Radiographic and Clinical Outcome in Early Juvenile Rheumatoid Arthritis and Juvenile Spondyloarthropathy: a 3-Year Prospective Study ANNE M

Home , Spondyloarthropathy

Radiographic and Clinical Outcome in Early Juvenile Rheumatoid Arthritis and Juvenile Spondyloarthropathy: A 3-Year Prospective Study ANNE M. SELVAAG, BERIT FLATØ, KNUT DALE, GUNHILD LIEN, ODD VINJE, ANNA SMERDEL-RAMOYA, and ØYSTEIN FØRRE

ABSTRACT. Objective. To describe radiographic findings at disease onset and 3-year followup in patients with juvenile rheumatoid arthritis (JRA) and juvenile spondyloarthropathy (JSpA), to assess radiographic progression and its predictors, and to prospectively assess clinical outcome and predictors of persistent disease at 3-year followup. Methods. A total of 197 patients with JRA/JSpA were examined every 6 months for 3 years. Radiographic examination was performed at baseline and 3-year followup of knees and ankles (all patients) and of other joints on clinical indication. Remission was defined as minimum 6 months without medication and no clinical signs of active disease. Results. Radiographic abnormalities were found in 88% of the patients at onset and in 81% after 3 years. Frequency of swelling/osteoporosis decreased and frequency of abnormal growth increased from baseline to followup. Knees, hands, and wrists had most frequently radiographic abnormalities. Radiographic progression occurred in 38% of the patients. Joints with swelling/osteoporosis on radiographs, young age, and a large number of mobility-restricted joints at baseline were predictors of radiographic progression. At 3 years, 26% of the patients were in remission and 75% had been treated with disease-modifying antirheumatic drugs. Reduced well-being, a large number of active joints and negative antinuclear antibody at baseline were predictors of persistent disease after 3 years. Conclusion. After 3 years most patients had radiographic abnormalities and persistent disease. Young age, many affected joints, reduced well-being, and negative antinuclear antibody at onset increased the risk of radiographic progression and persistent disease after 3 years. (First Release June 1 2006; J Rheumatol 2006;33:1382–91)

Key Indexing Terms: JUVENILE RHEUMATOID ARTHRITIS JUVENILE SPONDYLOARTHROPATHY RADIOGRAPHY REMISSION PROGNOSIS RISK FACTORS Chronic arthritis in childhood often causes changes such as destruction may occur earlier5,6. Few studies have examined growth disturbances and destruction of bone in the joint area. prospectively the early radiographic changes in juvenile Radiographic documentation of the joint damage is an impor- rheumatoid arthritis (JRA) and juvenile spondyloarthropathy tant tool for the rheumatologist when assessing disease. (JSpA) in a systematic way. Radiographic progression has Studies have shown that about 25% of children with child- usually been measured by comparing frequencies of changes hood arthritis have radiographic damage after 6–10 years of at onset and at followup and not in terms of the progression of 1-3 disease duration . Radiographic erosions are known to be a changes in the individual patient. 1,4 late sign of severe arthritis , but some studies indicate that Several attempts have been made to establish and evaluate standardized radiographic assessment methods for children with juvenile arthritis6-13, but an internationally acceptable From the Department of Rheumatology, Rikshospitalet University Hospital, Oslo, Norway. radiographic assessment standard is not available. More stud- Supported by the Norwegian Foundation for Health and Rehabilitation ies using repeat standardized radiographs from representative through the Norwegian Rheumatism Association, the Norwegian Society patient cohorts with JRA/JSpA are needed to contribute to for Rheumatology, the Olga Imerslund Foundation, the Solveig Amalie standardization of radiographic readings for childhood arthri- Husbys Memorial Foundation, the Grethe Harbitz Legacy, and the Department of Rheumatology, Rikshospitalet University Hospital. tis. Predictors of erosions in JRA/JSpA have been found in 2,3,14 A.M. Selvaag, MD, Research Fellow; B. Flatø, MD, PhD, Senior longterm studies , but predictors of radiographic changes Consultant; K. Dale, MD, PhD, Professor; G. Lien, MD, PhD, Research in early disease course have not been explored prospectively. Fellow; O. Vinje, MD, PhD, Senior Consultant; A. Smerdel-Ramoya, MSc, Little is known about HLA genes and their association with PhD; Ø. Førre, MD, PhD, Professor. radiographic findings in childhood arthritis14. Address reprint requests to Dr. A.M. Selvaag, Department of Rheumatology, Rikshospitalet University Hospital, 0027 Oslo, Norway. Disease activity in juvenile arthritis may continue for many E-mail: [email protected] years, but may also remit with minor or no sequelae. Thirty to Accepted for publication February 26, 2006. sixty percent of children with arthritis are in remission after

