Novel and Effective 28-Joint Disease Activity Measure for Rheumatoid Arthritis: The Disease Activity Score 28-Monocyte Chemotactic Protein-1

Remission is widely considered the primary therapeutic goal in rheumatoid arthritis (RA) management, and this target has increasingly become attainable with the advent of biologic treatments. The feasibility of achieving remission can be evaluated using various criteria, including the 28-joint Disease Activity Score (DAS28) index [1]. However, the 2011 remission criteria established by the American College of Rheumatology/European League Against Rheumatism (ACR/EULAR) [2] were demonstrated to outperform the DAS28-erythrocyte sedimentation rate (ESR) for patient global, physician global, morning stiffness, and functional status assessments [3].

Aletaha and Smolen have argued for abandoning DAS28-based remission criteria [4], citing the potential for residual joint swelling even when DAS28-ESR scores are < 2.0 [5] as well as the possibility of residual tissue and molecular abnormalities indicating active disease as valid reasons for doing so [6]. Nevertheless, we have developed an alternative, clinically useful DAS28-based approach: the DAS28 index supplemented with monocyte chemotactic protein-1 (DAS28-MCP-1) results [7, 8]. MCP-1 is produced locally by activated monocytes and fibroblasts at the site of inflammation. In addition, a MCP-1 antagonist reduces or prevents arthritis in MRL-lpr mice. It implies that MCP-1 is much involved in arthritic inflammation [7]. The four cut-off points of DAS28-MCP-1 for the remission, low, moderate, and high disease activity categories are < 2.2, ≤ 3.6, and ≤ 4.8, respectively [9]. Our findings indicate the high criterion validity of the DAS28-MCP-1, with strong intercorrelations and time–change correlations among various disease activity scores as well as favorable limits of agreement identified through Bland–Altman plots [8]. Furthermore, DAS28-MCP-1 scores were correlated with Health Assessment Questionnaire-Disability Index (HAQ-DI) scores, similarly to all other disease activity scores [8].

Swollen joint count (SJC) has been demonstrated to be more strongly correlated with the development of bone erosions than ESR and CRP in patients with RA [10]. The committee that established the 2011 ACR/EULAR remission criteria for RA reported that 10% of patients with a DAS28-ESR score of < 2.6 had ≥ 4 swollen joints and that 1 patient had > 20 swollen joints [11]. Similarly, our findings revealed that 5.2% of the included patients with RA with a DAS28-ESR score of < 2.6 had ≥ 4 swollen joints, with one patient presenting with 10 swollen joints [12]. By contrast, none of the patients with RA with a DAS28-CRP score of < 2.5 or a DAS28-MCP-1 score of < 2.2 had ≥ 4 swollen joints [12].

The relationship between MCP-1 levels in blood and SJC has been investigated in both earlier and more recent studies (Table 1). A 2001 study revealed that plasma MCP-1 levels, but not ESR or CRP levels, were significantly correlated with SJC [13]. In a 2021 study, serum MCP-1 levels were reported to be significantly correlated with SJC [14]. In two recent patient cohorts, plasma or serum MCP-1 levels were significantly correlated with SJC, whereas ESR or CRP exhibited significant correlations in only one of the two patient cohorts [12]. These findings suggest that MCP-1 levels provide a more accurate indication of SJC and are potentially more closely associated with bone erosion in RA than ESR and CRP are.

MCP-1 demonstrated a strong correlation with DAS28-ESR, DAS28-CRP, and DAS28-MCP-1 scores in the 2013 cohort, surpassing the correlation observed with ESR and CRP [12]. Additionally, MCP-1 exhibited a strong correlation with SDAI, CDAI, and DAS28-MCP-1 scores in the 2020 cohort, with a performance comparable to that of CRP and stronger than that of ESR [12].

We also observed that a significantly higher proportion of patients with RA (ranging from 57% to 92% differences) who met the DAS28-ESR remission criteria (< 2.6) had residual swollen joints than that of those who met the remission criteria according to the other disease activity scores [12]. Similarly, residual swollen joints were more frequently reported (with differences ranging from 71% to 94%) during patient visits when patients had achieved remission according to the DAS28-ESR criteria (< 2.6) than when patients had achieved remission according to the other four disease activity scores [12].

