Background/Purpose

The induction of remission is the primary target of RA therapy. Failing to achieve the patient global estimate of disease activity criterion (PGA<=1cm) has been shown to be the primary cause hindering patients to be classified as being in remission. Here, we aimed to determine factors that predict achieving the PGA criterion of remission in RA patients in usual clinical care.

Methods

We selected patients from a longitudinal RA database, who started MTX monotherapy or TNFi treatment in combination with MTX or leflunomide and had at least 6 months of follow-up. In univariate analysis, we tested core-set variables to identify candidate predictors of patient perceived remission (PGA≤1cm), which we then subjected to multivariate logistic regression analysis for the outcome: PGA≤1cm.

Results

Data of 172 patients receiving MTX (82% female, 65% rheumatoid factor positive, mean SDAI: 18.5±12.3) and of 112 patients on TNFi (82% female, 62% rheumatoid factor positive, mean SDAI: 19.7±13.3) were used for analysis. After 6 month of treatment 70% of those receiving MTX and 55% of the TNFi group were in a state of remission or low disease activity, based on the SDAI. Forty-seven percent of MTX and 34% of TNFi patients evaluated their PGA as being ≤1cm. Seventy-four percent of MTX and 58% of TNFi treated patients who had PGA≤1cm had also an evaluator global assessment of ≤1cm. The overlap of patient perceived remission and remission by SDAI was 49% in the MTX group and 46% in the TNFi group.

Univariate analyses in MTX treated patients showed an association of pain, HAQ scores, and TJC with the outcome PGA≤1cm. In the multivariate logistic regression model, the odds ratio (OR) was 0.46 for baseline HAQ (Table) and 0.98 for pain as predictors for achieving remission after 6 months of treatment. Higher HAQ and pain scores therefore lead to a lower odds for achieving PGA ≤1cm.

Patients with an improvement in HAQ within the first 3 month have a 3.5 (CI: 1-13) times higher odds to achieve PGA≤1cm after 6 months of treatment than patients who report a worsening in HAQ, which corresponds to a probability of 23% versus 8% to achieve PGA≤1cm.

In patients receiving TNFi-therapy, baseline HAQ (OR: 0.31) and pain (OR: 0.98) were shown to be predictive, i.e. a patient with a baseline HAQ of 1.25 or a pain score of 50mm has a 20% probability of achieving PGA remission, whereas a baseline HAQ of 0.5 or a pain score of 20mm coincides with a 40% probability. 

Table: Summary of the multivariate logistic regression model for the outcome PGA≤1cm

Conclusion

We demonstrated here that patients with poor function or high pain levels are likely to fail patient reported remission, as shown here for the patient global variable, which is part of the established remission criteria. These findings were independent of the treatment regimen. Improvement in function enhances the chances for achieving patient perceived remission after 6 months of DMARD treatment.