1382 The Journal of Rheumatology 2006; 33:7

Downloaded on September 29, 2021 from www.jrheum.org 7–10 years of disease duration2,3,15. Some of the variations in positive. Positive IgM rheumatoid factor (RF) was defined as titer ≥ 64 by the Rose-Waaler test. HLA-B27 was determined by serologic testing24. HLA- these frequencies were due to the lack of established criteria DRB1 typing was performed using the nonisotopic method, based on poly- for remission. Recently, consensus was reached on prelimi- merase chain reaction (PCR) and hybridization with oligonucleotide probes nary criteria for remission in selected categories of juvenile labeled with biotin25. The HLA-DRB1 associations studied included DR1 idiopathic arthritis (JIA)16. Studies have shown that most (DRB1*01), DR4 (DRB1*04), DR5 (DRB*11 and *12), and DR8 remissions occur within 5 years after onset15,17, but few stud- (DRB1*08) alleles. HLA-DPB1 alleles were typed by sequence-specific oligonucleotide probing using a PCR, as described26. The HLA-DPB1 asso- ies have followed patients prospectively to assess remission ciations studied included DP2 (DPB1*02) and DP3 (DPB1*03) alleles. rates and predictors in early disease course. Predictors of per- 2,14 Radiographic examination. Radiographs from 174 patients were obtained by sistent disease after longterm followup have been reported , conventional technology with high resolution. In 23 patients a digital tech- but little is known about predictors of persistent disease early nique was used in accordance with the EULAR guidelines27. in the disease course18. The knees and ankles of all the patients were radiographed at baseline and The aim of our study was to: (1) prospectively describe the followup. Radiographic examination was done on clinical indication for other joints/joint groups. Contralateral joints were always included. The following radiographic findings in peripheral joints in early JRA/JSpA 8 joints/joint groups were studied: shoulders, elbows, wrists (radiocarpal at onset and after 3 years, (2) define radiographic progression joints and carpus), hands (interphalangeal and metacarpophalangeal joints), based on a standardized grading system for radiographic read- hips, knees, ankles, and feet (tarsal joints, metatarsophalangeal and interpha- ings and identify predictors of progression from among langeal joints). The radiographs of these joints were scored in chronological patient characteristics and measures of disease activity, and order by one of 2 trained radiologists (KD and VJ) not blinded to patient data, according to the Dale radiographic classification system for JRA: grade 0 (3) assess 3-year short-term clinical outcome and predictors of (normal joints), grade 1 (juxtaarticular osteoporosis and/or periarticular soft persistent disease at 3 years. tissue swelling), grade 2 (growth abnormality, bony erosion not present), grade 3 (growth abnormality and marginal bony erosions), grade 4 (deforma- MATERIALS AND METHODS tion and severe erosions), and grade 5 (gross destruction and deformity)9,28,29. Patients. Two hundred twenty-seven patients with JRA or JSpA and less than Interobserver error has been found to be low for this classification system9. 18 months of disease duration were admitted to the Department of An abnormal radiograph was defined as grade 1 or higher. Radiographic pro- Rheumatology, Rikshospitalet University Hospital, Oslo, between March gression was defined as an increase in the radiographic grade between base- 1995 and December 1999. Of these, 197 (87%) patients (61% girls) partici- line and followup, or from the first to the last radiograph taken during this pated in the study and were examined by a pediatric rheumatologist every 6 interval. An increase from grade 0 to ≥ 2, from grade 1 to ≥ 2, or from grade months for a mean ± SD of 3.2 ± 0.4 years. 2 to ≥ 3 was considered to be progression. An increase from grade 0 to 1 was Thirty (13%) of the 227 patients with JRA/JSpA did not participate. The not considered to be radiographic progression. parents of 18 of these children chose not to participate, 4 parents did not know Measures of physical function. The impact of arthritis on physical function Norwegian well enough, and 8 were excluded due to incomplete data. The was assessed by the Norwegian version of the Childhood Health Assessment participants were comparable to the nonparticipants with regard to age, sex, Questionnaire (CHAQ)30,31. The CHAQ is a disease-specific instrument that onset type, and disease duration (data not shown). measures physical disability and gives an index with values between 0 (no One hundred eighty-five of the patients had JRA according to the disability) and 3 (severe disability). The parents completed the CHAQ. Pain 19 American College of Rheumatology criteria and 12 had JSpA [3 had juve- and parent’s global assessment of child’s well-being were measured on a 10 20 nile ankylosing spondylitis (JAS) , 4 had the syndrome of seronegative cm visual analog scale (VAS), where 0 means no pain or doing very well and 21 enthesopathy and arthropathy (SEA) , and 5 had juvenile psoriatic arthritis 10 means very severe pain or doing very poorly. 22 (JPsA) ]. No patients with JPsA had clinical sacroiliitis. Our protocol com- Statistical analyses. The differences between participants and nonparticipants prised patients with arthritis only, and excluded patients with enthesitis with- and between 2 patient groups were measured by the independent sample t test out arthritis. The JIA classification criteria were published in 1998, after the 23 and the chi-square test. One-way analysis of variance (ANOVA) was used to start of our study . Our patients were not reclassified according to these cri- compare the numbers of images showing abnormalities per patient in the dif- teria because this would have excluded 33 patients (17%) with undifferenti- ferent subtypes. ated arthritis. Logistic regression analyses were used to assess predictors of radi- Disease onset was defined as the date on which a physician documented ographic progression and persistent disease after 3 years. Initial univariate arthritis or systemic features. Remission was defined as using no medical tests analyzed the relation between outcome and patient characteristics (age, treatment and having no signs of active arthritis, systemic involvement, or female sex, disease duration at baseline), genetic markers (HLA-B27, DR1, active uveitis for the previous 6 months. DR4, DR5, DR8, DP2, DP3), and measures of disease activity assessed at The patients came from all parts of Norway except the counties Nord- baseline or within the first 6 months [disease subtype, disease duration before Trøndelag, Troms, and Finnmark. The study was approved by the Regional administration of disease-modifying antirheumatic drugs (DMARD), Waaler Ethics Committee for Medical Research. Informed consent was obtained from RF, ANA, CHAQ, pain, well-being, physician’s global assessment of disease all the participants. activity, number of mobility-restricted joints, number of active joints, CRP, Clinical and laboratory data. The patients were examined by one of 5 pedi- ESR, and radiographic findings]. atric rheumatologists (BF, DS, OV, GL, or AMS). The clinical examination Clinical variables were dichotomized with a cutpoint at the sum of the included registration of the number of joints with swelling, tenderness and means for those with and those without radiographic progression/remission limited range of motion, the number of active joints (swelling or both tender- divided by 2. Subsequent multivariate analyses were used to identify a set of ness and limited range of motion), and physician’s global assessment of dis- statistically significant predictors. Age, sex, and variables with a p value < ease activity (on a 5-point Likert scale, where 1 means inactive, 2 mild, 3 0.10 in the univariate analysis were analyzed as possible predictors in the moderate, 4 severe, and 5 very severe disease activity). Erythrocyte sedimen- multivariate analysis. Highly intercorrelated independent variables (r > 0.7) tation rate (ESR) and C-reactive protein (CRP) were measured by standard were avoided in the analysis. Backward and forward regression methods were methods. Antinuclear antibody (ANA) titer ≥ 64, measured by indirect used. The explanatory power of the variables was measured by the immunofluorescence using mouse liver sections as substrate, was considered Nagelkerkes R2. Odds ratios (OR) and 95% confidence intervals (CI) were

