Prognostic factors for radiographic progression, radiographic damage, and disability in juvenile idiopathic arthritis

Authors


Abstract

Objective

To investigate the rate of radiographic progression and identify prognostic factors of radiographic progression, radiographic damage, and physical disability in juvenile idiopathic arthritis (JIA).

Methods

Ninety-four JIA patients with a median disease duration of 1.1 years were followed up prospectively for a median of 4.5 years. Bilateral wrist radiographs were obtained at baseline, at 1 year, and at the last followup visit. Radiographic damage was assessed by the carpal length (Poznanski score), and physical disability by the Childhood Health Assessment Questionnaire (C-HAQ). Yearly radiographic progression, the Poznanski score at the final visit, and the C-HAQ score at the final visit were used as outcome measures. Baseline parameters included demographic, clinical, laboratory, and radiographic data.

Results

The mean ± SD Poznanski score was −1.2 ± 1.3 at baseline, −1.7 ± 1.8 at the 1-year visit, and −1.9 ± 2.2 at the final visit (P < 0.0001). Radiographic progression was greater during the first year (mean ± SD −0.5 ± 1.1) than between the 1-year visit and the final visit (−0.2 ± 1.3). The mean yearly radiographic progression during the entire study period was −0.1 ± 0.4. Logistic regression analysis revealed that radiographic progression during the first year was the only baseline parameter that was predictive of all 3 study outcomes. The final Poznanski score was also predicted by the baseline Poznanski score, whereas female sex was protective against radiographic progression.

Conclusion

We identified the prognostic factors for poorer outcome in polyarticular-course JIA. The changes in the early Poznanski score can be used to predict long-term joint damage and physical disability.

Juvenile idiopathic arthritis (JIA) is a heterogeneous condition whose clinical course varies widely and is difficult to predict (1). Although it was previously believed that the disease had a good to excellent outcome in most cases, recent studies have shown that a higher than expected percentage of patients may develop unremitting disease and that there is a considerable risk of developing progressive joint destruction and serious functional disability (2). Predicting outcome in JIA is crucial for its optimal clinical management. It would be desirable to be able to distinguish patients with a high likelihood of an untoward outcome so as to better manage the disease course and to institute appropriately aggressive therapy at an early stage. This has become even more important since new therapies that can be effective in the most severe and refractory forms of JIA are now available (3).

The degree of joint damage caused by the synovitis process is traditionally assessed radiographically. Radiography is an important clinical tool for the evaluation of disease progression (4) and is usually seen as the “gold standard” for measuring the disease-modifying potential of antirheumatic drugs (5). In adult patients with rheumatoid arthritis, serial plain radiographs have been largely used to measure the progression of joint damage in clinical practice, clinical research, and therapeutic trials. However, little information exists on the use of radiography in the investigation of disease outcome in children with JIA. Furthermore, the assessment of radiographic progression has never been included in controlled trials of second-line agents in JIA. This chiefly reflects the lack of established, validated radiographic scoring systems for use in subjects of pediatric ages.

It is commonly believed that the traditional scoring methods used for adult rheumatoid arthritis (6, 7), which are based on the assessment of joint space narrowing and erosions, may not be suitable for the evaluation of pediatric joint diseases. In contrast to adults, it is difficult to reliably determine cartilage loss and erosions in children by simple examination of radiographs because ossification is incomplete and joint space width varies with age (8). In 1978, Poznanski et al (9) published standards for normal carpal length in growing children, thus enabling a measurement of carpal size that is not dependent on the degree of ossification. The carpal length reflects a reduction in the joint space of the wrist, rather than erosive damage, and may therefore be a good method for identifying cartilage damage in the early phases of joint diseases. Previous studies have shown that the assessment of carpal length represents a suitable radiographic tool for investigating the efficacy of second-line drug therapy in JIA (10, 11).

The purpose of the present study was to determine the rate of radiographic progression, as measured by carpal length, in patients with polyarticular-course JIA. We also attempted to determine prognostic factors for radiographic progression, radiographic damage, and physical disability.

PATIENTS AND METHODS

Patient selection.

