 au/ml of 25 control sera was
taken as the cut-o€ limit. For some analyses, sera having a value
higher than the mean + 6SD (8.8 au/ml) were considered as
positive.
Statistical analysis
Student's t-test and the Z-test for proportions were used for intergroup comparisons and the chi-squared test was used to test the
di€erence in prevalence between males and females. The relationship of IgA RF with other quanti®able clinical parameters and IgM
RF was analysed using Pearson's correlation coecient test.
Results
Of the 82 patients, 47 had polyarticular, 18 pauciarticular and 17 systemic onset type of JRA. The mean age of
patients was 13.9 years and the median duration of
disease was 4 years (Table 1). Sixty-one children, 37 of
polyarticular and 12 each of pauciarticular and systemic
onset type had active disease at the time of inclusion in
the study. Seven of the 47 children with polyarticular
disease had classical RF as detected by latex agglutinition.
Thirty ®ve (42.7%) children were positive for IgA
RF. The prevalence in three di€erent types was: polyarticular, 21/47 (44.7%); pauciarticular, 9/18 (50%); and
systemic onset, 5/17 (29.4%). This di€erence was not
statistically signi®cant. The majority of positive sera in
pauciarticular and systemic onset type were marginally
above the cut-o€ limit of mean + 2SD of control
Table 1 Clinical pro®le of juvenile rheumatoid arthritis (JRA)
patients (Poly Polyarticular, Pauci Pauciarticular)
Parameters
Poly
Pauci
Systemic Total
Number
Mean age (years)
Range
Sex ratio (M:F)
Median disease
duration (years)
Range
47
15.3
5±35
21:26
5
18
12.5
6±23
14:4
1
17
11.5
4±18
12:5
3
82
13.9
4±35
47:35
4
0.5±20
0.25±9
0.2±8
0.2±20
Fig. 1 Scatter plot of immunoglobulin A (IgA) rheumatoid factor
(RF) level in the subtypes of juvenile rheumatoid arthritis (JRA). The
cut-o€ limit is shown at two levels, 2SD (solid line) and 6SD (dashed
line). A signi®cant (P < 0.05) number of patients with polyarticular
disease were positive compared to systemic onset and pauciarticular
subset at the higher cut-o€
(Fig. 1). To avoid the blunting e€ect of the low positivity upon any true di€erence that may be there, the cuto€ limit was set at a higher limit of mean + 6SD
(8.8 au/ml). The seroprevalence in the polyarticular
subset 16/47 (34%) was signi®cantly (P < 0.05) higher
than the pauciarticular (2/18; 11.1% ) and the systemic
onset group (1/17; 5.8%). The mean level of the IgA RF
was also signi®cantly (P < 0.05) higher in the polyarticular (9.6 + 11.5 au/ml) compared to the systemic
onset (4.4 + 72 au/ml) but not with the pauciarticular
subtype (5.8 + 7 au/ml).
IgA RF was present more often in females with
polyarticular disease (16/26, 61.5%) than males (5/21,
23.8%; P < 0.05). Patients with active disease had a
higher mean IgA RF (9.15 au/ml) than inactive disease
(4.6 au/ml, P < 0.05). No correlation was seen between
IgA RF and individual clinical and laboratory parameters of disease activity, such as tender joint count,
duration of morning sti€ness, haemoglobin (Hb), ESR
and CRP. IgA RF had good correlation with IgM
RF (r ˆ 0.706, P < 0.05).
Discussion
This study con®rms that the level of IgA RF is higher
in the polyarticular subset and serves to di€erentiate
polyarticular disease from the other subtypes of JRA.
The three subtypes of JRA have di€erent clinical expressions and characteristic pathogenic mechanisms of
each subset. In our continued e€ort to look for serological markers, which are distinctive of these subsets,
we ®rst looked for IgM RF [8] and subsequently hidden
IgM RF [9]. IgM RF identi®ed patients with deforming
disease but its distribution was seen in all the three
subsets of disease. Hidden IgM RF was also seen in all
the three subsets though it was signi®cantly associated
with early onset pauciarticular disease. The prevalence
of 42.7% patients is within the range of previous report
49
of 22±58% [5, 6, 7]. The presence of larger number of
patients with polyarticular disease, which has a higher
prevalence, may have shifted the proportion on the
higher side. In the three previous reports, about onethird of the patients were of polyarticular type while in
our study these constituted 57% of the total patients.
Our results show that a higher proportion of patients
with polyarticular disease are positive for IgA RF when
compared to the other two subsets. This di€erence is
evident only at a higher cut o€ level of mean + 6SD. As
ELISA is a sensitive assay, de®ning the optimum cut-o€
limit is crucial for use in clinical practice. Our observation is in accord with Walker et al. [6] that IgA RF
is positive mainly in polyarticular disease. This is in
contrast to that observed by both Saulsbury [5] and
Ramakrishnan [7], who found IgA RF mainly in the
pauciarticular subset though the di€erence between the
di€erent subsets was not statistically signi®cant. It is
possible that the low positivity of IgA RF in both these
reports and lower number of patients with polyarticular
disease have masked the true association of IgA RF with
polyarticular subsets as observed during the work reported here. As there is no di€erence in assay procedures, it is likely that the disease activity status, the
sex ratio and the optimal level that is taken as positive
are important considerations. The ethnic factor cannot
be ruled out when we compare our data with that of
children with JRA from western countries.
