Rheumatology Advance Access originally published online on February 1, 2006
Rheumatology 2006 45(3):355-356; doi:10.1093/rheumatology/kei246
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
LETTER TO THE EDITOR |
Prevalence and long-term significance of paraproteinaemia in rheumatoid arthritis
Departments of Rheumatology and 1 Clinical Biochemistry, Aberdeen Royal Infirmary, Aberdeen, 2 John N. Cumings Laboratory, National Hospital for Neurology and Neurosurgery, London and 3 Department of Rheumatology, Raigmore Hospital, Inverness, UK
Correspondence to: M. J. Garton. E-mail: mark.garton{at}tuht.scot.nhs.uk
SIR, Uncontrolled studies [1, 2] suggest monoclonal gammopathy (MG) is more common in rheumatoid arthritis (RA), while oligoclonal gammopathy (OG) is rarely reported. We describe cross-sectional and longitudinal studies assessing the prevalence, clinical features and prognostic significance of MG/OG in RA. We obtained approval from Joint Ethical Committees of the Grampian and Highland Health Boards, and informed written consent from subjects, according to the Declaration of Helsinki.
In 1993–1995, we obtained serum samples from 101 consecutive RA out-patients (26 male, 75 female) and 87 controls (32 male, 55 female) with non-inflammatory rheumatism, requiring orthopaedic surgery. RA patients were mostly seropositive (n = 79) and erosive (n = 81), 64 patients were currently using DMARDs and of these, 8 were on cytotoxic type of DMARDs (and by implication 56 were taking non-cytotoxic DMARDs). The same applies to the prior exposure group, with 16 having been exposed to DMARDs previously, of whom 8 had used cytotoxic DMARDs; physician-rated disease activity was assessed as mild (n = 57), moderate (n = 37) or severe (n = 6), 31 had at least one arthroplasty, and one male patient had previously been diagnosed with MG of uncertain significance (MGUS). Controls were slightly older than RA patients (mean ± S.D., 63.7 ± 14.4 vs 60.1 ± 11.7 yr, P>0.05). Using agarose gel electrophoresis (AGE) [3] and isoelectric focusing (IEF) for immunoglobulin (Ig) G [4], the combined prevalence of MG/OG was higher in RA patients than controls (17 vs 7%; difference 10%, 95% confidence limits 0.9%, 19%; uncorrected 
2 = 4.3, P = 0.038), and usually oligoclonal. MG was detected in 2% (0.2%, 7%) of RA patients using AGE and 3% (0.6%, 8%) using additional IEF; no controls [0% (0%, 4%)] had MG, using either technique. Paraproteinaemic patients tended to have more severe RA (P<0.05) and showed a non-significant excess of erosive disease. We separately studied records of six RA patients and six haematology out-patients (matched for age at diagnosis and disease duration) with MGUS (normal bone marrow, low serum paraprotein concentration; absent immunoparesis, Bence Jones proteinuria or osteolytic lesions). Mean ± S.D. baseline serum paraprotein concentrations were higher among controls than those with RA (15.0 ± 6.63 vs 5.67 ± 1.21 g/l, P = 0.018). Of the six RA-MGUS patients, five had an IgG 
band, and 1 an IgA 
, and RA had preceded paraprotein detection by a median (range) of 11.5 (1–27) yr; all had erosive disease, but two males were persistently seronegative. Pooled analysis of all RA patients (n = 106, as one RA patient appeared in both study groups) confirmed an association between paraproteinaemia and disease severity, and a trend towards an increased prevalence of erosive disease (Table 1). Cytotoxic exposure was uncommon among RA patients with MG (2/8) or OG (1/14).
