Search J Rheum

Advanced Search


Current Issue


Guidelines for Authors

Classified Ads


Search PubMed


Subscriber Registration

Guidelines for Website Users

JRheum Update Service

Contact Info

Tumor Necrosis Factor-a Antagonists Induce Lupus-like Syndrome in Patients with Scleroderma Overlap/Mixed Connective Tissue Disease

To the Editor:

Novel biologic agents that attenuate inflammation by targeting tumor necrosis factor-a (TNF-a) have gained acceptance in the treatment of rheumatoid arthritis (RA) and are being used in a variety of other inflammatory disorders, including Crohn's disease, adult onset Still's disease, uveitis, Wegener's granulomatosis, Beh�et's disease, scleroderma, Sj�gren's syndrome, sarcoidosis, pyoderma gangrenosum, and polymyositis/dermatomyositis1,2.

It now appears that although these therapies may suppress some autoimmune diseases, they may potentiate others. TNF-a clearly plays a deleterious role in the pathogenesis of RA; its role in lupus, however, is controversial. TNF-a blockade, in the form of either a soluble TNF-a receptor such as etanercept, or a monoclonal antibody directed against the TNF receptor such as infliximab, has been associated with the development of autoantibodies, and in some cases, a clinical lupus-like syndrome3,4. Patients treated with these agents for RA can develop subacute cutaneous lupus erythematosus or a drug-induced lupus syndrome5.

We describe 2 cases of patients with mixed connective tissue disease (MCTD) with prominent polyarthritis who developed clinical features consistent with a lupus-like syndrome and positive serological markers of lupus after treatment with a TNF-a antagonist. Discontinuation of the therapy in each case produced clinical remission, resolution of autoantibodies, and normalization of hypocomplementemia (Tables 1 and 2). Rechallenge with infliximab in the second case precipitated the recurrence of the lupus-like syndrome.

Table 1. Serial serologies of Patient 1.

Table 2. Serial serologies of Patient 2.

Case 1. A 44-year-old woman presented with Raynaud's phenomenon (RP), lymphadenopathy, constitutional symptoms, transient chest pain and arthralgias of her shoulders, knees, ankles, and small joints of the hands. Studies revealed a positive antinuclear antibody (1:640), positive anti-RNP, but no anti-dsDNA, anti-Smith, anti-Ro, or anti-La antibodies (Table 1). Despite treatment with hydroxychloroquine (HCQ), she developed significant puffy hands and feet, morning stiffness, joint pain, and mild dysphagia for solids. Her examination disclosed cervical and axillary lymphadenopathy, mild sclerodactyly, and a symmetrical rheumatoid-like polyarthritis. She failed treatment for her arthritis with a combination of a nonsteroidal antiinflammatory drug (NSAID), low dose prednisone, and weekly methotrexate (MTX). Etanercept was added in June 1999, and 9 months later she was able to discontinue prednisone and MTX. HCQ was reinstituted in July 2000 for mild flare with synovitis in the metacarpophalangeal and proximal interphalangeal joints. In May 2001, a febrile illness developed with malaise, progressive anemia, and intense polyarthritis. The etanercept was stopped and then reintroduced in June 2001 after these symptoms resolved. However, new investigations showed anti-double-stranded DNA and anti-Ro and anti-RNP antibodies, whereas anti-Sm and anti-La remained negative. In July 2001, she developed periorbital edema, a malar rash, an erythematous nonpalpable rash on her legs and trunk and persistent, worsening polyarthralgia with active synovitis, including new bilateral knee effusions. A drug-induced lupus syndrome was suspected; the etanercept was discontinued and she was treated with corticosteroids. Her fever, constitutional symptoms, polyarthritis, and rash resolved. However, after one month without etanercept, her polyarthritis returned, requiring a combination of NSAID, prednisone, and mycophenolate mofetil for control. One year after etanercept, the dsDNA antibodies had disappeared, and complements remained normal.

Case 2. A 38-year-old man presented with RP and polyarthralgias. Laboratory data revealed normal complements and a positive antinuclear antibody (1:640), positive anti-RNP, but no anti-dsDNA, anti-Smith, anti-Ro, or anti-La antibodies (Table 2). He was treated with amlodipine, HCQ, and prednisone. He then developed progressive shortness of breath, increasing myalgias and arthralgias, and dysphagia. Physical examination was remarkable for strikingly cyanotic fingers, diffuse lymphadenopathy, puffy fingers with synovitis, and sclerodactyly. He was treated with NSAID, HCQ, and MTX. Despite benefit, the MTX was discontinued after he developed persistent liver enzyme elevation. Active polyarthritis returned and infliximab therapy was instituted. Shortly after the third infliximab infusion, he developed a self-limited episode of flank pain, nausea, vomiting, and general arthralgias and myalgias that lasted about 24 hours. He received his fourth dose of infliximab 2 months later, and again he developed the same diffuse pain, chills, and vomiting. Hypocomplementemia and and new anti-dsDNA antibodies were detected. He was switched to low dose prednisone and leflunomide, with good control. One year after discontinuing infliximab, the anti-dsDNA antibodies had disappeared, and complements normalized (Table 2).

This is the first report of the development of a lupus-like syndrome with specific autoantibodies and hypocomplementemia in MCTD treated with etanercept or infliximab. Although these agents are popular therapeutic options for RA, their use in the treatment of polyarthritis associated with MCTD is not reported. A subset of patients with scleroderma/MCTD will have RA-like polyarthritis requiring aggressive disease-modifying therapy to control the inflammatory process. Both these patients experienced excellent control of their arthritis, but developed an acute illness, in association with administration of a TNF-a antagonist. They both resolved their clinical syndrome, autoantibody production, and hypocomplementemia after discontinuing the drug. This temporal association strongly suggests a causal relationship between the use of drug, the induction of autoantibodies, and the lupus-like syndrome.