Beginning in December 1986, all consecutive patients who were seen at the Department of Pediatrics, Istituto di Ricovero e Cura a Carattere Scientifico Policlinico S. Matteo in Pavia and fulfilled the International League of Associations for Rheumatology revised criteria for JIA (12), and had polyarthritis with bilateral wrist involvement underwent a biannual clinical assessment and a yearly wrist radiograph. To be eligible for this study, at least 3 wrist films had to be available for review by December 2001: the first radiograph obtained at study entry (baseline), the second after 1 year, and the third at the last followup visit (final).

Clinical assessment.

Patient characteristics recorded at baseline included: age at onset, sex, disease duration, and JIA subtype. The following assessments were made at baseline and every 6 months until the end of the study: physician's global assessment of overall disease activity, measured on a 10-cm visual analog scale (VAS) (0 = no activity; 10 = maximum activity); parent's global assessment of the child's overall well-being, measured on a 10-cm VAS (0 = very good; 10 = very poor); Childhood Health Assessment Questionnaire (C-HAQ), Italian version (13), (0 = best; 3 = worst); number of joints with pain upon movement/tenderness; number of swollen joints; number of joints with limited range of motion; number of joints with active arthritis (defined as the number of joints with swelling or, if no swelling was present, with limitation of movement, with either pain upon movement/tenderness); erythrocyte sedimentation rate (ESR; Westergren method); and C-reactive protein (CRP) level (by nephelometry).

In the first years of the study, functional ability was measured by either the Modified Lee Index (14) or the Juvenile Arthritis Functional Assessment Report (JAFAR) (15). In order to standardize scores from all functional ability tools, scores from the JAFAR and the Modified Lee Index were proportionally converted to the 0–3 scale of the C-HAQ. Because we previously observed a very high correlation among the 3 instruments when administered to the same patient on the same day (16), we believe that combining scores from the different instruments for purposes of analysis is valid. The articular indices were assessed in a total of 67 joints (those included in the normal clinical evaluation) and were assessed in each patient by the same investigator (AR or SV) throughout the study.

Radiographic assessment.

Standard radiographs of both wrists in posteroanterior view were made at the baseline, 1-year, and final visits using 3M XDA film and a 3M Trimax T16 cassette (3M Company, St. Paul, MN), with settings of 50 mA, 0.03 seconds, and 43–47 kV. Radiographic damage was scored according to the Poznanski method (9) by 2 independent investigators (FR and FT) who were unaware of the clinical data. This method is based on measurements of the radiometacarpal (RM) width, which is the distance from the base of the third metacarpal bone to the midpoint of the distal growth plate of the radius, and the maximum length of the second metacarpal bone (M2).

Prior to the measurements, the 2 observers were trained in the technique of RM and M2 determination by an experienced radiologist (GB), who supervised all radiographic assessments. In cases in which advanced carpometacarpal erosions made it difficult to define the bone ends, the last film that allowed a reliable assessment was used as the final radiograph. If advanced destruction occurred within the first 2 years of followup, then the patient was excluded from the analysis. Because the Poznanski method cannot be used when there is radiographic closure of the growth plates of the second metacarpus, patients who had apparent radiographic closure of the second metacarpal growth plate within the first 2 years of followup were excluded. In patients who had closure of the growth plate after the second year of followup, the wrist radiograph performed immediately before demonstration of the closure was used as the final radiograph. The RM and M2 measurements were made to the nearest 0.1 mm by using a precision gauge (DialMax; Swiss Precision, Solothurn, Switzerland). The scores assigned by the first investigator (FR) were used for the analyses; the scores assigned by the second investigator (FT) were used to validate the scores of the first investigator.

The measures obtained for RM and M2 were plotted against each other using the normative charts of Poznanski et al (9). For each wrist, the number of standard deviations between the expected and the observed RM width for the measured M2 length was calculated according to the formulas reported by Poznanski et al (9). The RM/M2 score, which represents the carpal length and will be referred to as the Poznanski score, reflects the amount of radiographic damage in the wrist. The greater the negative value of the Poznanski score, the more severe the radiographic damage. In each pair of wrists, the score for the more damaged side was used in the analyses. We made this choice because we believed that separate assessments of the right and left wrists would be less meaningful clinically and that the use of the arithmetic sum or mean of the 2 wrists could affect the reliability of the estimation of the true amount of damage in patients with a marked difference in the severity of radiographic changes between the 2 wrists (i.e., with positive versus negative Poznanski scores).

