In this real-life cohort, we show the potential of blood MRP8/14 levels to predict the disease course of JIA for 1.5 years after diagnosis. In addition, we showed for the first time that MRP8/14 measured in plasma might be superior to that measured in serum when assessing disease activity in newly diagnosed JIA patients.
To predict the disease course, we demonstrated that blood MRP8/14 levels in newly-diagnosed DMARD-naive JIA patients were associated with the need for medication later. One third of our JIA patients were without DMARDs one year after the diagnosis, which is in line with the proportion of patients with the less aggressive oligoarticular JIA subtype in this study. Patients who were using sDMARDs at one year had higher blood MRP8/14 levels at the baseline compared to non-medicated patients, reflecting higher levels of inflammation. MRP8/14 levels in patients who used bDMARDs were higher than levels in non-medicated patients, but the difference was not statistically significant. This could be explained by the small number of the bDMARD users.
The potential role of MRP8/14 in prognostic evaluation emerged when comparing the blood MRP8/14 levels in patients who achieved clinically inactive disease status while off medication for at least six months during the 1.5-year follow-up period and those who started MTX treatment, regardless of treatment response. Lower levels were associated with sustainable inactive disease without systemic medication. The MRP8/14 biomarkers and their relationships to outcome within the first year after JIA diagnosis were also studied in the German Inception Cohort of Newly Diagnosed Patients with JIA (ICON-JIA). In contrast with our results, they found no association between MRP8/14 levels measured at baseline and disease activity (cJADAS ≤ 1 or active joint count < 1) at 12 months (17). However, their cohort differed from ours in that a large proportion of their patients (156 of 212) were treated with sDMARDs, and their cohort consisted of patients with diagnosed with JIA recently (less than 12 months before inclusion). They also divided the material more roughly than we did, as they included only two groups: active and inactive disease. In our study, disease activity was assessed more accurately, using a continuous variable, PGA. Our observation of the role of MRP8/14 as a prognostic tool is notable because of the importance of attempting to identify patients who may not need systemic medication to avoid exposing them to potential treatment side effects, as well as identifying patients who require urgent treatment interventions.
We were unable to replicate the previously observed associations in JIA patients between increased pre-treatment MRP8/14 levels and good responses to systemic drug therapy (16) (12, 22). Our results are in line with Barendregt et al.’s recent study, in which the researchers found no difference in baseline MRP8/14 levels between JIA patients who responded to treatment and those who did not (19). In our JIA population, more than half of the patients started MTX therapy within eight months of diagnosis. Thus, the duration of the sampling to the start of medication varied between the patients, and inflammatory activity might have changed during that period. This could be one explanation. Another might be that in our real-life cohort, we set the endpoint as remission on MTX, which is a substantial target to attain. Our study population might also reflect real-life situations more than the earlier studies did because we included all the consecutive patients with new-onset arthritis from a single hospital district area, including patients with very low disease activity. The preliminary core criteria for pediatric arthritis used in many other studies (12, 16, 22) are insufficient in practice when achieving long-term remission should be the goal.
We analyzed associations between the PGA and laboratory markers in the univariate model and found relation between the PGA and P-MRP8/14, S-MRP8/14, CRP and ESR (Table 2). Recent studies have also shown an association between MRP8/14 and disease activity markers such as ESR (12, 23) and CRP (16), but only a weak correlation with a physician’s visual analog scale (VAS) (16) or a weak to no correlation with the number of active joints (12, 16). The association with PGA and calprotectin was obtained in a cohort of rheumatoid arthritis patients in which calprotectin correlated more strongly with the PGA than with other clinical parameters (i.e., swollen or tender joints, patient’s global VAS) and was more strongly associated with calprotectin in plasma than in serum (20). We investigated the effect of combining laboratory markers and S-MRP8/14 or P-MRP8/14 on the PGA using multivariate linear regression. We also demonstrated that a more effective set was to combine the laboratory markers with P-MRP8/14 rather than S-MRP8/14. When combining CRP and P-MRP8/14, it seemed that P-MRP8/14 might be a better marker of disease activity than CRP in newly-onset JIA (Table 2). In S-MRP8/14, that kind of behavior was not observed. This highlights the utility of P-MRP8/14 as a diagnostic tool for disease activity in patients with newly-onset JIA. Moreover, in a recent study, La et al. found that MRP8/14 has more specificity than CRP does as a diagnostic tool and marker of disease activity for JIA (24).
Clinicians assess disease activity and evaluate patients’ condition at each visit to the rheumatology clinic. Laboratory parameters constitute one evaluation tool, but they generally do not work well for overall assessment. In our study, we assessed the patients’ overall clinical condition using the PGA as assessed on a VAS. The PGA is a general assessment of overall disease activity that can be performed easily in everyday practice. It involves the subjective opinion of a clinician, and it does not require any knowledge of scoring methods. PGA estimation has been demonstrated to be a more responsive outcome measure in children with JIA than other variables used in clinical trials of JIA (25, 26). Although there is no accurate score for this parameter, Falcone et al. (27) established a good inter-observer agreement on the PGA with a wide spectrum of disease activity and severity among JIA patients. In our cohort, we tested inter-observer agreement by defining the PGA while reading the patients’ medical records. The estimation of disease activity was quite similar between the physicians in our unit (PK and PV). A substantial limitation in our analyses was that we studied the association of MRP8/14 only with the PGA and not also with the Juvenile Arthritis Disease Activity Score (JADAS) (28). This was the case because the patient/parent assessment of global well-being at the time of diagnosis was only available for 60% of JIA patients.
Most recent JIA studies focusing on MRP8/14 were assayed using serum samples (7–9, 11–14, 16). Considerably fewer studies used plasma samples (10, 22, 29, 30). To our knowledge, this population-based study is the first to compare MRP8/14 between serum and plasma samples in JIA. Our finding regarding the superiority of assaying of P-MRP8/14 in the cohort of JIA patients is in line with Nordal et al.’s findings regarding adult rheumatoid arthritis patients (20). As in that study, we also identified lower concentrations of P-MRP8/14 than S-MRP8/14. Nordal et al. assumed that this might be due to the increased in vitro release of MRP8/14 from activated neutrophils during the handling of blood for serum sampling. This can also lead to incorrectly high levels of S-MRP8/14 in patients with mild disease. This observation would partly explain the superiority of plasma samples compared to serum samples, as the former are more stable during handling, and the risks of artifacts are lower.
Our prospective real-life study of new-onset treatment-naïve JIA patients demonstrates the potential of MRP8/14 in the prediction of the disease course. When we have the means to identify patients with an aggressive disease course, even at the time of diagnosis, we can quickly target them with medical interventions probably avoid subsequent consequences. In addition, patients with a mild disease course can avoid exposure to potentially harmful medication side effects.