TNF-a antagonists can be added to the more than 80 drugs that are thought to cause a lupus-like syndrome. It is of interest that both RA and systemic lupus erythematosus are associated with high serum concentrations of TNF-a. Studies in murine models of lupus provide conflicting explanations about whether TNF-a is pathogenic or protective6,7. In addition, in humans there is a correlation between higher serum levels of the soluble TNF-a receptors and increased lupus disease activity8. It is not clear why the use of a TNF-a antagonist in MCTD would trigger a lupus-like syndrome.

Although case reports have limitations in their interpretation, the temporal association of a lupus-like illness with serologic changes and hypocomplementemia in these 2 cases is compelling evidence of a causative relationship. TNF-a antagonists must be administered with caution if used to treat MCTD-associated polyarthritis.

LISA CHRISTOPHER-STINE, MD, Instructor in Medicine; FREDRICK WIGLEY, MD, Professor of Medicine, Division of Rheumatology, Johns Hopkins University, 1830 East Monument Street, Suite 7500, Baltimore, Maryland 21205, USA. Address reprint requests to Dr. Christopher-Stine.


1. Tutuncu Z, Morgan GJ Jr, Kavanaugh A. Anti-TNF therapy for other inflammatory conditions. Clin Exp Rheumatol 2002;206 Suppl 28:S146-51.

2. Fitzcharles MA, Clayton D, Menard HA. The use of infliximab in academic rheumatology practice: an audit of early clinical
experience. J Rheumatol 2002;29:2525-30.

3. Charles PJ, Smeenk RJ, Delong J, Feldman M, Maini RN. Assessment of autoantibodies to double-stranded DNA induced in rheumatoid arthritis patients following treatment with infliximab, a monoclonal antibody to tumor necrosis factor alpha: findings in open-label and randomized placebo-controlled trials. Arthritis Rheum 2000;43:283-90.

4. Shakoor N, Michalska M, Harris CA, Block JA. Drug-induced
systemic lupus erythematosus associated with etanercept therapy. Lancet 2002;359:579-80.

5. DeBandt MJ, Descamps V, Meyer O. Two cases of
etanercept-induced systemic lupus erythematosus in patients with rheumatoid arthritis [abstract]. Ann Rheum Dis 2001;60:175.

6. Gordon C, Ranges GE, Greenspan JS, Wofsky D. Chronic therapy with recombinant tumor necrosis factor in autoimmune NZB/NFI mice. Clin Immunol Immunopathol 1989;52:41-4.

7. Jacob CO, McDevitt HO. Tumor necrosis factor-alpha in murine autoimmune "lupus" nephritis. Nature 1988;331:356-8.

8. Aderka D, Wysenbeek A, Engelmann H, et al. Correlation between serum levels of soluble tumor necrosis factor receptor and disease activity in systemic lupus erythematosus. Arthritis Rheum 1993;36:1111-20.

Ocular Inflammatory Disease in Patients with Rheumatoid Arthritis Taking Etanercept: Is Discontinuation of Etanercept Necessary?

To the Editor:

Eye involvement is a frequent event in pediatric and adult rheumatologic patients. Tumor necrosis factor (TNF) may play a role in experimental uveo-retinitis. In scleritis, TNF and interleukin 1 released by the local inflammatory cell infiltrate may be associated with sclera destruction1,2. Treatment with anti-TNF, therefore, is a logical approach to this problem. Indeed, such treatment has been reported to be successful in intractable cases of juvenile rheumatoid arthritis (JRA) associated uveitis, seronegative spondyloarthropathies, rheumatoid scleritis, and Beh�et's disease3-6.

There is one report, however, of anti-TNF treatment precipitating inflammatory eye disease5. Clinicians may face the dilemma of whether discontinuation of etanercept is necessary. We describe 2 cases of anterior uveitis and severe scleritis occurring in patients undergoing chronic etanercept treatment. In our patients, local and systemic immunosuppressive

treatment were curative without discontinuation of etanercept. The patients remained asymptomatic for one and 2 years, respectively.

The first patient was a 48-year-old white woman with seropositive RA and Sj�gren's syndrome of 6 years' duration. She had been taking etanercept monotherapy (25 mg twice weekly) for 3 years, with remission of the arthritis. There were no extraarticular manifestations other than sicca syndrome. Eighteen months after the start of etanercept treatment, and while the disease remained in remission, she developed left eye scleritis, necessitating local and systemic steroids, local cyclosporine, and azathioprine treatment. HLA-B27 was absent and erythrocyte sedimentation rate (ESR) was 28 mm/h. There was resolution and nonrecurrence even after discontinuation of azathioprine. Etanercept was continued throughout the course of the illness.

The second case was a 58-year-old white woman with RA of 20 years' duration. She had been taking methotrexate (MTX) for 10 years, in combination with etanercept (25 mg twice weekly) for 3 years. Her RA was erosive, seropositive, and nodular with significant hand and feet deformities, but with no tender joints. She had no other extraarticular manifestations. Twenty months after the start of etanercept treatment, she developed severe anterior uveitis necessitating local and systemic steroids. HLA-B27 was absent, ESR was 30 mm/h, and there was no increase of synovitis or number of tender joints. There was resolution and nonrecurrence. Etanercept was continued throughout the course of the illness.