Radiographic progression during the first year and radiographic progression during the entire study period were then determined by calculating the change in the Poznanski score between the radiographs obtained at baseline and those obtained at 1 year as well as those obtained at the final visit, respectively. Because the duration of followup varied among the study patients, the radiographic change between the baseline and the final radiographs was divided by the number of years of followup for each patient to yield the yearly radiographic progression. A positive value for radiographic progression indicates improvement, whereas a negative value indicates worsening.

To investigate the reproducibility of the radiographic scoring method we used, we assessed the intraclass interreader correlation coefficients by comparing the values obtained for the RM width and the M2 length by the 2 observers for all radiographs. We assessed the intraclass intrareader correlation coefficients by asking 1 of the 2 examiners (FR) to interpret all radiographs for a second time in a blinded manner, 3 months after the previous review. The intraclass interreader and intrareader correlation coefficients for the RM and M2 measurements were very high, ranging from 0.97 to 0.99. The bias estimate and the 95% confidence interval (95% CI) between the 2 observers, calculated according to the method of Bland and Altman (17), for the Poznanski score were −0.041 and −1.72, 1.72, respectively, in the less damaged wrist and −0.071 and −3.72, 3.57, respectively, in the more damaged wrist. The bias estimate and 95% CI between the 2 observers for the progression scores from the first to the last radiograph, calculated according to the same method, were 0.009 and −0.32, 0.34, respectively (Figure 1).

Figure 1.

Progression scores plotted according to the method of Bland and Altman. Broken lines show 95% limits of agreement.

To verify whether our patients developed an abnormal shortening of M2 in comparison with the impairment of the patient's height, which often occurs in JIA, we compared the M2 length in the worse wrist with the reported norms and evaluated the relationship between the shortening of the M2 length and the reduction in height (as compared with the norms) at both the baseline assessment and the final assessment. Furthermore, we evaluated whether a shortening of the M2 length was related to age at disease onset, sex, disease duration, amount of radiographic damage, and rate of radiographic progression. For these purposes, Z scores (units of standard deviation above or below the normal mean for each measurement) were derived for the M2 length (18) and the patient's height (19) by interpolation from published norms.

Statistical analysis.

Comparison of clinical features between patients who were included and those who were excluded from the study was made by the Mann-Whitney U test, in the case of quantitative variables, and by the chi-square test, in the case of qualitative variables. Comparison of radiographic progression in the different time intervals was performed with the nonparametric Friedman's test for repeated measures. The relationships between Z scores for the M2 length and the patient's height and between the M2 length Z score and the continuous variables were evaluated by the correlation method. The relationship between the M2 length Z score and sex was assessed by the Mann-Whitney U test.

Univariate analyses of the relationship between the baseline variables and the 3 study outcomes were undertaken using the Mann-Whitney U test for comparison of 2 groups (e.g., males versus females), analysis of variance (Kruskal-Wallis test) for comparison of more than 2 groups (e.g., the different JIA subtypes), or the correlation method for continuous variables.

Multiple logistic regression was used to find relevant independent prognostic variables for the 3 study outcomes. Before the application of logistic regression procedures, the study outcomes were dichotomized to binary outcomes. The yearly radiographic progression was dichotomized as the absence or presence of progression, with absence being represented by positive values and presence by negative values. Cut points for the final Poznanski score were −2 SD or more (better outcome) and less than −2 SD (poorer outcome). Cut points for the final C-HAQ score were ≤0.5 (better outcome) and >0.5 (poorer outcome) (20). The prognostic variables were selected using univariate analyses (P < 0.05 for any outcome). Although they were not statistically significant in the univariate analyses, the disease duration and the M2 length Z score at baseline were also entered into the multivariate models because they were considered a priori to be of foremost clinical importance for the study outcomes. For each logistic regression model, the sensitivity and specificity in predicting the outcome and the percentage of patients correctly classified were calculated. The Statistica software package (StatSoft, Tulsa, OK) was used for the univariate analyses, and the Stata software package (Stata, College Station, TX) was used for the multivariate analyses.

RESULTS

A total of 110 patients eligible for the study were identified. Sixteen of them were excluded for the following reasons: apparent radiographic closure of the second metacarpal growth plate within the first 2 years of followup (9 patients), lack of clinical data (4 patients), lost radiographs (2 patients), and advanced bone changes making the RM and M2 measurements unreliable within the first 2 years of followup (1 patient). These 16 patients were comparable to the remaining 94 patients with regard to sex ratio, JIA subtype distribution, disease duration at first observation, and time lag between disease onset and start of second-line drug treatment. However, the excluded group had a significantly older age at disease onset, partly reflecting the fact that more than half of them (9 of 16) had already developed radiographic closure of the second metacarpal growth plate during the early phase of the study (data not shown). The 110 potentially eligible patients were part of a group of ∼350 JIA patients who were seen during the study period. The ineligible patients did not have bilateral wrist involvement or lacked the necessary radiographs.