As indicated in Ref. [12], we examined medication changes by evaluating the patients with RA who achieved remission according to various disease activity scores at baseline (Month 0, Month 3, or Month 6). Medication use at baseline was then compared with use 6 months later to determine whether medication doses had increased by ≥ 30% or whether new medications had been prescribed. No significant difference in the percentage of patients requiring increased medication doses or new medications was observed between the patients with RA who did not achieve remission according to the 2005 modified ARA definition of remission [15] and those who did among the patients with RA who met the DAS28-ESR, DAS28-MCP-1, DAS28-CRP, SDAI, or CDAI remission criteria (Table 2) [12]. However, among the patients who met the DAS28-ESR remission criteria, significant differences in medication changes were observed between those who also met the 2011 ACR/EULAR remission criteria and those who did not (Table 2) [12]. By contrast, no significant medication change was noted among the patients with RA who met the DAS28-MCP-1, SDAI, or CDAI remission criteria, regardless of whether they met the 2011 ACR/EULAR remission criteria (Table 2) [12]. The mechanism underlying this finding warrants further investigation.

Finally, the DAS28-MCP-1, DAS28-CRP, SDAI, and CDAI criteria significantly outperformed the DAS28-ESR criteria, with results that aligned more closely with the 2011 ACR/EULAR remission criteria (Table 2) [12]. Furthermore, compared with the DAS28-ESR, SDAI, and CDAI criteria, the DAS28-MCP-1 criteria yielded results that aligned significantly more closely with the 2005 modified ARA remission criteria (Table 2) [12].

In the context of RA treatment, remission is generally the primary goal of a treat-to-target strategy [16], and effective management of RA can lead to substantial improvements in health-related quality of life (HRQoL) [17]. Moreover, a study assessing 17 remission definitions revealed that the proportion of patients achieving an HAQ score of ≤ 0.5 was the highest among the patients who achieved Boolean-defined remission (88.8%) and lowest among those who achieved DAS28-ESR-defined remission (67.5%) [18]. These findings suggest that DAS28-ESR definitions of remission are inadequate for accurately assessing the functional status of patients with RA.

Notably, the correlation between immunoregulatory cells or cytokines and DAS28-MCP-1 scores differed markedly from that observed with other disease activity measures in patients with RA [9]. For example, the number of M2 macrophages was lower in the DAS28-MCP-1 < 2.2 subgroup than in the DAS28-MCP-1 ≥ 2.2 subgroup; this pattern was not observed when remission DAS28-ESR scores (< 2.6), DAS28-CRP scores (< 2.5), and SDAI scores (≤ 3.3) and their corresponding nonremission scores were compared [9]. This finding supports the use of DAS28-MCP-1 < 2.2 for fulfilling the 2005 modified ARA remission criteria [8], and a study demonstrated that among the 10 cytokines and 3 cell types it examined, only the number of M2 macrophages significantly differed between the remission (lower numbers) and nonremission (higher numbers) subgroups [9]. Moreover, sTNF-R1 was positively and significantly correlated with DAS28-MCP-1 scores, although with a low correlation coefficient [9]. This result further supports the use of DAS28-MCP-1 < 2.2 for fulfilling the 2011 ACR/EULAR remission criteria [8], with recent results indicating that sTNF-R1 levels were significantly lower in a 2011 ACR/EULAR remission subgroup than in a 2011 ACR/EULAR nonremission subgroup [9]. Therefore, a lower number of M2 macrophages and reduced sTNF-R1 levels may link remission DAS28-MCP-1 scores to the 2005 modified ARA and 2011 ACR/EULAR remission criteria, respectively.

The statement on the number of anti-inflammatory M2 macrophages seems to be contradictory to the concept that M2 macrophages are anti-inflammatory in action. However, the plasma MCP-1 levels in the DAS28-MCP-1 < 2.2 remission subgroup (78.92 pg/mL [57.32, 131.41], median [25%, 75% percentiles]) are correspondently lower than those in the DAS28-MCP-1 ≧ 2.2 nonremission subgroup (105.50 pg/mL [78.08, 148.60], median [25%, 75% percentiles]; p = 0.003) [9]. This latter result is compatible with the lower number of M2 macrophages in the DAS28-MCP-1 < 2.2 remission subgroup, considering that macrophages are one of the major sources of MCP-1 production [13]. Moreover, the result showing that the median number of M2 macrophages in the 2005 modified ARA remission subgroup (1850/mL) is lower than the median number of M2 macrophages in the 2005 modified ARA nonremission subgroup (3200/mL; p = 0.021, by Mann–Whitney U test) also reinforces such a trend [9]. Likewise, the temporal change of plasma sTNF-R1 levels at Month 0 (M0), M6, and M12 is exactly the same as that of M2 macrophages [9]. Nevertheless, because no such results have been reported yet in the literature, it has to wait for further study reports from different researchers to solve such a conflict.