Our cases indicate that etanercept treatment, in the standard dosages, was selectively effective in improving the arthritis, but could not prevent ocular involvement. Perhaps the pathophysiologic process of ocular involvement is different, and TNF-a may not be the predominant responsible cytokine. It is of interest that infliximab rather than etanercept was found to be a more consistently beneficial treatment for patients with JRA with MTX-resistant uveitis and HLA-B27 associated chronic uveitis4,6. This may possibly indicate a different cytolytic mechanism. The data, however, are not sufficient to determine the superiority of infliximab over etanercept in the treatment of uveitis.

We believe that discontinuation of etanercept is not necessary in the management of ocular rheumatoid involvement in patients receiving etanercept.

ATHAN N. TILIAKOS, MSIII, University Hospital Health System - Richmond Heights Hospital, Richmond Heights, Ohio; NICHOLAS A. TILIAKOS, MD, FACR, 705 Walther Road, Lawrenceville, Georgia 30045, USA.


1. Dick AD, Duncan L, Hale G, et al. Neutralizing TNF-alpha activity modulates T-cell phenotype and function in experimental autoimmune uveoretinitis. J Autoimmunity 1998;11:255-64.

2. Wakefield D, Lloyd A. The role of cytokines in the pathogenesis of inflammatory eye disease. Cytokine 1992;4:1-5.

3. Reiff A, Takei S, Sadeghi S, et al. Etanercept therapy in children with treatment-resistant uveitis. Arthritis Rheum 2001;44:1411-5.

4. El-Shabrawi Y, Herman J. Case series of selective anti-tumor necrosis factor alpha therapy using infliximab in patients with nonresponsive chronic HLA-B27 associated anterior uveitis: comment on the articles by Brandt et al. Arthritis Rheum 2002;46:2821-2.

5. Smith JR, Levinson RD, Holland GN, et al. Differential efficacy of tumor necrosis factor inhibition in the management of inflammatory eye disease and associated rheumatic disease. Arthritis Rheum 2001;45:252-7.

6. Sfikakis PP, Theodossiadis PG, Katsiari CG, Kaklamanis P, Markomichelakis NN. Effect of infliximab on sight-threatening panuveitis in Behcet's disease [letter]. Lancet 2001;358:295-6.

Effect of Weather Exposure on Rheumatoid Arthritis

To the Editor:

Patients with arthritis often complain that certain weather conditions, in particular cold and dampness, aggravate their disease symptoms. However, studies on influences of weather on arthritis patients have reported conflicting results. In all studies, the extent to which the participating patients were actually exposed to the weather was not taken into account. This is a serious shortcoming since, except for barometric pressure, considerable differences will often exist between indoor and outdoor conditions. For example, outdoor frost will have little effect on patients staying indoors all day.

As I am myself a patient with rheumatoid arthritis (RA) [since 1979, male, 56 years old, fulfilling the American College of Rheumatology criteria1, positive rheumatoid factor (RF)] and live in the marine climate of The Netherlands, I looked for effects of the time I daily spent outdoors on subjective as well as objective variables of my RA. The beneficial effect of the time spent outdoors on joint pain score and erythrocyte sedimentation rate (ESR) has been reported2.

Over a period of 4 years (1999-2002), I daily recorded the time I spent outdoors (24 days excepted) and calculated the monthly means (Figure 1A). In addition to being outdoors for longer times, I undertook other measures in attempts to lower the 24 h ambient temperature and humidity: not wearing pyjamas3, keeping the bedroom window open3, not wearing a coat outdoors at temperatures above 0�C, and sleeping on a ventilated spring mattress. These measures were all in effect during the study. Blood samples were taken regularly (n = 38, on Wednesdays between 11:00 AM and 12:00 noon) for determination of RF (latex-nephelometry, Figure 1B) and ESR as a measure of disease activity.


Figure 1. Monthly mean 24 h time outdoors (A), rheumatoid factor (B), and erythrocyte sedimentation rate (C) from January 1, 1999, to December 31, 2002, in a patient with RA. Shading indicates periods with monthly mean 24 h time outdoors = 2.5 h. Error bars in A indicate SD. Horizontal bars in C indicate medication: azathioprine 125 mg daily (aza), MTX 7.5 mg, later 10 mg, weekly.

Based on the level of the monthly mean time spent outdoors, the study was subdivided into long outdoor periods (>= 2.5 h, shaded area in Figure 1) and short outdoor periods (< 2.5 h, unshaded). Although at a clinically low level, the RF virtually always decreased during the long periods outdoors, and increased during the short periods (Figure 1). The percentage of change in RF in the interval between 2 successive determinations was negatively correlated with the mean 24 h time spent outdoors during the interval (r = -0.44, p < 0.01).

During the second half of the study (from January 1, 2001) the RF as well as ESR were negatively correlated with the mean 24 h time spent outdoors in the 4 weeks preceding each determination. The data from 1999 and 2000 were not used for these calculations because the medication was changed twice (horizontal bars in Figure 1C), with subsequent lowering of the ESR. For RF as well as the ESR, the negative correlation was maximal when the data were shifted back in time by 51 days with respect to the time-outdoors data (r = -0.57, p < 0.01 for the correlation with RF and with ESR). This indicates that an increase in weather exposure is followed 51 days later on average by a decrease in RF and ESR and vice versa. The similarity in the correlation of the time spent outdoors with the RF and ESR supports the conclusion of a recent study that RF concentrations reflect disease activity in RA4.