Selvaag, et al: Radiographic changes in early JRA/JSpA 1383

Downloaded on September 29, 2021 from www.jrheum.org calculated. The only variables with missing values were DR8 and ANA [35 tissue swelling and/or osteoporosis (grade 1) decreased from (18%) missing] and the missing values were replaced using median 24% at baseline to 11% at followup (p < 0.001). Growth substitution. To study the disease course, we performed ANOVA for repeated meas- abnormalities (grade 2), the most frequent radiographic find- urements on each of the study variables (CHAQ, global assessment of well- ing at baseline, increased from 30% at baseline to 44% at fol- being, pain, physician’s global assessment of disease activity, number of lowup (p = 0.005). Joint erosions (grade 3) were found in 3 mobility-restricted and active joints, ESR, and CRP as dependent variables). joints/joint groups (0%) at baseline and in 18 (3%) joint Remission and months of followup (0, 6, 12, 18, 24, 30, 36) were used as groups at followup (p = 0.025). No patient had deformation fixed factors and identification numbers as random factors (nested to being in remission or not after 3 years). Residuals were normally distributed. This and severe erosions (grade 4) or gross destruction and defor- analysis shows whether there is a difference in mean estimated level between mity (grade 5) in our study. those who went into remission and those who did not during the 3 years. We For joints/joint groups with clinical arthritis, abnormalities applied ANOVA to each variable to examine whether those who went into were demonstrated in 76% of the radiographs at baseline and remission and those who did not improved during the 3 years (corrections for in 74% at followup (Table 1). For the radiographs taken of multiple comparisons were not performed). For all analyses, p values < 0.05 (2-tailed tests) were considered statisti- joints without clinical arthritis, we found that 25% of the cally significant. All statistical analyses were performed by the SPSS software radiographs had abnormalities at baseline, increasing to 28% program (SPSS Inc., Chicago, IL, USA) versions 10.0/11.0. at followup. The frequency of swelling/osteoporosis decreased from RESULTS baseline to followup in the knees (33% vs 5%; p < 0.001) and The patient group comprised 197 patients (121 females and 76 the ankles (28% vs 12%; p = 0.004) (Table 1). The frequency males) with a mean ± SD age of 6.9 ± 4.2 years at baseline. of growth abnormalities increased from baseline to followup One hundred eleven patients (56%) had oligoarthritis [16 in the knees (43% vs 61%; p < 0.001) and the ankles (25% vs (8%) of these developed extended oligoarthritis], 55 (28%) 32%; p = 0.028). Radiographic changes were most often had RF-negative polyarthritis, 5 (3%) had RF-positive pol- found in the knees and hands (76% and 64%, respectively) at yarthritis, 14 (7%) had systemic arthritis, 7 (4%) had baseline, and in the wrists, knees and elbows (74%, 66%, and JAS/SEA, and 5 (3%) had JPsA. The mean ± SD disease dura- 74%) at followup (Table 1). Joint erosions were mainly found tion was 4.7 ± 3.6 months at baseline and 42.6 ± 7.4 months in the wrists (5, 10%) and hands (5, 9%) at followup, but were at followup after a mean of 3.2 ± 0.4 years. The mean symp- also found in the ankles (3, 2%), elbows (2, 11%), shoulders tom duration at baseline was 10.1 ± 9.9 months. Positive ANA (1, 33%), feet (1, 2%), and knees (1, 1%). was found in 53 patients (33%) out of 162 tested and HLA- A total of 174 patients (88%) had radiographic abnormali- B27 was positive in 43 patients (22%). ties in one or more of the joint groups at baseline and 159 At followup, 147 patients (75%) had been or were current- patients (81%) had abnormal findings at followup. Fifty-nine ly being treated with one or more DMARD. Methotrexate was patients (30%) had abnormalities in one joint/joint group, 63 the most frequently used DMARD (51%). The mean disease (32%) in 2, 22 (11%) in 3, 6 (3%) in 4, 8 (4%) in 5, and one duration before DMARD was 6.6 ± 7.4 months. Eighty-five (1%) in 7 joints/joint groups. patients (43%) were treated with corticosteroids at some time Abnormal growth caused leg length discrepancies > 5 mm in during the 3 years of followup. 59 (30%) of the patients during the study, but persisted in only The 197 patients had a total of 498 joints/joint groups (one 22 patients (11%) at followup (17 of 22 had been treated with or more joints of the maximum 8 joint groups) with clinical intraarticular corticosteroid injections in the knee). Eight signs of arthritis at baseline or during the 3 years of followup. patients (4%) had leg length discrepancies > 10 mm and 4 The knees were the joint group most frequently affected (77% patients had ≥ 15 mm. Two patients underwent epiphysiodesis. of the patients), followed by ankles (58%), hands (34%), Erosions were found in 12 patients at followup, 6 patients wrists (30%), feet (23%), elbows (21%), hips (7%), and shoul- had 2 erosions, and 6 patients had one erosion each. Thirteen ders (2%). There was a mean of 2.5 involved joint groups per of the 18 erosions (72%) were in the upper limb. Three of the patient. erosions were also seen at baseline. Thus 15 erosions devel- Radiographically evaluated joints/joint groups. A total of oped during the 3 years. 1297 radiographs were obtained, corresponding to 6.6 images Radiographic abnormalities in disease subtypes. We also per patient. Radiographs were obtained for 88% of the looked at the distribution of radiographic changes per patient joints/joint groups with clinical arthritis at baseline and for according to JRA subtype (Figure 1). At baseline the number 73% at followup. The most complete radiographic data were of radiographs with growth abnormalities was higher in RF- obtained for knees (95% of joints with arthritis at baseline and negative polyarthritis (1.80) than in persistent oligoarthritis followup) and ankles (94% of joints with arthritis at baseline (0.86; p = 0.001) or systemic arthritis (0.36; p = 0.007). The and 95% at followup; data not shown). difference between RF-negative polyarthritis and persistent Radiographic abnormalities were found in 55% of the 744 oligoarthritis was still significant at followup (1.07 vs 1.01; p radiographs obtained at baseline and 58% of the 553 radi- = 0.004). The number of radiographs with erosions was high- ographs obtained at followup (Table 1). The frequency of soft er in RF-positive polyarthritis (0.40) than in all other subtypes Personal non-commercial use only. The Journal of Rheumatology Copyright © 2006. All rights reserved.

1384 The Journal of Rheumatology 2006; 33:7

Downloaded on September 29, 2021 from www.jrheum.org Table 1. Radiographic findings at baseline and 3-year followup in 197 patients with juvenile rheumatoid arthritis and juvenile spondyloarthropathy.