|
Subjects with screen-detected MG/OG and the 12 established MGUS patients were followed up after a median (range) of 114 (11–139) months and 116 (66–148) months, respectively. No MG/OG subject showed clinical evidence of a plasma cell dyscrasia, but deaths occurred in 4/17 RA patients (two cardiorespiratory, two with disease complications), and 2/6 controls (one cardiorespiratory, one unknown). In the six RA-MGUS patients, paraprotein concentrations gradually increased in three, fell in two (becoming consistently undetectable in one) and remained faint (<5 g/l) in one. One male RA patient (IgA
) developed immunoparesis, but none suffered malignant transformation and overall the mean ± S.D. serum paraprotein concentration at follow-up (6.3 ± 3.9 g/l) was similar to baseline. In MGUS controls, mean serum paraprotein concentrations increased during follow-up to 22.3 ± 12.1 g/l, with three subjects developing overt myeloma and one Waldenström's macroglobulinaemia. MG has been reported in 0.9% of patients with inflammatory joint disease [1] and 1.7% of patients with classical RA and high-titre rheumatoid factor [2]. Interpretation of historical uncontrolled studies is difficult (the prevalence of MG increases more than 10-fold with advancing age [5]), but contemporary comparisons [1, 5] suggest a modest increase in risk. We found a higher absolute risk in our RA patients (partly due to the improved diagnostic sensitivity of IEF [6]), but lacked power to detect significant excess compared with controls. However, the combined prevalence of MG/OG was increased, particularly in those with more severe clinical disease. None of our eight RA patients with MG experienced malignant transformation during follow-up, compared with 7/23 reported by Kelly et al. [7]. This probably reflects important differences in prognostic variables, including baseline serum paraprotein concentrations [8], which may also explain the poor outcome in our MGUS controls. Serum oligoclonal bands often indicate systemic inflammation, but prevalence data are lacking. In RA, they are probably synthesized intrasynovially by a few B-cell clones [9], responding to excessive antigenic stimulation, and perhaps mirroring disease activity. Their antigenic specificity in RA is unknown although Cruz et al. [10] reported oligoclonal antibodies directed against measles epitopes in 9/10 RA patients. Whether OG might disappear during effective therapy of RA in unclear, but MG can obviously diminish or become undetectable on prolonged follow-up.
In conclusion, MG/OG does seem more prevalent in RA, and may indicate more severe disease. The risk of malignant progression in RA patients with MG is most likely determined by baseline paraprotein concentrations, as in the general population.
The authors wish to thank the orthopaedic department at Woodend Hospital, Aberdeen, and Dr Dominic Culligan and Dr Neil Liggett for their assistance with the study.
The authors have declared no conflict of interest.
References
- Dryll A, Ryckewaert A, Kahn MF et al. Rhumatismes inflammatoires et paraproteine a propos de neuf observations personneles. Rev Rheum 1967;34:155–68.
- Abadi I, Korngold L, Freyberg R. Monoclonal immunoglobulins in rheumatoid arthritis. Arthritis Rheum 1968;9:90.
- Jeppsson JO, Laurell CB, Franzen B. Agarose gel electrophoresis: proposed selected method. Clin Chem 1979;25:629–38.
[Free Full Text] - Keir G, Luxton R, Thompson EJ. Isoelectric focussing of cerebrospinal fluid IgG: an annotated update. Ann Clin Biochem 1990;27:436–43.
- Axelsson U, Bachman R, Hallen J. Frequency of pathological proteins (M components) in 6,995 sera from an adult population. Acta Med Scand 1966;179:235–47.[Web of Science][Medline]
- Schipper HI, Bertram G, Kaboth U. Microheterogeneity of paraproteins. I. Diagnostic value of isoelectric focusing followed by immunoblotting. Clin Chim Acta 1988;171:271–8.[Medline]
- Kelly CA, Baird G, Foster H, Hosker H, Griffiths ID. Prognostic significance of paraproteinaemia in rheumatoid arthritis. Ann Rheum Dis 1991;50:290–4.
[Abstract/Free Full Text] - Kyle RA, Therneau TM, Rajkumar SV et al. A long-term study of prognosis in monoclonal gammopathy of undetermined significance. N Engl J Med 2002;346:564–9.
[Abstract/Free Full Text] - Carpenter AB, Huczko E, Eisenbeis CH Jr, Kelly RH. Evidence for locally synthesised and clonally restricted immunoglobulin in the synovial fluid from rheumatoid arthritis patients. Clin Chim Acta 1990;193:1–12.[Medline]
- Cruz M, Hansen K, Ernerudh J, Steere AC, Link H. Lyme arthritis: oligoclonal anti-Borrelia burgdorferi IgG antibodies occur in joint fluid and serum. Scand J Immunol 1991;33:61–71.[CrossRef][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||