Furthermore, plasma sTNF-R1 levels in the DAS28-MCP-1 < 2.2 remission subgroup (482.01 ± 207.31 pg/mL, mean ± SD) are slightly lower than those in the DAS28-MCP-1 ≧ 2.2 nonremission subgroup (491.33 ± 176.29 pg/mL), but not significantly different (p = 0.772) [9]. The sTNF-R1 result does not contradict (though not support) the statement on the number of M2 macrophages. Yet, plasma sTNF-R1 levels correlated significantly and positively with DAS28-MCP-1 scores, indicating that low plasma sTNF-R1 levels exist with low DAS28-MCP-1 scores [9]. Additionally, plasma sTNF-R1 levels are significantly lower in the 2011 ACR/EULAR remission subgroup than those in the 2011 ACR/EULAR nonremission subgroup [9]. Hence, this result further supports that low plasma sTNF-R1 levels match with low DAS28-MCP-1 scores [9]. These results are similar to the report showing that lower sTNF-R1 levels were found in the RA patients remaining in remission than those having disease flare after biologics were discontinued [19]. In particular, sTNF-R1 has been shown to induce monocyte apoptosis besides acting through TNF-α binding, which surely makes such a relationship between macrophages/sTNF-R1 and DAS28-MCP-1 scores more complicated [20].

A study reported lower plasma IL-5 levels in a DAS28-ESR < 2.6 subgroup than in a DAS28-ESR ≥ 2.6 subgroup [9]. The lower IL-5 levels in the DAS28-ESR < 2.6 subgroup may support the fulfillment of the 2011 ACR/EULAR remission criteria; significantly lower IL-5 levels were observed in the 2011 ACR/EULAR remission subgroup than in the nonremission subgroup [9]. These findings suggest that distinct immuno-inflammatory abnormalities persist across different DAS28-based categories, particularly in relation to the fulfillment of the 2005 modified ARA and 2011 ACR/EULAR remission criteria.

Those authors have done receiver-operating-characteristics curve analysis for changes (Month 3 − Month 0 [M3 − M0], M6 − M0, M9 − M0, and M12 − M0) in DAS28 and SDAI scores against changes (M12 − M0) in the bone erosion of both hand and feet x-ray films [8]. All area under the curves (AUCs) were low (from 0.422 to 0.557), that is, having poor diagnostic accuracy. Moreover, all DAS28 and SDAI scores against modified total Sharp score offered all AUCs less than 0.5 [8]. Therefore, a future study with a longer follow-up time for roentgenographic changes is much needed.

The only limitation of those studies on establishing the DAS28-MCP-1 formula [7-9, 12] is that no criteria have yet been established to assess the treatment response in patients with RA. This should be the final step in fully validating the DAS28-MCP-1 formula.

Nevertheless, the newly devised DAS28-MCP-1 formula has advantages over DAS28-ESR in terms of fulfilling two different remission definitions [8]. Low DAS28-MCP-1 scores correspond with low sTNF-R1 levels that indicate 2011 ACR/EULAR remission [9], which can furnish a basis for further research of sTNF-R1 behind the DAS28-MCP-1 formula. Such a connection with an immunoregulatory cytokines has not been seen in DAS28-CRP and SDAI scores' linkage [9]. Moreover, that DAS28-MCP-1 scores < 2.2 in remission fulfilled 2005 modified ARA remission is significantly better than that of SDAI scores ≦ 3.3 in remission [8]. Therefore, the DAS28-MCP-1 formula provide different aspects of advantages over DAS28-CRP or SDAI formulae. Besides, “The cost of MCP-1 assay (by ELISA assay, R&D systems, Minneapolis, MN, USA) is 0.58 times lower than the cost of (high sensitivity) CRP examination by turbimetry” has been stated earlier [8]. With these advantages of DAS28-MCP1 over DAS28-ESR, DAS28-CRP, and SDAI in mind, it is worth and optimistic for the development of a rapid and high-throughput method for detecting MCP-1 in the future. That is, demand drives development.

In summary, the results obtained through the proposed DAS28-MCP-1 formula are consistent with other commonly used disease activity measure scores. Specifically, the DAS28-MCP-1 is consistent with the 2011 ACR/EULAR remission criteria, which are also compatible with SDAI, CDAI, and DAS28-CRP scores. In this context, the DAS28-MCP-1 significantly outperformed the DAS28-ESR. Notably, both DAS28-MCP-1 and DAS28-CRP scores are consistent with the 2005 modified ARA definition of remission—the most stringent definition of RA remission—and significantly surpassed DAS28-ESR, SDAI, and CDAI scores. Distinct immuno-inflammatory abnormalities, particularly in terms of immunoregulatory cells and cytokines, were observed across different DAS28-formula categories, particularly for the fulfillment of the 2005 modified ARA and 2011 ACR/EULAR remission criteria. Therefore, the DAS28-MCP-1 score warrants considerable attention for use in the evaluation of disease activity among patients with RA.

The author takes full responsibility for this article.

The author declares no conflicts of interest.