In general, outdoor activities were light (sitting and cycling, in the estimated ratio 1:1). Due to airy clothing with subsequent increased ventilation and cooling of the microclimate close to the skin, being outdoors was often a chilly experience. This observation shows that the beneficial effect of being outdoors is in agreement with the previously reported positive correlation between the meteorological temperature/humidity complex and pain in 88 patients with RA in The Netherlands5. The correlation of the RF as well as ESR with variations in weather exposure objectively illustrates linkage between weather and disease activity. Moreover, being outdoors seems to be a way to reduce the disease activity of my RA. The underlying mechanism is unclear. As described2, the effect might be related to skin vasoconstriction, or to increased production of cortisol.

For future investigations on weather effects on RA (or other disease), I recommend that the daily duration of weather exposure be taken into account. The results in this study justify similar research on a larger scale and in different climates. If the effects I found on myself can be reproduced in other patients with RA, being outdoors for longer times might be a valuable addition to the treatment of RA.

WIEBE R. PATBERG, MSc, Department of Medical Physiology, University of Groningen, PO Box 196, 9700 AD Groningen, The Netherlands


I thank Prof. H.W.G.M. Boddeke and Prof. J.J. Rasker for their helpful comments.


1. Arnett FC, Edworthy SM, Bloch DA, et al. The American Rheumatism Association 1987 revised criteria for the classification of rheumatoid arthritis. Arthritis Rheum 1988;31:315-24.

2. Patberg WR, Rasker JJ. Beneficial effect of being outdoors in rheumatoid arthritis [letter]. J Rheumatol 2002;29:202-4 (correction appears in J Rheumatol 2002;29:864).

3. Patberg WR. Effect of weather on daily pain score in rheumatoid arthritis [letter]. Lancet 1987;8555:386-7.

4. Knijff E, Drossaers W, Verhoeven A, et al. Rheumatoid factor measured by fluoroimmunoassay: a responsive measure of rheumatoid arthritis disease activity that is associated with joint damage. Ann Rheum Dis 2002;61:603-7.

5. Patberg WR, Nienhuis RLF, Veringa F. Relation between meteorological factors and pain in rheumatoid arthritis in a marine climate. J Rheumatol 1985;12:711-5.

Polymyalgia Rheumatica and Biopsy-Proven Giant Cell Arteritis Exhibit Different HLA-DRB1* Associations

To the Editor:

Jacobsen et al have confirmed the reported association between HLA-DRB1*04 alleles and biopsy-proven giant cell arteritis (GCA). Moreover, they also found a lack of significant association between isolated polymyalgia rheumatica (PMR) and HLA-DRB1*04 alleles in the Scandinavian population1.

Perhaps these authors were not aware of our observations reported 4 years ago2. In the Lugo region of Northwest Spain, temporal artery biopsy with histological examination is a general procedure for patients presenting with GCA manifestations. In addition, longterm followup of patients diagnosed with isolated PMR, without any clinical feature of GCA, is generally performed. This allows a rigorous distinction between biopsy-proven GCA, associated or not with PMR manifestations, and isolated PMR. In our series of 86 isolated patients with PMR this condition was associated with HLA-DRB1*13/14 (previously DR6), although HLA-DRB1*0401 and *0404 alleles were marginally increased in frequency compared with controls2. Unlike isolated PMR, in our series of 62 cases of biopsy-proven GCA an association with HLA-DRB1*04 alleles, in particular with *0401 and *0404, was observed2. This association between GCA and HLA-DRB1*04 alleles was independent of the presence or absence of associated PMR. Thus, our findings suggest that most of these associations with HLA-DRB1*04 alleles are primary with biopsy-proven GCA, regardless of co-occurrence of PMR. As with isolated PMR, the associations of HLA with GCA in Northwest Spain were primarily with HLA-DRB1 and not HLA-DQA1 or HLA-DQB1 alleles2.

Further studies of our group have underlined the role of HLA-DRB1*04 alleles as markers of disease severity in biopsy-proven GCA and PMR3,4. Indeed, patients with biopsy-proven GCA carrying HLA-DRB1*04 alleles exhibited a significantly increased risk of ischemic visual manifestations3. Also, relapses of PMR in isolated PMR patients were more commonly observed in those carrying HLA-DRB1*0401 alleles, in particular *04014. However, both GCA and PMR are polygenic diseases where other genes may play a role in both clinical expression and disease severity5,6. Due to this, besides a higher number of patients with biopsy-proven GCA and isolated PMR tested in our studies, the different expression of other genes may explain, at least in part, the differences in terms of disease severity observed between our series and the one of Jacobsen, et al1.

MIGUEL A. GONZALEZ-GAY, MD, PhD; CARLOS GARCIA-PORRUA, MD, PhD,. Hospital Xeral-Calde, Lugo, Spain; WILLIAM E.R. OLLIER, PhD, Centre for Integrated Genomic Research, University of Manchester, Manchester, UK.


1. Jacobsen S, Baslund B, Madsen HO, Tvede N, Svejgaard A, Garred P. Mannose-binding lectin variant alleles and HLA-DR4 alleles are associated with giant cell arteritis. J Rheumatol 2002;29:2148-53.

2. Dababneh A, Gonzalez-Gay MA, Garcia-Porrua C, Hajeer A, Thomson W, Ollier W. Giant cell arteritis and polymyalgia rheumatica can be differentiated by distinct patterns of HLA class II association. J Rheumatol 1998;25:2140-5.

3. Gonzalez-Gay MA, Garcia-Porrua C, Llorca J, et al. Visual manifestations of giant cell arteritis. Trends and clinical spectrum in 161 patients. Medicine (Baltimore) 2000;79:283-92.

4. Gonzalez-Gay MA, Garcia-Porrua C, Vazquez-Caruncho M, Dababneh A, Hajeer A, Ollier WE. The spectrum of polymyalgia rheumatica in northwestern Spain: incidence and analysis of variables associated with relapse in a 10 year study. J Rheumatol 1999;26:1326-32.