Baseline 3 Year Followup Joints Total No. of Abnormal, Grade 1, Grade 2, Grade 3, Total no. of Abnormal, Grade 1, Grade 2, Grade 3, Radiographs no. (%) no. (%) no. (%) no. (%) Radiographs no. (%) no. (%) no. (%) no. (%)

Total 743 405 (55) 181 (24) 221 (30) 3 (0) 554 322 (58) 60 (11) 244 (44) 18 (3) Clinical arthritis* Yes 437 330 (76) 146 (33) 181 (41) 3 (1) 366 270 (74) 47 (13) 206 (56) 17 (5) No 306 75 (25) 35 (11) 40 (13) 0 (0) 188 52 (28) 13 (7) 38 (20) 1 (1) Joint group† Knees 180 137 (76) 60 (33) 77 (43) 0 (0) 187 124 (66) 9 (5) 114 (61) 1 (1) Ankles 167 87 (52) 46 (28) 41 (25) 0 (0) 185 86 (46) 23 (12) 60 (32) 3 (2) Hands 100 64 (64) 29 (29) 34 (34) 1 (1) 55 36 (65) 8 (15) 23 (42) 5 (9) Wrists 100 46 (46) 16 (16) 29 (29) 1 (1) 50 37 (74) 10 (20) 22 (44) 5 (10) Feet 105 41 (39) 14 (13) 26 (25) 1 (1) 46 20 (43) 4 (9) 15 (33) 1 (2) Elbows 50 24 (48) 13 (26) 11 (22) 0 (0) 19 14 (74) 4 (21) 8 (42) 2 (11) Shoulders 10 1 (10) 1 (10) 0 (0) 0 (0) 3 1 (33) 0 (0) 0 (0) 1 (33) Hips 31 5 (16) 2 (6) 3 (10) 0 (0) 9 4 (44) 2 (22) 2 (22) 0 (0)

Grade 1: swelling/osteoporosis; grade 2: abnormal growth; grade 3: abnormal growth and erosions. * Joints with or without signs of clinical arthritis (swelling or both tenderness and limited range of motion) at baseline or during followup. † One or more joints out of a maximum of 8 radiographed joint sites (range 0-8) (hands include metacarpophalangeal and interphalangeal joints, wrists include radiocarpal joints and carpus, and feet include tarsal joints, metatarsophalangeal, and interphalangeal joints). The contralateral joint is included on every radiograph.

Figure 1. The number of radiographs with growth abnormalities (grade 2) per patient (y axis) in each subtype of JRA at baseline and followup. *p = 0.001 vs persistent oligoarthritis. †p = 0.007 vs systemic arthritis. ‡p = 0.004 vs persistent oligoarthritis. at baseline (p < 0.001; Figure 2). At followup, joint erosions joints/joint groups showed radiographic progression (one were more frequent in radiographs of RF-positive polyarthri- joint/joint group in 45 patients, 2 in 23 patients, 3 in 5 patients, tis than in those of persistent oligoarthritis (0.60 vs 0.03; p = and 4 in 2 patients), and 358 joints/joint groups were 0.022). Nine of the 12 patients with erosions had polyarticular unchanged. The radiographic progression was in the form of disease course. abnormal growth in 102 (89%) of the joints/joint groups and Radiographic progression. Seventy-five patients (38%) had in the form of developing joint erosions in 12 (11%) of the radiographic progression at 3-year followup. The frequencies joints/joint groups. of patients with radiographic progression were 34% in per- Predictors of radiographic progression. Predictors of radio- sistent oligoarthritis, 44% in extended oligoarthritis, 49% in graphic progression among patient characteristics, HLA alle- RF-negative polyarthritis, 40% in RF-positive polyarthritis, les, and disease variables at baseline were analyzed by logis- 21% in systemic arthritis, and 33% in the JSpA group (not sig- tic regression analyses. Joints with swelling/osteoporosis on nificant). radiographs, young age at onset, and a large number of joints Nine hundred forty-four (73%) of the radiographs were with limitation of motion at baseline were predictors of radio- matched from baseline to followup. One hundred fourteen graphic progression (Table 2). Personal non-commercial use only. The Journal of Rheumatology Copyright © 2006. All rights reserved.

Selvaag, et al: Radiographic changes in early JRA/JSpA 1385

Downloaded on September 29, 2021 from www.jrheum.org Figure 2. The number of radiographs with erosions (grade 3) per patient (y axis) in each subtype of JRA at baseline and followup. *p < 0.001 vs all other subtypes. †p = 0.036 vs persistent oligoarthritis.

Table 2. Patient characteristics and disease variables measured at baseline predicting radiographic progression after 3 years in juvenile rheumatoid arthritis and juvenile spondyloarthropathy patients (n = 197).

Characteristics/ Univariate Analysis*, p Multivariate Analysis*†,p Disease Variables OR (95% CI) OR (95% CI)

Onset age, yrs 0.90 (0.84–0.97) 0.008 0.89 (0.82–0.97) 0.009 Female 1.58 (0.86–2.89) 0.138 RF-negative polyarthritis 1.89 (0.92–3.29) 0.049 CHAQ, ≥ 0.6 1.70 (0.94–3.07) 0.079 No. of joints with LOM, ≥ 3 2.88 (1.52–5.46) 0.001 2.49 (1.17–5.30) 0.018 No. of active joints, ≥ 4 2.87 (1.50–5.48) 0.001 Radiographic findings Swelling/osteoporosis 9.31 (4.53–19.11) < 0.001 7.95 (3.80–16.64) < 0.001 Growth abnormality 0.43 (0.24–0.78) 0.006