There was a good correlation of IgA RF with activity
of disease but it was not apparent with individual clinical
and laboratory variables of activity such as tender joint,
morning sti€ness or ESR. This suggests that a single
parameter is inadequate to assess disease activity particularly in a disease as heterogeneous as JRA, A previous study [6] of children with polyarticular JRA had
also found a higher prevalence of IgA RF in children
with functional class III/IV compared to class I/II;
however, functional class is a result of activity and end
organ damage and thus may not be a good marker of
disease activity alone.
The level of IgA RF showed a good correlation with
that of IgM RF, similar to an earlier report [6]. Taking
account of our earlier association of IgM RF with
deforming joint disease and the present observation
of association of IgA RF with polyarticular subset, the
simultaneous presence will probably indicate a severe
polyarticular disease. It may be suggested that the same
stimulus is responsible for production of both isotypes
of RA. In contrast, there is evidence for independent
expression at in¯ammatory sites in RA [12]. IgA RF in
the absence of IgM RF has been found in other conditions, such as Henoch Schonlein purpura [13] and IgA
nephropathy [14]. In adult RA, co-occurrence of IgM
RF and IgA RF indicates a poor prognosis. In conclusion, this study con®rms an earlier report that a higher
level of IgA RF is able to distinguish polyarticular disease from systemic and pauciarticular subset of JRA.
De®ning an appropriate cut±o€ level particularly when
a sensitive test is employed is crucial in obtaining this
distinction.
References
1. Lawrence JK, Moore TL, Osborn TG, Nesher G, Madson KL
Kinsella MB (1993) Autoantibody studies in juvenile rheumatoid arthritis. Semin Arthritis Rheum 22:265±274
2. Withrinton RH, Teitsson L, Valdimarsson H, Seifert MH
(1984) Prospective study of early rheumatoid arthritis. Association of rheumatoid factor isotypes with ¯uctuations in disease activity. Ann Rheum Dis 43:679±685
3. Aranson JA, Jonsson TH, Brekkon A, Sigurjonsson K, Valdimarson H (1987) Relation between bone erosion and rheumatoid factor isotypes. Ann Rheum Dis 46:380±384
4. Swedler W, Wallman J, Frolich CJ, Teodorescu M (1997)
Routine measurement of lgM, IgG and IgA rheumatoid factors. High sensitivity, speci®city and predictive value for
rheumatoid arthritis. J Rheumatol 24:1037±1044
5. Salusbury FT (1990) Prevalence of IgM, IgA and IgG rheumatoid factors in juvenile rheumatoid arthritis. Clin Exp
Rheumatol 8:513±517
6. Walker SM, McCurdy DK, Shaham B, Brisk R, Weitting K,
Arora Y, Lehman TJ, Hanson V, Bernstein B (1990) High
prevalence of IgA rheumatoid factor in severe polyarticular
onset juvenile rheumatoid arthritis, but not in systemic onset
or polyarticular disease. Arthritis Rheum 33:199±204
7. Ramakrishnan TP, Howite NT, Wedgwood JF, Hatam L,
Valacer DJ, Bonagura VR (1991) The major rheumatoid factor
cross reactive idiotype and IgA rheumatoid factor in juvenile
rheumatoid arthritis. J Rheumatol 18:1068±1072
8. Aggarwal A, Dabadghao S, Naik S, Misra R (1994) Rheumatoid factor estimation by enzyme linked immunosorbent
assay (ELISA) delineate a subset of deforming joint disease in
juvenile chronic arthritis. Rheumatol Int 14:135±138
9. Dhanda D, Misra R, Naveed M, Pandey CM (1999) Hidden
rheumatoid factor estimation using high performance liquid
chromatography: clinical relevance in Asian Indian children
with juvenile chronic arthritis. Br J Rheumatol, in press
10. Brewer EJ, Bass J, Baum J, Cassidy IT, Fink C, Jacobs J,
Hanson V, Levinson JE, Schaller J, Stillman JS (1977) Current
proposed revision of JRA criteria. Arthritis Rheum 20:195±199
11. Faith A, Pontesilli O, Unger A, Panayi GS, Johns P (1982)
ELISA assays for IgM and IgG rheumatoid factors. J Immunol Methods 55:169±177
12. Koopman WJ, Miller RK, Crago SS, Mestecky J,
Schronhenloher RE (1986) IgA rheumatoid factor: evidence
for independent expression at local sites of tissue in¯ammation. Ann NY Acad Sci 409:258±272
13. Saulsbury FT (1986) IgA rheumatoid factor in Henoch
Schonlein purpura. J Pediatr 108:71±76
14. Czerkinsky C, Koopman WJ, Jackson S, Collins JE, Crago SS,
Schrohenloher RE, Julian BA, Galla JH, Mestecky J (1986)
Circulating immune complexes and immunoglobulin rheumatoid factor in patients with mesangial immunoglobulin A
nephropathy. J Clin Invest 77:1931±1938
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.