The baseline demographic and clinical characteristics of the 94 patients who were available for radiologic analysis are shown in Table 1. Thirty-five patients had systemic arthritis, 30 had polyarthritis, 24 had extended oligoarthritis, 2 had enthesitis-related arthritis, and 3 had psoriatic arthritis. All disease status measures varied widely along their ranges, but the individual mean values indicated that, on average, the entire patient group had rather active disease at the start of the study. The median duration of followup from baseline to the final radiographic assessment was 4.5 years (range 2–13.5 years). During the study period, all but 5 patients had received 1 or more second-line medications: 87 patients took methotrexate, 24 took cyclosporin A, 13 took sulfasalazine, 3 took etanercept, 3 took cyclophosphamide, 2 took azathioprine, and 2 took hydroxychloroquine.

Table 1. Main clinical features of the 94 patients at baseline
Baseline variableNo. of patientsMedianMinimumMaximum
  • *

    Three patients were rheumatoid factor positive.

  • Range of 0 (best) to 10 (worst).

  • Range of 0 (best) to 3 (worst).

  • §

    Normal <15 mm/hour.

  • Normal <0.3 mg/dl; all values below the threshold were equalized to 0.1 mg/dl.

  • #

    For this study, a Poznanski score below −2 was defined as abnormal.

No. of males/females60/34
Disease subtype    
 Systemic35*
 Extended oligoarticular24*
 Polyarticular30*
 Enthesitis-related arthritis2*
 Psoriatic arthritis3*
Age at onset, years4.60.714.2
Disease duration, years1.106.1
Physician's global assessment of overall disease activity6.30.210
Parent's global assessment of the child's overall well-being2.2010
No. of joints with active disease7048
No. of joints with limited range of motion5036
Childhood Health Assessment Questionnaire score0.403
Erythrocyte sedimentation rate, mm/hour§464139
C-reactive protein, mg/dl2.40.118
Poznanski score#−1.2−4.41.7
Z score for length of the second metacarpal bone−0.3−7.03.0

In the whole cohort, the amount of radiographic damage increased significantly over time (Figure 2). The mean ± SD Poznanski score was −1.2 ± 1.3 at baseline, −1.7 ± 1.8 at the 1-year visit, and −1.9 ± 2.2 at the final visit (P < 0.0001). The mean radiographic progression during the first year (−0.5 ± 1.1) was greater than the mean radiographic progression during the remaining followup period (−0.2 ± 1.3). The mean yearly radiographic progression during the entire followup period was −0.1 ± 0.4. The mean Z scores for M2 length and for height were −0.3 ± 1.5 and −0.2 ± 1.6, respectively, at baseline, and −0.7 ± 1.9 and −1.0 ± 1.8, respectively, at the final visit. The Z scores for M2 length and for height were strongly correlated, both at the baseline visit (r = 0.84, P < 0.0001) and at the final visit (r = 0.82, P < 0.0001), indicating that there was no disproportionate retardation of growth of M2 in comparison with the impairment of height throughout the disease course. The Z score for M2 length correlated with male sex and with disease duration, but not with age at disease onset (data not shown).

Figure 2.

Poznanski scores over the study period. The amount of radiographic damage increased significantly over time (P < 0.0001).

The results obtained by univariate analyses of the relationship between the baseline variables and the yearly radiographic progression, the final Poznanski score, and the final C-HAQ score are presented in Table 2. Radiographic progression in the first year was consistently the strongest correlate for all 3 outcomes. Male sex was the only other variable that was related to all 3 outcomes, whereas the systemic subtype of JIA was related to the yearly radiographic progression and the final Poznanski score, but not the final C-HAQ score. Other significant relationships were those between the baseline and final Poznanski scores, between the baseline and final C-HAQ scores, and between the number of joints with active disease at baseline and the final Poznanski score. Notably, the baseline M2 length Z score was not related to any of the 3 outcomes. The baseline M2 length Z score was not related to the baseline Poznanski score, whereas its final level was related to both the yearly radiographic progression and the final Poznanski score, suggesting that patients with greater radiographic progression developed a more severe retardation of growth of the M2 bone (data not shown).