In patients with oligoarticular-onset JRA, the most impor- Figure 3 shows the course of important disease variables tant predictors of radiographic progression were joints with over 3 years for patients who achieved remission and for those swelling/osteoporosis on radiographs at onset and the pres- who did not. The estimated mean levels of the disease vari- ence of DRB1*08 (OR 14.53, 95% CI 4.94–42.75, p < 0.001, ables were significantly different between those with and 2 and OR 3.11, 95% CI 1.07–9.02, p = 0.037; Nagelkerkes R = those without remission in terms of CHAQ score, well-being, 37.1%). In patients with polyarticular-onset JRA the most pain, physician’s global assessment of disease activity, num- important predictors of radiographic progression were joints ber of mobility-restricted or active joints, and ESR (p < 0.001 with swelling/osteoporosis on radiographs and elevated CRP for all differences except for ESR, p = 0.014), but not for CRP at baseline (OR 39.7, 95% CI 3.56–442.80, p = 0.005, and OR (p = 0.065). 2 1.03, 95% CI 1.00–1.08, p = 0.046; Nagelkerkes R = 44.7%). Patients who went into remission showed improvements Remission. Fifty-two patients (26%) were in remission with- within 12 months in physician’s global assessment of disease out medication after 3 years. The remission rate was 39% in activity, number of joints with limitation of motion, number of persistent oligoarthritis, 6% in extended oligoarthritis, 16% in active joints, and ESR (p ≤ 0.001; Figure 3). Patients with RF-negative polyarthritis, 0% in RF-positive polyarthritis, persistent disease showed improvements in physician’s global 29% in systemic arthritis, 0% in JAS/SEA, and 20% in JPsA. assessment of disease activity and ESR at 12 months (p < The remission rate in RF-negative polyarthritis was signifi- 0.001). There was no significant improvement in patient- cantly lower than in persistent oligoarthritis (p = 0.004). reported disease variables in either group of patients. Personal non-commercial use only. The Journal of Rheumatology Copyright © 2006. All rights reserved.

1386 The Journal of Rheumatology 2006; 33:7

Downloaded on September 29, 2021 from www.jrheum.org Figure 3. The course of different disease variables during 3 years of followup of patients with juvenile arthritis, showing those who achieved remission and those who did not. The x axis represents the number of months in the study. The y axis represents the measurement scale for each variable. During the first year an improvement occurred in physician-reported variables and in ESR. *p ≤ 0.001 for improvement from baseline to 12 months.

Predictors of persistent disease. Logistic regression analyses erosions were confirmed was 32.8 months ± 15.3 (range 3–50 were used to identify predictors of persistent disease after 3 months). years. Reduced well-being, a large number of active joints, At followup, 52% of the patients had a CHAQ disability and negative ANA at baseline were the most important pre- index of 0. Forty-three percent of the patients with radi- dictors of persistent disease (Table 3). ographic progression compared with 57% of the patients with In patients with persistent oligoarthritis, high levels of pain no radiographic progression had a CHAQ disability index of at baseline predicted persistent disease after 3 years (OR 2.8, 0 (p = 0.045). The patients with radiographic progression had 2 95% CI 1.1–7.2, p = 0.030; Nagelkerkes R = 8.1%). We did a mean CHAQ score of 0.44 ± 0.57 at followup compared not find any predictors of persistent disease in extended with 0.30 ± 0.49 in those who had no progression (p = 0.065). oligoarthritis, or RF-negative or RF-positive polyarthritis (data not shown). DISCUSSION Relation between radiographic progression, remission, and This was a prospective study of radiographic changes and disability. Radiographic progression was found in 27% of the their relation to HLA genes and disease progression in a patients in remission and 42% of the patients with persistent cohort of patients with early JRA/JSpA. Radiographic abnor- disease (p = 0.054). All the patients with erosions had persist- malities were present in 81% of the patients after 3 years of ent disease after 3 years. Mean disease duration before joint followup. Radiographic progression occurred in 38% and

Selvaag, et al: Radiographic changes in early JRA/JSpA 1387

Downloaded on September 29, 2021 from www.jrheum.org Table 3. Patient characteristics and disease variables assessed at baseline in patients with juvenile rheumatoid arthritis and juvenile spondyloarthropathy that predicted persistent disease after 3 years (n = 197).

Characteristics/ Univariate Analysis*, p Multivariate Analysis*†,p Disesae Variables OR (95% CI) OR (95% CI)

Onset age, yrs 1.07 (0.99–1.16) 0.108 Female 1.53 (0.81–2.91) 0.192 RF-negative polyarthritis 2.22 (1.00–4.94) 0.050 ANA, ≥ 64 0.47 (0.24–0.93) 0.030 0.42 (0.20–0.89) 0.023 CHAQ, ≥ 0.45 3.21 (1.58–6.52) 0.001 Pain, ≥ 22 mm 3.21 (1.44–7.13) 0.004 Well-being, ≥ 24 mm 3.30 (1.63–6.70) 0.001 3.06 (1.80–6.34) 0.003 No. of joints with LOM, ≥ 3 2.64 (1.15–6.04) 0.022 No. of active joints, ≥ 3 2.86 (1.39–5.90) 0.004 3.17 (1.47–6.80) 0.003 Physician global assessment of disease activity 1.81 (0.92–3.57) 0.088

OR: Adjusted odds ratio per unit change in the variable. CHAQ: Childhood Health Assessment Questionnaire. LOM: limitation of motion. * Results of logistic regression analyses with persistent active disease after 3 yrs as the dependent variable. Persistent disease was present in 145 patients (73.6%). *† Final model including variables significantly associated with persistent disease in the univariate analysis (p < 0.100) as independent variables. Disease duration, disease duration before disease modifying antirheumatic drugs, DR1, DR4, DR5, DR8, DP2, DP3, HLA-B27, Waaler RF, ESR, CRP, and radiographic findings were not significantly associated with persistent disease (data not shown). Clinical variables were dichotomized. Nagelkerkes R2 = 17.6%. remission in 26%. Radiographic findings of swelling/osteo- “joint space narrowing” as a separate category in our grading porosis, young age, a large number of mobility-restricted system9. joints, and elevated CRP at baseline and the presence of Radiographs of knees and ankles were taken for all the DRB1*08 were predictors of radiographic progression. patients, while radiographs of other joints were taken on clin- Predictors of persistent disease at 3 years were reduced well- ical indication. If radiographs had been taken of all joints/joint being, a large number of active joints, and negative ANA at groups both at baseline and at followup, this would have given baseline. a more exact radiological picture, but concerns regarding the Our patient group is a referral-based cohort that represents total radiation dose made this problematic. about half of the estimated incidence of JRA and JSpA After 3 years of followup, 81% of our patients had abnor- patients in Norway during the study period. These patients mal radiographs of at least one joint/joint group, and more probably have more severe disease than unselected cohorts than half of all radiographs showed abnormalities. This is in from the general population. However, our patient group is line with a study of JIA patients at a median of 24 months of comparable to the JRA/juvenile chronic arthritis (JCA) disease duration, which found radiographic abnormalities in at patients in a number of epidemiological studies as regards least one joint in 87% of the patients and in half of the radio- age, sex, and distribution of onset type32-34. graphs of clinically involved joints12. In a group of patients The high frequency of HLA-B27 in our patient group is in with polyarticular JRA, 61% of radiographs taken at onset accordance with another Norwegian study35. In the were abnormal, but these radiographs were of the hand and Norwegian population the frequency of HLA-B27 varies from wrist only11. Another study reported that 27% of the patients 10% in the south to 16% in the north36. If patients with with pauciarticular, 67% of those with polyarticular, and 75% JAS/SEA are excluded, the frequency of HLA-B27 in our of those with systemic JRA had abnormal radiographs in study is 19%. A linkage disequilibrium between HLA-DR8 terms of joint space narrowing 5 years after onset38. Although and HLA-B27 in Norwegian patients with JRA has been the definition of radiographic abnormality varies between found and may explain the excess of HLA-B27 (R. Ploski, studies, radiographic abnormalities seem to be frequent in personal communication). However, our study covers the first early childhood arthritis. years of the disease, and our patient group with oligoarthritis There are few studies describing early radiographic pro- could contain patients who will develop JSpA later. gression in a referral-based cohort. We found radiographic The Dale radiographic grading system has been used in progression in 38% of the patients, including 44% of those previous studies3,14,17. The system distinguishes between with extended oligoarthritis, 49% of those with RF-negative arthritis with and without erosions. In some recent studies polyarthritis, and 40% of those with RF-positive polyarthritis. joint space narrowing has been included in the description of Our results are consistent with those of a recently published radiographic changes12,37. Joint space narrowing can to some study where 46% of the patients with polyarticular JRA extent be explained by the enlargement of the epiphysis due to showed radiographic progression as measured by the JRA and the subsequent alteration in the bone/cartilage ratio. Poznanski score39. For this reason and because the true joint cartilage may be Abnormal skeletal growth was the most common radio- diminished due to secondary osteoarthritis, we did not include graphic finding in our study, and the frequency increased from Personal non-commercial use only. The Journal of Rheumatology Copyright © 2006. All rights reserved.