5. Gonzalez-Gay MA. Genetic epidemiology. Giant cell arteritis and polymyalgia rheumatica. Arthritis Res 2001;3:154-7.

6. Gonzalez-Gay MA, Hajeer AH, Dababneh A, et al . IL-6 promoter polymorphism and phenotypic expression of polymyalgia rheumatica in biopsy-proven giant cell arteritis. Clin Exp Rheumatol 2002;20:179-84.

Biological and Clinical Markers of Disease Activity in Ankylosing Spondylitis

To the Editor:

The inflammatory response involved in ankylosing spondylitis (AS) involves special features1 that distinguish it from other inflammatory rheumatic diseases: preferential location in the enthesis2 (which undergoes ossification) and sacroiliac joints, a prominent role of microorganisms in reactive arthritis, clear genetic predisposition regarding HLA-B27, and exquisite sensitivity to nonsteroidal antiinflammatory drugs (NSAID). In patients with AS, clinical assessment of disease activity is difficult3 due, in part, to the relative inaccessibility of the body areas usually involved (i.e., the enthesis and other rachidial spinal structures) and overall because the correlation between the most frequently used serum markers, nonspecific biological indicators of inflammation4, and the activity of the disease are controversial (except in those clinical forms with peripheral involvement). While this has been widely documented5, no new biological measures allowing one to accurately discriminate the clinical activity of AS have been reported. Indeed, although some of these various inflammatory mediators may always be present, high concentrations of inflammation mediators are present in only 34-64% of patients with severe disease6.

We investigated whether biochemical variables are related to clinical features of disease activity in AS. Forty-two consecutive outpatients fulfilling the modified New York criteria for AS7 were recruited. A single researcher examined all patients. Disease activity was assessed using visual analog scales (VAS) for night pain (NP), axial pain (AP) as rachialgia that was relieved with motion and increased by resting, and pain/swelling in peripheral joints (PP) during the week prior to the clinical visit. In addition, both the physician and the patients judged disease activity separately. Thus, the disease was considered to be active with a VAS measure = 40 mm and inactive otherwise. AP and PP are 2 relevant components of the Bath AS Disease Activity Index (BASDAI)8,9.

On the same day, blood samples were collected for the determination of acute phase reactants: erythrocyte sedimentation rate (ESR), platelets, hemoglobin (Hb), serum concentrations of C-reactive protein (CRP), alpha-acid glycoprotein (AAG), and albumin (ALB).

We studied 42 patients (32 men/10 women) with mean age 40 � 12 years and disease duration 16 � 11 years. Thirty-six patients were HLA-B27 positive. All had bilateral sacroiliitis. All had axial involvement, either isolated (n = 12) or associated with peripheral involvement (n = 30). In spite of this, we analyzed the 2 groups (axial/mixed) separately, and the differences obtained were not statistically significant.

Comparisons of means between the active and inactive groups are shown in Table 1. Fifteen patients had AP; however, none of their biochemical indicators was significantly different from the patients without AP. Twenty patients had PP/swelling; their mean Hb levels were lower and their mean ESR, platelets, CRP, and AAG levels were significantly higher than in the other patients. Fifteen patients complained of NP; in these, only the mean AAG level (127.4 � 27.9 mg/dl) was greater (p = 0.006) than in those without NP (98.6 � 24.7 mg/dl). Thirty patients judged that their disease was active at the time of the consultation; the mean AAG level (114.3 � 32.3 mg/dl) was significantly (p = 0.048) higher than in the other 12 individuals (95.3 � 11.7 mg/dl). After examining the patients, and in absence of information about their biochemical measures, the physician judged there was active disease in 21 patients; their mean Hb levels were lower and the mean levels of ESR, platelets, CRP, and AAG were significantly higher than in those with inactive disease.

Table 1. Comparisons of means between active and inactive disease groups.

Spearman's rho correlation coefficients between the clinical and biochemical disease activity measures are shown in Table 2.

Table 2. Correlations between clinical and biochemical disease activity measures.

The inflammatory clinical variables we evaluated (NP, inflammatory spinal pain, and peripheral arthritis) are all considered essential in assessment of clinical activity status in AS9,10. Presently, the BASDAI is a good instrument to measure disease activity in AS, but when we began our study it was not frequently used in our clinical practice; for this reason we use VAS to measure certain clinical disease activity variables. We combined the BASDAI components with additional information of special interest in clinical practice, i.e., NP and patient and physician global assessment. Specifically, NP is a good indicator of inflammatory status, as are morning stiffness and CRP serum levels11; in addition, NP, inflammatory spinal pain, peripheral arthritis, and the patient's global assessment are some of the specific indicators used for clinical record-keeping by the Assessments in AS Working Group12.

In our analysis, we assumed biological markers to be objective indicators, and each clinical symptom of VAS a subjective indicator of disease activity. From our results it follows that the presence of pain (particularly NP or PP) and patient global assessment and physician assessment of the activity status are correlated with the mean increase in AAG levels. Dougados, et al11 also consider NP to be an accurate indicator of inflammation. According to Taylor, et al13, NP is positively correlated with sacroiliac sclerosis, but negatively with ankylosis. On the other hand, we found assessment of activity status by the physician and PP to be the clinical variable most strongly correlated with the objective biological elements. Our results reveal that PP (a component of BASDAI) is associated with the biochemical markers of inflammation in patients with AS. Similarly, subjective assessments of disease by a physician correlate well with biological markers of disease activity.

