Patient Baseline Characteristics

The total patient population for the RA-BE-REAL study is n = 1074 from five European countries. However, in this analysis, a total of n = 673 patients in remission/LDA (n = 364) and those who did not achieve remission/LDA (n = 309) with no missing CDAI data at 3 months were included. The mean age and gender were similar across all three cohorts and irrespective of the patients’ remission/LDA status at 3 months. At baseline, patients in cohort B-TNFi and in remission/LDA had a shorter disease duration, while cohort A-Baricitinib and cohort B-OMA patients were similar. Cohort B-TNFi patients were more often naïve to b/tsDMARDs. In the group that did not achieve remission/LDA, the disease duration was also shorter for cohort B-TNFi and longer in cohort B-OMA.

At baseline, the patients in remission/LDA had a numerically lower mean CDAI score (cohort A 20.1 versus cohort B-TNF 19.7 and cohort B-OMA 18.3) than those who did not achieve remission/LDA (cohort A- baricitinib 28.5, cohort B-TNFi 27.1 and cohort B-OMA 27.3). Pain > 20 mm at baseline was present in 87% of cohort A-baricitinib, 86% of cohort B-TNFi and 84% of cohort B-OMA patients who achieved remission/LDA at 3 months. In the subgroup that did not achieve remission/LDA, pain > 20 mm at baseline was numerically more frequent: cohort A-baricitinib (97%), cohort B-TNFi (90%) and cohort B-OMA (95%) (Table 1).

Table 1 Patient demographics and disease characteristics at baseline by remission or LDA status at 3 monthsPain Assessments—Pain VAS Scores ≤ 10 mm, ≤ 20 mm and > 20 mm

Amongst the patients in remission/LDA, 40.5% in cohort A-Baricitinib, 41.4% in cohort B-TNFi and 33.3% in cohort B-OMA reported no pain (≤ 10 mm) at 3 months. In the subgroup that achieved remission/LDA, a noticeable numerical decrease in the percentage of patients in all cohorts reporting pain VAS ≤ 10 mm was observed over time up to 24 months (Table 2).

Table 2 Percentage of patients who met the thresholds of remaining pain (VAS) over 24 months

Of those patients who did not achieve remission/LDA, only 2.3%, 4.9% and 1.3% in cohort A, cohort B-TNFi and cohort B-OMA, respectively, reported pain VAS ≤ 10 mm at 3 months.

At 3 months, pain VAS scores of > 20 mm were observed by 37.9% in cohort A-baricitinib, 45.0% in cohort B-TNFi and 44.4% in cohort B-OMA for the patients in remission/LDA and in those who did not achieve remission/LDA, 91.5% in cohort A, 91.3% in B-TNFi and 94.7% in cohort B-OMA. At 24 months, pain VAS scores of > 20 mm were less frequent and decreased approximately equally for all cohorts (A: 34.7%, B-TNFi: 39.6%, B-OMA: 30.2%) in remission/LDA. The proportion of patients who did not achieve remission/LDA and reported pain VAS scores of > 20 mm at 24 months was 50.8% (cohort A), 56.3% (cohort B-TNFi) and 57.9% (cohort B-OMA) (Table 2).

Pain Threshold Improvements ≥ 70, 50 and 30%Pain Threshold Improvements ≥ 70%

When analyzed by thresholds of improvement from baseline, almost half (45.8%) of the cohort A-Baricitinib patients in remission/LDA reported a ≥ 70% pain improvement by 3 months, while 41.4% and 39.7% of patients in cohort B-TNFi and cohort B-OMA, respectively, also observed pain improvement ≥ 70%. In the group that did not achieve remission/LDA, 6.9% and 6.8% in cohort A-Baricitinib and cohort B-TNFi, respectively, and 2.6% (cohort B-OMA) reported ≥ 70% pain improvement. From 6 to 24 months, the proportion of patients reporting a ≥ 70% pain improvement remained relatively stable for all cohorts in both subgroups (Fig. 1).

Fig. 1

Percentage of patients who achieved improvements in pain VAS measurements ≥ 30%, ≥ 50% and ≥ 70% at 24 months. Patients who achieved remission/LDA are represented in graphs A, C and E and those who did not achieve remission/LDA represented in graphs B, D and F. BARI baricitinib, LDA low disease activity, OMA other mode of action, TNFi tumour necrosis factor inhibitor, VAS visual analogue scale

Pain Threshold Improvements ≥ 50%

Additionally, for the group that achieved remission/LDA, more patients in cohort A-baricitinib achieved ≥ 50% pain improvement (66.8%) at 3 months. This level was achieved by 56.8% in cohort B-TNFi and 61.9% in cohort B-OMA. For the group that did not achieve remission/LDA, 17.7% in cohort A, 20.4% in cohort B-TNFi and 14.5% in cohort B-OMA reported ≥ 50% pain improvement. The percentage of patients achieving ≥ 50% pain improvement at 24 months were 38.6% (cohort A), 36.0% (cohort B-TNFi) and 20.6% (cohort B-OMA) in the group that achieved remission/LDA and 19.2% (cohort A), 21.4% (cohort B-TNFi) and 10.5% (cohort B-OMA) in the group that did not achieve remission/LDA.