1388 The Journal of Rheumatology 2006; 33:7

Downloaded on September 29, 2021 from www.jrheum.org 30% at onset to 44% at followup. Previous studies have oligoarticular JRA, 2–4 affected joints at disease onset were shown that the frequency of growth abnormalities varies from found to be a predictor of extended disease course17. In a 10% to 48% in patients with JRA during the first 2–3 years of recent study of predictors of persistent disease in JRA after 15 disease12,37,40. Abnormal growth in the form of enlargement years of disease duration, the number of affected joints in the of the epiphysis tends to be most pronounced in the distal first 6 months was found to be one of several predictors14. On femur, the proximal tibia, and the radial head41. Active arthri- the other hand, a study of German patients with JCA/JSpA tis in a knee may influence leg length, especially in younger found that joint involvement at onset had no influence on the patients42,43. There were growth abnormalities in 61% of the probability of remission after 10 years15. knees and leg length discrepancies in 30% of our patients. A Well-being has not previously been reported as a predictor lower frequency (19%) of leg length discrepancy was found in of persistent JRA/JSpA, but has been shown to be an impor- a German study, after 7 years of followup15, but the longer fol- tant predictor of physical function after 3 years of childhood lowup period meant that the patients were correspondingly arthritis54. older than ours. In a recent study, patients with ANA-negative polyarthritis Knees, hands, and wrists were the most frequently (radio- were found to have more severe disease during the first 2 graphically) affected joints in our study. This is in line with years after disease onset than ANA-positive patients55. ANA the findings of other studies12,41,44,45. In contrast, one study positivity has been associated with less disability and an found that hips had a high frequency of radiographic joint absence of radiographic abnormalities in JRA/JIA12,56, but abnormalities, but these were measured as joint space narrow- with longer disease duration and an extended disease course in ing and erosions and thus were not fully comparable with our oligoarticular JRA57,58. The prognostic value of ANA status findings37. therefore merits further investigation. The joints most frequently affected by erosions in our We found a tendency toward a higher CHAQ disability study were hands and wrists, which has also been found in score among those who had radiographic progression than other studies4,12,31,37,40,46. We also found, in accordance with among those without progression. A recent study reported a previous reports, that patients with RF-positive polyarthritis good correlation between the probability of radiographic had a higher frequency of erosions than patients with the other abnormalities and overall severity score12. Other studies have 14,31,37,41,46 subtypes . found correlations between erosions or joint space narrowing Joints showing swelling/osteoporosis on radiographs, and a higher CHAQ disability index3,37. young age, and a large number of mobility-restricted joints at In this prospective study most patients had radiographic baseline were found in our study to be predictors of radio- abnormalities and persistent disease after 3 years. We found graphic progression. Most of this progression consisted of the Dale radiographic classification system to be suited to growth abnormalities. Joints with swelling and osteoporosis at describing radiographic findings. Further longitudinal studies onset have been found to be prone to growth disturbances if of the frequencies of radiographic abnormalities and radio- 47 the disease activity continues for 6 months or more , and graphic progression in representative cohorts of patients with growth abnormalities have been reported to be more common JIA should be made with a view to arriving at an internation- 42,47,48 among young patients . Our finding of a higher frequen- ally acceptable radiographic grading system for joint abnor- cy of growth abnormalities in patients with polyarticular dis- malities in juvenile arthritis. The importance of well-being, ease is also consistent with the findings of previous studies, number of active joints, and ANA as predictors for the where patients with polyarthritis have been reported to be longterm outcome of JRA and JSpA should be further 5,31,49 more likely to develop radiographic abnormalities . explored. The remission rate in our study corresponds well with other studies in which 28% to 39% of the patients with ACKNOWLEDGMENT JRA/JCA were found to be in remission after 7–10 The authors thank Dag Sørskaar for examining patients, Virginia Johnston for years2,50–52. As regards subtypes, our patients with persistent grading the radiographs, Helga Vonheim Bruaseth and Berit Halmrast for oligoarthritis had the highest remission rate, the patients with assistance with data collection, and Geir Aamodt and Thore Egeland for help with the statistics. systemic arthritis had medium-high remission rates, and the patients with polyarticular course and those with JSpA had the REFERENCES lowest rates. Previously these differences in remission rates 1. Cassidy JT, Martel W. Juvenile rheumatoid arthritis: between subtypes of juvenile arthritis have been found in clinicoradiologic correlations. Arthritis Rheum 1977;20 Suppl patients with longer disease duration2,14,15,50,51,53. In a recent 2:207-11. study where episodes of clinical remission were measured 2. Gare BA, Fasth A. The natural history of juvenile chronic arthritis: a population based cohort study. I. Onset and disease process. during followup, the same differences in remission rates were 52 J Rheumatol 1995;22:295-307. found among JIA subgroups . 3. Flato B, Aasland A, Vinje O, Forre O. Outcome and predictive Few studies have assessed predictors of persistent disease factors in juvenile rheumatoid arthritis and juvenile in early juvenile arthritis in a multivariate way. In a study of spondyloarthropathy. J Rheumatol 1998;25:366-75. Personal non-commercial use only. The Journal of Rheumatology Copyright © 2006. All rights reserved.