The acute phase reactants most frequently used as indicators of inflammatory activity in AS are CRP and ESR. Some investigators have reported increased ESR and CRP values in patients with active AS and markedly increased levels in cases of peripheral arthritis relative to axial forms14. By contrast, other authors have found increased ESR and CRP values only in peripheral forms of AS15,16. Our results reveal that alpha-acid glycoprotein concentration might be an additional, very useful disease activity marker because of its good correlation with NP, PP, physician assessment, ESR values, and CRP serum levels.

MAR�A CARMEN MU�OZ-VILLANUEVA, MD, Researcher and Clinical Methodologist; ELISA MU�OZ-GOMARIZ, BD, Researcher; ALEJANDRO ESCUDERO-CONTRERAS, MD, Rheumatologist; VER�NICA P�REZ-GUIJO, MD, Rheumatologist; EDUARDO
COLLANTES-EST�VEZ, MD, Chief of Rheumatology Service, Servicio de Reumatolog�a, Hospital Universitario Reina Sof�a, Avda. Men�ndez Pidal, s/n 14004, C�rdoba, Spain.

This work was supported by a grant-in-aid from the Health Council of Andalusia Board (SAS 157/99).


1. Amor B. Inflammation in spondylarthropathies. Rev Rhum Mal Osteoartic 1989;56:8-11.

2. McGonagle D, Khan MA, Marzo-Ortega H, O'Connor P, Gibbon W, Emery P. Enthesis in spondylarthropathy. Curr Opin Rheumatol 1999;11:244-50.

3. Van der Linden SJ, van der Heijde DM. Ankylosing spondylitis. Rheum Dis Clin North Am 1998;24:663-77.

4. McGonagle D, Gibbon W, Emery P. Classification of inflammatory arthritis by enthesitis. Lancet 1998;352:1137-40.

5. Volanakis JE. Acute-phase proteins in rheumatic diseases. In: Koopman WJ, editor. Arthritis and allied conditions. 13th ed. Baltimore: Williams and Wilkins; 1997:505-15.

6. Sheehan NJ, Slavin BM, Donovan MP, Mount JN, Mathews JA. Lack of correlation between clinical disease activity and
erythrocyte sedimentation rate, acute phase proteins or protease inhibitors in ankylosing spondylitis. Br J Rheumatol

7. Van der Linden SJ, Valkenburg HA, Cats A. Evaluation of
diagnostic criteria for ankylosing spondylitis. A proposal for
modification of the New York criteria. Arthritis Rheum 1984;27:361-8.

8. Garret SL, Jenkinson TR, Whitelock HC, Kennedy LG, Gainsford G, Calin A. A new approach to defining disease status in
ankylosing spondylitis. The Bath Ankylosing Spondylitis Disease Activity Index (BASDAI). J Rheumatol 1994;21:2286-91.

9. Calin A, Nakache JP, Gueguen A, Zeidler H, Mielants H, Dougados M. Defining disease activity in ankylosing spondylitis: is a
combination of variables (Bath Ankylosing Spondylitis Disease Activity Index) an appropriate instrument? Rheumatology Oxford 1999;38:878-82.

10. Dougados M. L'int�r�t de la mesure en pratique quotidienne. Rev Rhum Mal Osteoartic 1998;65:3-4.

11. Dougados M, Gueguen A, Nakache JP, et al. Clinical relevance of C-reactive protein in axial involvement of ankylosing spondylitis.
J Rheumatol 1999;26:971-4.

12. Van der Heijde D, Calin A, Dougados M, Khan MA, van der Linden SJ, Bellamy N, ASAS Working Group. Selection of
instruments in the core set for DC-ART, SMARD, physical therapy, and clinical record keeping in ankylosing spondylitis. Progress Report of the ASAS Working Group. J Rheumatol 1999;26:951-4.

13. Taylor HG, Wardle T, Beswick EJ, Dawes PT. The relationship of clinical and laboratory measurements to radiological change in ankylosing spondylitis. Br J Rheumatol 1991;30:330-5.

14. Smith GW, James V, Mackenzie DA, et al. Ankylosing spondylitis and secretor status: a re-evaluation. Br J Rheumatol

15. Spoorenberg A, van der Heijde D, de Klerk E, et al. Relative value of erythrocyte sedimentation rate and C-reactive protein in
assessments of disease activity in ankylosing spondylitis.
J Rheumatol 1999;26:980-4.

16. Kerr HE, Sturrock RD. Clinical aspects, outcome assessment,
disease course, and extra-articular features of
spondyloarthropathies. Curr Opin Rheumatol 1999;11:235-23.

Ceiling Effects and the Schober Test

To the Editor:

In patients with ankylosing spondylitis (AS), assessments of restrictions in lumbar range of motion most often measure lumbar flexion, using the Schober test or one of its modifications. Lumbar extension is rarely measured, despite evidence indicating that lumbar extension is restricted earlier in the course of AS than lumbar flexion1,2. If extension is restricted before flexion, some patients who have a normal Schober test may have limited spinal extension. These patients may have little ability to demonstrate further improvement in lumbar flexion, but would be able to demonstrate improvement in lumbar range of motion if extension was measured. The inability of a measure to indicate further improvement is known as a ceiling effect3. We examined the extent to which measurement of lumbar flexibility was subject to a ceiling effect because these measures included only the Schober test and not measurement of lumbar extension. To do so, we determined the prevalence of restrictions in lumbar extension among patients with AS who had a normal or mildly restricted Schober test.