Table 2. Relationship between the baseline variables and the yearly radiographic progression, the final Poznanski score, and the final score on the C-HAQ
Baseline variableYearly radiographic progressionFinal Poznanski scoreFinal C-HAQ score
rPrPrP
  • *

    The strongest correlations for all 3 outcomes were male sex (versus female sex) and radiographic progression in the first year. Systemic disease subtype (versus all other disease subtypes) was related to the yearly radiographic progression and the final Poznanski score, but not the final score on the Childhood Health Assessment Questionnaire (C-HAQ).

  • Mann-Whitney U test.

  • Analysis of variance.

Male sex*0.00030.0030.005
Systemic disease subtype*0.0020.00090.96
Age at onset0.090.370.0040.970.050.73
Disease duration−0.110.270.010.360.070.62
Physician's global assessment of overall disease activity−0.210.330.0050.980.190.52
Parent's global assessment of the child's overall well-being−0.100.580.300.090.010.67
No. of joints with active disease−0.130.200.220.030.250.06
No. of joints with limited range of motion−0.110.300.160.130.220.11
C-HAQ score−0.160.200.060.640.330.03
Erythrocyte sedimentation rate0.080.450.0550.620.020.89
C-reactive protein level−0.130.270.0350.770.060.68
Poznanski score0.080.470.58<0.00010.200.14
Z score for length of the second metacarpal bone0.180.080.130.230.090.51
Radiographic progression in the first year*0.62<0.00010.59<0.00010.390.003

Table 3 shows the baseline parameters identified by multivariate logistic regression models that had as dependent variables the yearly radiographic progression, the final Poznanski score, and the final C-HAQ score, as well as the sensitivity and specificity of each model in predicting the study outcome and the percentage of patients correctly classified. Radiographic progression in the first year entered the regression in all 3 models, with odds ratios (ORs) of 14.32 for the yearly radiographic progression, 6.49 for the final Poznanski score, and 8.42 for the final C-HAQ score. Female sex was protective against yearly radiographic progression, with an OR of 0.14, and the baseline Poznanski score affected the final Poznanski score, with an OR of 9.28. Other baseline variables, such as JIA subtype, age at disease onset, disease duration, and measures of disease activity and physical disability, had no significant predictive power, nor was the Z score for the M2 length related to radiographic outcome. With regard to disease duration, the distribution was similar in the systemic (median 0.9 years), polyarticular (median 0.8 years), and extended oligoarticular (median 1.6 years) subtypes (P = 0.19), suggesting that the apparent predictive power of early radiographic progression could apply to all JIA subtypes.

Table 3. Logistic regression models of baseline variables as predictors of the 3 study outcomes*
OutcomePredictorOdds ratio95% CIPSensitivitySpecificity% correctly classified
  • *

    95% CI = 95% confidence interval; C-HAQ = Childhood Health Assessment Questionnaire.

Yearly radiographic progression (no/yes) (n = 91)Radiographic progression in the first year (no/yes)14.324.51–45.52<0.00010.920.6683.5
 Sex (female/male)0.140.032–0.600.003   
Final Poznanski score (−2 SD or more/less than −2 SD) (n = 91)Baseline Poznanski score (greater than −2 SD/−2 SD or less)9.282.63–32.740.00010.350.9872.5
Radiographic progression in the first year (no/yes)6.491.89–22.310.0006   
Final C-HAQ score (≤0.5/>0.5) (n = 57)Radiographic progression in the first year (no/yes)8.421.70–41.650.0020.910.4663.2

The results of the logistic regression analyses (which implied dichotomization of study outcomes into binary outcomes) were comparable to those obtained with multiple regression analyses (i.e., handling the outcomes as continuous variables) (data not shown). Notably, the yearly radiographic progression was strongly correlated with the degree of joint damage at the final visit, as measured by the Poznanski score (P < 0.0001), and with the level of physical disability at the final visit, as measured by the C-HAQ score (P = 0.0004) and the number of joints with limited range of motion (P = 0.0003). Among the clinical indicators of disease activity at the final assessment, the yearly radiographic progression was correlated with the number of joints with active disease (P < 0.0001), but not with the physician's or the parent's global assessments, the ESR, or the CRP (Table 4).