Pain Threshold Improvements ≥ 30%

At month 3, 80.5% in cohort A, 71.2% in cohort B-TNFi and 79.4% in cohort B-OMA reported ≥ 30% pain improvement for patients who achieved remission/LDA. Among those who did not achieve remission/LDA, 33.8% in cohort A, 31.1% in cohort B-TFNi and 32.9% in cohort B-OMA reported ≥ 30% pain improvement. At month 24, the percentage of patients who achieved remission/LDA and reported ≥ 30% pain improvement was similar across cohorts, with cohort A observing 46.8% improvement, 45.9% for cohort B-TNFi and 31.7% for cohort B-OMA. For patients who did not achieve remission/LDA, 28.5% (cohort A), 33.0% (cohort B-TNFi) and 22.4% (cohort B-OMA) achieved ≥ 30% pain improvement. Of note, for those patients who attained remission/LDA, more patients treated with baricitinib (cohort A) achieved numerically greater pain improvements across all pain thresholds at each time point (Fig. 1).

Change From Baseline in Pain VAS Measurements

The mean change from baseline of pain measurement at 3 months was numerically greater for cohort A (− 32.6 mm) than for cohort B-TNFi and cohort B-OMA (− 27.3 mm and − 28.0 mm, respectively). At 24 months, the observed mean changes from baseline measurements were − 26.2 mm (cohort A), − 20.8 mm (cohort B-TNFi) and − 16.0 mm (cohort B-OMA). Patients who did not achieve remission/LDA reported a mean change from baseline of − 9.6 mm (cohort A) and − 8.6 mm (cohort B-TNFi and cohort B-OMA) at 3 months (Fig. 2). Patients treated with baricitinib observed early improvements in pain regardless of disease activity response, and further improvements were observed for those who achieved remission/LDA. For those who received TNFi, pain was quickly controlled, but no further improvement was observed for those who achieved remission/LDA versus those who did not achieve remission/LDA. Lastly, patients who received OMA demonstrated better pain control over time for those patients who achieved remission/LDA.

Fig. 2

Mean change from baseline in pain VAS measurements for A patients who achieved remission/LDA and B patients who did not achieve remission up to 24 months. BARI baricitinib, LDA low disease activity, OMA other mode of action, TNFi tumour necrosis factor inhibitor, VAS visual analogue scale

Change in CDAI Score

The mean change from baseline in CDAI scores for patients who achieved remission/LDA at 3 months was − 15.7 for cohort A-baricitinib, − 15.1 for cohort B-TNFi and − 13.2 for cohort B-OMA. These scores remained relatively stable up until 24 months. For those patients who did not achieve remission/LDA, the mean change from baseline CDAI scores reported were − 8.6 (cohort A), − 5.0 (cohort B-TNFi) and − 7.8 (cohort B-OMA). At 24 months, the mean change in CDAI scores were − 15.8 (cohort A), − 12.5 (cohort B-TNFi) and − 11.3 (cohort B-OMA); it took longer for patients who did not achieve remission/LDA to see a noticeable change in CDAI scores compared with the subgroup that achieved remission/LDA. Patients in the subgroup that achieved remission/LDA observed numerically greater changes in CDAI scores at 3 months compared with the patients who did not achieve remission/LDA; however, a similar CDAI score was observed by both those who achieved remission and those who did not achieve remission at 24 months. For both subgroups, patients in cohort A observed the greatest numerical decrease in CDAI scores (Fig. 3A and B).

Fig. 3

Mean change in CDAI scores for all three cohorts. A Patients who achieved remission/LDA and B patients who did not achieve remission up to 24 months. BARI baricitinib, CDAI Clinical Disease Activity Index, LDA low disease activity, OMA other mode of action, TNFi tumour necrosis factor inhibitor

Residual Pain

At 3 months, the proportion of patients with residual pain (VAS ≤ 20 mm) in remission/LDA was 62.1% for cohort A-baricitinib, 55.0% for cohort B-TNFi and 55.6% for cohort B-OMA. For those who did not achieve remission/LDA, residual pain (VAS ≤ 20 mm) was observed in 8.5% and 8.7% of patients in cohort A and cohort B-TNFi, respectively, and 5.3% in cohort B-OMA (Table 2). At 24 months, in the group that achieved remission/LDA, less residual pain was experienced by patients in cohort A (34.7%), cohort B-OMA (38.7%) and cohort B-TNFi (27.0%). In the group that did not achieve remission/LDA, an increase in residual pain was observed by cohort B (OMA: 6.6% and TNFi: 16.5%) and 9.2%, reported by cohort A, at 24 months (Table 2).

Associating Factors with Pain Improvement

Based on the baseline characteristics of patients, a logistic regression model was used to determine the factors that predicted achieving ≥ 30% pain improvement at 3 months. Patients treated with baricitinib (cohort A) were significantly more likely to achieve greater pain improvement compared with those treated with TNFi or OMA (cohort B). Patients in cohort A were 71% more likely to observe pain improvement compared with patients in cohort B-TNFi and B-OMA (p = 0.001). Patients who were previously treated with b/tsDMARDs treatments were also more likely to observe greater pain improvement compared with those who were naïve to b/tsDMARDs, with patients who received > 2 b/tsDMARDs being most likely to experience pain improvement (p = < 0.001). Other factors associated with achievement of pain improvement were body mass index and, inversely associated, swollen joint count (Table 3). For every 1 unit increase in BMI, the odds of achieving ≥ 30% pain improvement at 3 months increases by 4% when all other factors are constant, indicating a relatively small positive association between BMI and achieving ≥ 30% pain improvement at 3 months. Similarly, for every 1 unit decrease in swollen joint count, the odds of achieving ≥ 30% pain improvement decreases by 4%, indicating small negative association between swollen joint count and achieving ≥ 30% pain improvement at 3 months.

Table 3 Factors associated with achievement of ≥ 30% pain improvement at 3 months