We measured the Schober test of lumbar flexion and Smythe's test of lumbar extension in 230 patients with AS on their enrollment in a prospective cohort study4,5. The Schober test was measured as the increase with forward flexion of a 10-centimeter segment marked on the patient's back with the inferior mark at the level of the posterior superior iliac spines6. Smythe's test was recorded as the contraction of three 10-cm segments marked on the patient's back, with the inferior mark at the level of the posterior superior iliac spines, measured while the patient pushed up from a prone position and arched his back7. For this analysis, we used the contraction of the lowest 10-cm segment. The mean of 2 attempts of each maneuver was recorded. One rheumatologist (SK) performed 85% of the examinations, and the remainder were performed by a second examiner (MMW). Twenty patients were examined by both examiners, and intraclass correlations (> 0.80) showed good interobserver agreement.

Of the 230 patients, 22 patients (9.5%; mean age 43.2 � 11.7 yrs; mean duration of AS 16.1 � 11.3 yrs; 73% men) had a normal Schober test with an increase of 5.0 cm or more, and 43 patients (18.7%; mean age 42.4 � 12.6 yrs; mean duration of AS 12.3 � 11.4 yrs; 70% men) had a Schober test of 4.0-4.9 cm (Table 1). Among 22 patients with a normal Schober test, 2 patients (9%) had markedly limited spinal extension (< 1.5 cm contraction), while an additional 3 patients (13%) had limited extension of lesser degrees. Among 43 patients with a mildly restricted Schober test, 3 (7%) had marked limited spinal extension, and another 25 patients (58%) had limited extension of lesser degrees.

Table 1. Contraction with lumbar extension of the lowest 10-cm segment of Smythe's test in patients with AS who had a normal Schober test (increase of 5.0 cm) or mildly restricted Schober test (increase of 4.0-4.9 cm).

We chose to focus our analysis on contraction of the lower segment of the Smythe test, as this region corresponds with that measured in the Schober test and because it has a greater ability to indicate contraction than the middle or upper segments of the Smythe test. However, extension was also limited in the middle and upper segments of the Smythe test, with contraction of less than 1.5 cm in the upper segment in 5 of the 22 patients with a normal Schober test, and contraction of less than 1.5 cm in the middle segment in 2 of the 22 patients with a normal Schober test.

Our findings support studies that indicate that lumbar extension can be limited in patients with AS who have normal lumbar flexion. This distinction may be present in more than 20% of patients who have a normal Schober test, and in more than one-half of patients with a mildly limited Schober test. In these patients, the Schober test may not be able detect much further improvement in lumbar flexibility, but improvement could potentially be detected by measuring changes in lumbar extension. In intervention studies, reliance on the Schober test alone may underestimate the effect of treatment in these patients, while measuring both lumbar flexion and extension would provide greater opportunity to detect improvement in lumbar flexibility.

MICHAEL M. WARD, MD, MPH, Intramural Research Program, National Institute of Arthritis and Musculoskeletal and Skin Diseases, National Institutes of Health, US Department of Health and Human Services, Building 10, Room 9S205, 10 Center Drive MSC 1828, Bethesda, Maryland; E-mail:; SUSANA KUZIS, MD, Camino Medical Group, Sunnyvale, California, USA.


1. Dudley Hart F, Strickland D, Cliffe P. Measurement of spinal mobility. Ann Rheum Dis 1974;33:136-9.

2. Gran JT, Husby G, Hordvik M, Stormer J, Romberg-Andersen O. Radiological changes in men and women with ankylosing
spondylitis. Ann Rheum Dis 1984;43:570-5.

3. Streiner DL, Norman GR. Health measurement scales. A practical guide to their development and use. 2nd ed. New York: Oxford University Press; 1995.

4. Ward MM, Kuzis S. Validity and sensitivity to change of
spondylitis-specific measures of functional disability. J Rheumatol 1999;26:121-7.

5. Ward MM. Functional disability predicts total costs in patients with ankylosing spondylitis. Arthritis Rheum 2002;46:223-31.

6. ARA Glossary Committee. Dictionary of the rheumatic diseases. Vol. I: Signs and symptoms. New York: Contact Press; 1982.

7. Miller MH, Lee P, Smythe HA, Goldsmith CH. Measurement of spinal mobility in the sagittal plane: new skin contraction method compared with established methods. J Rheumatol 1984;11:507-11.

Fibromyalgia and Pramipexole: Promise and Precaution

To the Editor:

Despite 13 years of research after the establishment of criteria1, fibromyalgia syndrome (FM) remains difficult to understand and even more difficult to treat effectively. However, 2 articles provide important insight into pathogenesis and lead to a new treatment paradigm involving pramipexole.

First, Moldofsky and colleagues describe FM symptoms in healthy controls after an auditory arousal from stage IV sleep over 4 nights2. Second, Yunus and Aldag identified restless legs syndrome (RLS) disproportionately in subjects with FM versus healthy controls (31 % vs 2%, respectively)3. Therefore, FM could be the predictable consequence of prolonged and intense stage IV sleep deprivation, but reduction of the arousal (perhaps RLS as one example) was more important than induction of stage IV sleep with antidepressants.

Lorazepam and clonazepam (2 mg qhs) reduce RLS4. In a retrospective chart review, they also reduced FM tenderness scores for 82% of 202 patients at 2 weeks5 and for 62% of 174 patients at one year6 when added to nighttime antidepressants. By intention-to-treat (ITT) analysis at one year, 54% had =  50% decrease in tenderness score using either clonazepam or lorazepam 2 mg qhs.

Pramipexole is a second-generation dopamine 2 (D2) receptor agonist, approved by the US Food and Drug Administration in 1997 for Parkinson's disease, that is remarkably effective treatment for RLS7. It has strong affinity for the dopamine 3 receptor subtype in the D2 family, mild affinity for the central alpha-2 adrenoreceptor (target of clonidine), and no affinity for other dopamine, benzodiazepine, norepinepherine, or serotonin receptors.