Selvaag, et al: Radiographic changes in early JRA/JSpA 1389

Downloaded on September 29, 2021 from www.jrheum.org 4. Williams RA, Ansell BM. Radiological findings in seropositive 24. Hill AV, Allsopp CE, Kwiatkowski D, Anstey NM, Greenwood BM, juvenile chronic arthritis (juvenile rheumatoid arthritis) with McMichael AJ. HLA class I typing by PCR: HLA-B27 and an particular reference to progression. Ann Rheum Dis African B27 subtype. Lancet 1991;337:640-2. 1985;44:685-93. 25. Verduyn W, Doxiadis II, Anholts J, et al. Biotinylated DRB 5. Levinson JE, Wallace CA. Dismantling the pyramid. J Rheumatol sequence-specific oligonucleotides. Comparison to serologic 1992;19 Suppl 33:6-10. HLA-DR typing of organ donors in eurotransplant. Hum Immunol 6. Magni-Manzoni S, Rossi F, Pistorio A, et al. Prognostic factors for 1993;37:59-67. radiographic progression, radiographic damage, and disability in 26. Ronningen KS, Spurkland A, Markussen G, Iwe T, Vartdal F, juvenile idiopathic arthritis. Arthritis Rheum 2003;48:3509-17. Thorsby E. Distribution of HLA class II alleles among Norwegian 7. Pettersson H, Rydholm U. Radiologic classification of joint Caucasians. Hum Immunol 1990;29:275-81. destruction in juvenile chronic arthritis. Acta Radiol Diagn Stockh 27. van Riel PLCM, van Gestel AM, Scott DL, on behalf of the 1985;26:719-22. EULAR Standing Committee for International Clinical Studies 8. Poznanski AK, Hernandez RJ, Guire KE, Bereza UL, Garn SM. Including Therapeutic Trials. The EULAR handbook of clinical Carpal length in children — a useful measurement in the diagnosis assessments in rheumatoid arthritis. Alpen an den Rijn, The of rheumatoid arthritis and some congenital malformation Netherlands: Van Zuiden Communications B.V.; 2000. syndromes. Radiology 1978;129:661-8. 28. Larsen A, Dale K, Eek M. Radiographic evaluation of rheumatoid 9. Dale K, Paus AC, Laires K. A radiographic classification system in arthritis and related conditions by standard reference films. Acta juvenile rheumatoid arthritis applied to the knee. Eur Radiol Radiol Diagn Stockh 1977;18:481-91. 1994;4:27-32. 29. Larsen A, Dale K. Standardized radiological evaluation of 10. Cohen PA, Job-Deslandre CH, Lalande G, Adamsbaum C. rheumatoid arthritis. In: Dumonde DC, Steward MW, editors. Overview of the radiology of juvenile idiopathic arthritis (JIA). Laboratory test in the rheumatic diseases: standardization in Eur J Radiol 2000;33:94-101. clinical-laboratory investigation. Lancaster: MTP Press Limited; 11. Mason T, Reed AM, Nelson AM, et al. Frequency of abnormal hand 1979. and wrist radiographs at time of diagnosis of polyarticular juvenile 30. Singh G, Athreya BH, Fries JF, Goldsmith DP. Measurement of rheumatoid arthritis. J Rheumatol 2002;29:2214-8. health status in children with juvenile rheumatoid arthritis. Arthritis 12. Van Rossum MA, Zwinderman AH, Boers M, et al. Radiologic Rheum 1994;37:1761-9. features in juvenile idiopathic arthritis: A first step in the 31. Flato B, Sorskaar D, Vinje O, et al. Measuring disability in early development of a standardized assessment method. Arthritis Rheum juvenile rheumatoid arthritis: evaluation of a Norwegian version of 2003;48:507-15. the childhood Health Assessment Questionnaire. J Rheumatol 13. Van Rossum MA, Boers M, Zwinderman AH, et al. Development 1998;25:1851-8. of a standardized method of assessment of radiographs and 32. Gare BA, Fasth A. Epidemiology of juvenile chronic arthritis in radiographic change in juvenile idiopathic arthritis: introduction of southwestern Sweden: a 5-year prospective population study. the Dijkstra composite score. Arthritis Rheum 2005;52:2865-72. Pediatrics 1992;90:950-8. 14. Flato B, Lien G, Smerdel A, et al. Prognostic factors in juvenile 33. Peterson LS, Mason T, Nelson AM, O’Fallon WM, Gabriel SE. rheumatoid arthritis: a case-control study revealing early predictors Psychosocial outcomes and health status of adults who have had and outcome after 14.9 years. J Rheumatol 2003;30:386-93. juvenile rheumatoid arthritis: a controlled, population-based study. 15. Minden K, Kiessling U, Listing J, et al. Prognosis of patients with Arthritis Rheum 1997;40:2235-40. juvenile chronic arthritis and juvenile spondyloarthropathy. 34. Moe N, Rygg M. Epidemiology of juvenile chronic arthritis in J Rheumatol 2000;27:2256-63. northern Norway: a ten-year retrospective study. Clin Exp 16. Wallace CA, Ruperto N, Giannini E. Preliminary criteria for Rheumatol 1998;16:99-101. clinical remission for select categories of juvenile idiopathic 35. Flato B, Smerdel A, Johnston V, et al. The influence of patient arthritis. J Rheumatol 2004;31:2290-4. characteristics, disease variables, and HLA alleles on the 17. Guillaume S, Prieur AM, Coste J, Job-Deslandre C. Long-term development of radiographically evident sacroiliitis in juvenile outcome and prognosis in oligoarticular-onset juvenile idiopathic idiopathic arthritis. Arthritis Rheum 2002;46:986-94. arthritis. Arthritis Rheum 2000;43:1858-65. 36. Gran JT, Mellby AS, Husby G. The prevalence of HLA-B27 in 18. Modesto C, Woo P, Garcia-Consuegra J, et al. Systemic onset Northern Norway. Scand J Rheumatol 1984;13:173-6. juvenile chronic arthritis, polyarticular pattern and hip involvement 37. Oen K, Reed M, Malleson PN, et al. Radiologic outcome and its as markers for a bad prognosis. Clin Exp Rheumatol relationship to functional disability in juvenile rheumatoid arthritis. 2001;19:211-7. J Rheumatol 2003;30:832-40. 19. Brewer EJJ, Bass J, Baum J, et al. Current proposed revision of 38. Bowyer SL, Roettcher PA, Higgins GC, et al. Health status of JRA criteria. JRA Criteria Subcommittee of the Diagnostic and patients with juvenile rheumatoid arthritis at 1 and 5 years after Therapeutic Criteria Committee of the American Rheumatism diagnosis. J Rheumatol 2003;30:394-400. Section of The Arthritis Foundation. Arthritis Rheum 1977;20 39. Mason T, Reed AM, Nelson AM, Thomas KB. Radiographic Suppl 2:195-9. progression in children with polyarticular juvenile rheumatoid 20. van der Linden S, Valkenburg HA, Cats A. Evaluation of diagnostic arthritis: a pilot study. Ann Rheum Dis 2005;64:491-3. criteria for ankylosing spondylitis. A proposal for modification of 40. Lang BA, Schneider R, Reilly BJ, Silverman ED, Laxer RM. the New York criteria. Arthritis Rheum 1984;27:361-8. Radiologic features of systemic onset juvenile rheumatoid arthritis. 21. Rosenberg AM, Petty RE. A syndrome of seronegative J Rheumatol 1995;22:168-73. enthesopathy and arthropathy in children. Arthritis Rheum 41. Ansell BM, Kent P.A. Radiological changes in juvenile chronic 1982;25:1041-7. polyarthritis. Skeletal Radiology 1977;1:129-44. 22. Southwood TR, Petty RE, Malleson PN, et al. Psoriatic arthritis in 42. Simon S, Whiffen J, Shapiro F. Leg-length discrepancies in children. Arthritis Rheum 1989;32:1007-13. monoarticular and pauciarticular juvenile rheumatoid arthritis. 23. Petty RE, Southwood TR, Baum J, et al. Revision of the proposed J Bone Joint Surg Am 1981;63:209-15. classification criteria for juvenile idiopathic arthritis: Durban, 1997. 43. Vostrejs M, Hollister JR. Muscle atrophy and leg length J Rheumatol 1998;25:1991-4. discrepancies in pauciarticular juvenile rheumatoid arthritis. Am