Table 4. Relationship between the variables at the final visit and the yearly radiographic progression
Variable at final visitrP
Physician's global assessment of overall disease activity−0.050.72
Parent's global assessment of the child's overall well-being−0.190.20
No. of joints with active disease−0.41<0.0001
No. of joints with limited range of motion−0.370.0003
Childhood Health Assessment Questionnaire score−0.450.0004
Erythrocyte sedimentation rate−0.180.12
C-reactive protein level−0.100.37
Poznanski score0.72<0.0001

DISCUSSION

We investigated the rate of radiographic progression in 94 patients with JIA, all of whom had polyarthritis with symmetric involvement of the wrists. The wrist joint is affected in ∼60% of JIA patients, being second only to the frequency of knee joint involvement (21), but this frequency is much higher if only the polyarticular forms of JIA are considered. The wrist, together with the hip, is the site most vulnerable to changes seen on radiographs in patients with JIA (22, 23). Furthermore, wrist disease has been associated with a more severe course of arthritis (24, 25), a poorer functional outcome (23), and a lower likelihood of a short-term therapeutic response (26). Notably, all but 5 patients had resistant disease, which was judged severe enough to warrant treatment with a second-line drug, typically, methotrexate. Thus, our patient cohort had most of the clinical features that identify the subset of JIA patients who are at risk of joint destruction and poor functional outcome (2).

The potential for deterioration during the disease process was further demonstrated by the fact that patients experienced, on average, a significant increase in radiographic damage over time. Radiographic progression, however, was more pronounced during the first year of observation than during the remaining period of followup. These findings support the view that in polyarticular JIA, radiographic damage is common and occurs early (2). Similarly, radiographic damage has been shown to develop early in the disease course in adults with rheumatoid arthritis (27–29), with the rate of progression decreasing during the subsequent years of disease (30, 31). It must be noted that since the Poznanski score is a measure of cartilage loss, its sensitivity can be reduced in cases of advanced cartilage thinning. We therefore cannot exclude the possibility that a ceiling effect could partly account for the observed reduction in radiographic progression after the first year of followup. However, an increase in the number of patients who improved over the years of study could also have played a role. It is well known that the regenerative capacity of articular cartilage is better in growing children than in adults (10). Furthermore, the majority of the study patients had received second-line therapy with methotrexate, which has been shown to have a “disease-modifying” effect in JIA (10, 11).

Several studies have examined factors that may affect radiographic outcome in JIA. However, the results of these studies are not easily compared with ours because the development of radiographic damage has rarely been assessed by serial radiographs and by a standardized scoring system in JIA. Reported correlates of radiographic joint destruction include early onset, early radiographic abnormalities, presence of IgM rheumatoid factor, presence of HLA–B27, symmetric or more severe arthritis, the duration of active disease, and in patients with systemic disease, persistent systemic symptoms and thrombocytosis at and beyond 6 months (2, 21, 32–34). We found that the radiographic progression in the first year of followup was independently related to both the yearly radiographic progression during the entire study period and the amount of radiographic damage at the final visit. The degree of damage at baseline was also associated with an increase in total damage. Since patients had different disease durations at study entry, however, it was likely that the amount of radiographic damage at baseline would be different. Whether the presence of radiographic damage at baseline reflected a rapid deterioration or a relatively late diagnosis cannot be determined from our data. Nevertheless, it can be concluded that either the presence of early damage or the early development of damage predicts further deterioration. Because there may often be a significant lag time between clinical manifestations and the occurrence of radiographic changes and because cartilage loss may develop earlier than expected in JIA patients, the initial radiographic assessment for the purposes of clinical predictions should be done early in the disease course (i.e., within 6–12 months after initial presentation of symptoms).

Our female patients appeared to be protected against long-term radiographic progression as compared with our male patients. An association between male sex and a worse radiographic outcome has previously been reported (11, 35). However, other studies have found that female sex is the strongest risk factor for long-term disability (23, 33, 36), or they have failed to identify sex as one of the significant predictors of any outcome (20). In our study, the relatively greater likelihood of radiographic progression among the male patients might be partly explained by the fact that there were more male patients in the systemic disease subtype (43%) than in the other disease subtypes (28% in all other groups). Indeed, the systemic subtype (as compared with the other subtypes) was found to be strongly correlated with both long-term radiographic progression and joint damage by univariate analyses, although it did not enter any of the logistic regression models. Among JIA patient cohorts followed up for median intervals of 6–9 years, the frequency of destructive changes has been reported to be higher in the systemic subtype than in the other subtypes, except for the rheumatoid factor–positive polyarticular subtype (23). This latter subtype accounted for only 3 patients in our series.