A retrospective chart review of pramipexole for 166 patients with FM revealed encouraging results8. Patients added pramipexole to their best regimen to date and increased by 0.125-0.25 mg weekly, similar to its use for RLS. For those who continued pramipexole for more than 7 days (n = 127), the tenderness score decreased 54% at a mean dose of 1.55 mg qhs for 2-12 months (mean 4 mo). Inefficacy correlated with seeing a psychiatrist (p < 0.05, chi-square test), but not with age, pretreatment tenderness score, or disability. Twenty-three percent quickly discontinued for non-serious intolerances, usually nausea or anxiety. There were no serious adverse events even with doses up to 6.0 mg qhs. By ITT analysis (n = 166), 58% achieved =  50% decrease in tenderness score, even though 22 (13%) discontinued pramipexole before they could be evaluated. Other measures of FM activity were not collected. Further, for 19 patients unresponsive or intolerant to pramipexole, ITT analysis showed that 74% achieved = 50% decreased tenderness score after adding the other known dopamine 3-specific agonist, ropinirole, for a mean of 4 months9. Unfortunately, 13 of 19 also eventually discontinued for non-serious intolerances, especially nausea.

Since then, a strategy to decrease gastrointestinal intolerance has been very helpful. Of 89 consecutive patients given pramipexole scheduled to increase by 0.25 mg weekly to 2.0 mg qhs, 57 (75%) noted some nausea when interviewed at 3 weeks (0.75 mg dose). However, 39 of 57 continued treatment and controlled nausea by using one of 4 branded proton pump inhibitors (PPI) qhs at supratherapeutic doses (3 tablets). Of 16 who discontinued pramipexole, 8 refused the PPI and 6 took < 3 tablets qhs. Diarrhea and other PPI intolerances were infrequent. At 8 weeks (pramipexole 2.0 mg qhs), PPI use was not different from pretreatment use (16% vs 15%, respectively). Individual PPI success was unpredictable, ranging from 33% to 54%.

Finally, much higher pramipexole doses have been prescribed, as the most therapeutic dosage for FM appears to be 4.5 mg qhs achieved over 12 weeks. With careful monitoring, doses have been increased up to 10.5 mg qhs in select, previously injured, narcotic-dependent patients. A retrospective chart review of usual clinical practice for consecutive patients taking = 2.25 mg qhs (2.25-10.5 mg qhs) revealed 195 patients taking a mean dose of 4.2 mg for 7.6 months (range 2-25 mo). Patients discontinued because of cost (0.5%), inefficacy (5%), and intolerance (4%), including nausea (1.6%), daytime somnolence (1.1%), anxiety (0.8%), headache (0.2%), and dizziness (0.2%). Certainly, more patients had intolerances, but discontinuation rates were unexpectedly low, possibly because the patients were a select group (already tolerating = 2.25 mg qhs) or because the efficacy of pramipexole outweighed the degree of intolerance.

In summary, pramipexole may become an important option for patients with FM, but further studies are required to confirm these results and improve its application. Nevertheless, manipulating dopamine-related central nervous system regulatory mechanisms must be gradual, and initial nausea prevented whenever possible. These preliminary safety data are incomplete, but to date, there are no published data at single pramipexole doses > 2.25 mg for any condition. Until more sophisticated, randomized, blinded, placebo-controlled studies are completed, these data may illustrate potential benefits and important precautions when considering pramipexole for FM.

ANDREW J. HOLMAN, MD, Pacific Rheumatology Associates, 4300 Talbot Road South, Suite 101, Seattle, Washington 98055, USA.


1. Wolfe F, Smythe HA, Yunus MB, et al. The American College of Rheumatology 1990 criteria for the classification of fibromyalgia. Report of the Multicenter Criteria Committee. Arthritis Rheum 1990;33:160-72.

2. Moldofsky H, Scarisbrick P, England R, Smythe H. Musculoskeletal symptoms and non-REM sleep disturbance in patients with "fibrositis syndrome" and healthy subjects. Psychosom Med 1975;37:341-51.

3. Yunus M, Aldag J. Restless legs syndrome and leg cramps in fibromyalgia syndrome: a controlled study. BMJ 1996;312:1339.

4. Saletu M, Anderer P, Saletu-Zyhlarz G, et al. Restless legs
syndrome and periodic limb movement: acute placebo-controlled sleep laboratory studies with clonazepam. Eur Neuropsychopharmacol 2001;11:153-61.

5. Holman AJ. Effect of lorazepam on pain score for refractory fibromyalgia [abstract]. Arthritis Rheum 1998;41 Suppl:A1359.

6. Holman AJ. Safety and efficacy of lorazepam for fibromyalgia after one year [abstract]. Arthritis Rheum 1999;42 Suppl:A487.

7. Lin SC, Kaplan J, Burger CD, Fredrickson PA. Effect of
pramipexole in treatment of resistant restless legs syndrome. Mayo Clin Proc 1998;73:497-500.

8. Holman AJ. Safety and efficacy of the dopamine agonist,
pramipexole, on pain score for refractory fibromyalgia [abstract]. Arthritis Rheum 2000;43 Suppl:A1599.

9. Holman AJ. Ropinirole, a dopamine agonist, for refractory fibromyalgia: preliminary observations [letter]. J Clin Rheumatol 2003;9:277-9.

Return to December 2003 Table of Contents

© 2003. The Journal of Rheumatology Publishing Company Limited.
All rights reserved.