1390 The Journal of Rheumatology 2006; 33:7

Downloaded on September 29, 2021 from www.jrheum.org J Dis Child 1988;142:343-5. 53. Minden K, Niewerth M, Listing J, et al. Long-term outcome in 44. Michels H, Hafner R, Morhart R, Schuchmann L, Truckenbrodt H. patients with juvenile idiopathic arthritis. Arthritis Rheum Five year follow-up of a prospective cohort of juvenile chronic 2002;46:2392-401. arthritis with recent onset. Clin Rheumatol 1987;6 Suppl 2:87-92. 54. Selvaag AM, Lien G, Sorskaar D, Vinje O, Forre O, Flato B. Early 45. Chaplin D, Pulkki T, Saarimaa A, Vainio K. Wrist and finger disease course and predictors of disability in juvenile rheumatoid deformities in juvenile rheumatoid arthritis. Acta Rheumatol Scand arthritis and juvenile spondyloarthropathy: a 3 year prospective 1969;15:206-23. study. J Rheumatol 2005;32:1122-30. 46. Laaksonen AL. A prognostic study of juvenile rheumatoid arthritis. 55. Ravelli A, Felici E, Magni-Manzoni S, et al. Patients with Analysis of 544 cases. Acta Paediatr Scand 1966;1-163. antinuclear antibody-positive juvenile idiopathic arthritis constitute 47. Rosberg G, Laine V. Natural history of radiological changes of knee a homogeneous subgroup irrespective of the course of joint disease. joint in juvenile rheumatoid arthritis. Acta Paediatr Scand Arthritis Rheum 2005;52:826-32. 1967;56:671-5. 56. Ruperto N, Ravelli A, Levinson JE, et al. Long-term health 48. Fiszman P, Ansell BM, Renton P. Radiological assessment of knees outcomes and quality of life in American and Italian inception in juvenile chronic arthritis (juvenile rheumatoid arthritis). Scand cohorts of patients with juvenile rheumatoid arthritis. II. Early J Rheumatol 1981;10:145-52. predictors of outcome. J Rheumatol 1997;24:952-8. 49. Rennebohm R, Correll JK. Comprehensive management of juvenile 57. Oen K, Malleson PN, Cabral DA, et al. Early predictors of rheumatoid arthritis. Nurs Clin North Am 1984;19:647-62. longterm outcome in patients with juvenile rheumatoid arthritis: 50. Oen K, Malleson PN, Cabral DA, Rosenberg AM, Petty RE, subset-specific correlations. J Rheumatol 2003;30:585-93. Cheang M. Disease course and outcome of juvenile rheumatoid 58. Al Matar MJ, Petty RE, Tucker LB, Malleson PN, Schroeder ML, arthritis in a multicenter cohort. J Rheumatol 2002;29:1989-99. Cabral DA. The early pattern of joint involvement predicts disease 51. Fantini F, Gerloni V, Gattinara M, Cimaz R, Arnoldi C, Lupi E. progression in children with oligoarticular (pauciarticular) juvenile Remission in juvenile chronic arthritis: a cohort study of 683 rheumatoid arthritis. Arthritis Rheum 2002;46:2708-15. consecutive cases with a mean 10 year followup. J Rheumatol 2003;30:579-84. 52. Wallace CA, Huang B, Bandeira M, Ravelli A, Giannini EH. Patterns of clinical remission in select categories of juvenile idiopathic arthritis. Arthritis Rheum 2005;52:3554-62.

Selvaag, et al: Radiographic changes in early JRA/JSpA 1391

Downloaded on September 29, 2021 from www.jrheum.org