In the few studies that have examined the relationship between radiographic damage and disability in JIA, significant correlations between joint changes and greater levels of disability, as assessed by the Steinbrocker criteria or the C-HAQ, have been reported (for review, see ref. 23). We found that early progression of damage predicted long-term disability, as assessed by the C-HAQ. This observation was strengthened by the strong correlation between the yearly radiographic progression and both the C-HAQ score and the number of joints with limited range of motion at the final visit. These results indicate that there is a close relationship between structural damage and functional impairment over the course of JIA and suggest that the early progression of damage may help to identify at an early stage those patients who are at greater risk of poor functional outcome.

Among the measures of disease activity at the final visit, neither the subjective assessments made by the physicians or the children's parents nor the ESR or CRP values were related to the yearly radiographic progression. However, the yearly radiographic progression was strongly related to the number of joints with active disease, which suggests that greater articular damage was associated with continued joint inflammation. Studies in adult patients have shown that cumulative joint inflammation increases the risk of progression of joint damage (37). Thus, therapy aimed at suppressing joint inflammation early in the course of JIA may prevent the progression of joint damage.

Our study indicates that the Poznanski method is a reliable and sensitive instrument for assessing the progression of radiographic damage in patients with JIA. It has the advantages of being simple, quick, and reproducible, and it requires little specific training. Thus, it can be easily applied not only in therapeutic trials, but also in clinical settings. Availability of standards for normal carpal length is another advantage in studies of growing children. Furthermore, since the RM width is not dependent on the degree of ossification of the carpal bones, its measurement helps to overcome the problem of advanced skeletal maturation, which occurs frequently in JIA (38).

The Poznanski method has some disadvantages, however; it can be used only in patients with wrist involvement, it is unreliable in cases of advanced carpometacarpal erosions, making it difficult to define the bone ends, and it cannot be used once there is radiographic closure of the growth plates of the second metacarpal bone. Concerning this latter point, Zerin et al (39) reported that in none of their JIA patients with apparent radiographic closure of the growth plates of the second metacarpal bones did this occur prematurely in comparison with the published norms. Another potential problem with the Poznanski method is whether isolated measurement of the wrist represents a good surrogate measure of the severity of erosive joint disease in other joints throughout the body. Using a different scoring system, van Rossum et al (34) found that abnormalities seen on radiographs of the hand did not correlate well with radiographic abnormalities of other joint groups.

Because JIA patients are often small for their age and their bones are correspondingly small, age-related standards are not reliable. In the Poznanski method, therefore, the carpal width (RM) is compared with the length of an adjacent bone (M2) rather than with age. Indeed, it is believed that the RM width is more closely correlated with stature than with age, and stature correlates well with M2 (10). This approach allows a measure of joint size that is relatively independent of the size of the child. Zerin et al (39), however, reported that patients with polyarticular JIA may develop early retardation of the growth of the M2 bone that is disproportionately more severe, as well as more rapid in its progression over time, than the impairment of the patient's height. They also found that this disproportion increases with increasing duration of disease. We did not confirm these findings because in our patients, the Z scores for the M2 length and the height were strongly correlated throughout the disease course, indicating that there was no disproportionate retardation of growth of the M2 bone compared with the patient's height. Our finding of a relationship between growth retardation of the M2 bone and disease duration and radiographic progression is not surprising because it is well known that patients with severe, longstanding polyarticular JIA, particularly of systemic onset, most often experience impaired growth in terms of overall height (and, therefore, in terms of M2 length) (1).

In summary, our patients with polyarticular-course JIA experienced, on average, a significant radiographic progression over time, as measured by the Poznanski score. Radiographic progression was more pronounced during the first year of observation. Changes in the early Poznanski score were consistently predictive of yearly radiographic progression, long-term joint damage, and physical disability. These findings indicate that the Poznanski score is a meaningful outcome measure in JIA and that its measurement early in the disease course can help to identify those patients who are at greater risk of joint destruction and poor functional outcome.

Ancillary