Results

A total of 60 patients with RA were included in the study. The baseline demographic, clinical, and therapeutic characteristics of the study population are summarized in Table 1.

Variables | N = 60
Age (years)1 | 52.6 ± 11.6
Female3 | 58 (96.7)
Disease duration (years)2 | 15.5 (11.0-24.0)
Anti-CCP positivity 3 | 53 (88.3)
Erosive disease3 | 52 (86.7)
Hypertension3 | 10 (16.7)
Diabetes mellitus3 | 22 (36.7)
Fibromyalgia3 | 21 (35.0)
Osteoporosis3 | 41 (68.3)
HAQ2 | 1 (0.8-1.5)
Previous biologic therapy3 | 23 (38.3)
Concomitant csDMARDs
None3 | 25 (41.7)
Methotrexate3 | 18 (30.0)
Leflunomide3 | 7 (11.7)
Sulfasalazine3 | 10 (16.7)
Baseline disease activity
DAS28-ESR1 | 5.45 ± 0.84
DAS28-CRP1 | 5.08 ± 0.90
SDAI1 | 30.53 ± 12.16
ESR (mm/h)2 | 36.0 (21.5-64.3)
CRP (mg/L)2 | 20.9 (11.7-33.9)
Pain VAS1 | 6.22 ± 1.15
Current biologic therapy
Rituximab3 | 29 (48.3)
Anti-TNF3 | 31 (51.7)
Type of anti-TNF agent (n = 31)
Etanercept3 | 12 (38.7)
Infliximab3 | 6 (19.4)
Certolizumab pegol3 | 7 (22.6)
Golimumab3 | 6 (19.4)

The mean age was 52.6 ± 11.6 years, and the cohort was predominantly female (58 patients, 96.7%). Patients had long-standing disease, with a median disease duration of 15.5 years (11.0-24.0). Anti-CCP antibodies were positive in 53 patients (88.3%), and erosive disease was present in 52 patients (86.7%).

The most common comorbidities were osteoporosis (41 patients, 68.3%), diabetes mellitus (22 patients, 36.7%), and fibromyalgia (21 patients, 35.0%), whereas hypertension was reported in 10 patients (16.7%). The median Health Assessment Questionnaire (HAQ) score was 1.0 (0.8-1.5).

Regarding concomitant treatment, 35 patients (58.3%) received a csDMARD, most commonly methotrexate (18 patients, 30.0%), whereas 25 patients (41.7%) received biologic therapy (rituximab or a TNF inhibitor) without concomitant csDMARDs.

At M0, disease activity was high, with a mean DAS28-ESR of 5.45 ± 0.84, a mean DAS28-CRP of 5.08 ± 0.90, and a mean SDAI of 30.53 ± 12.16. The median ESR and CRP values were 36.0 mm/h (21.5-64.3) and 20.9 mg/L (11.7-33.9), respectively. The mean pain visual analog scale (VAS) score was 6.22 ± 1.15.

Previous exposure to biologic therapy was documented in 23 patients (38.3%). At biologic therapy initiation, 29 patients (48.3%) received rituximab, whereas 31 patients (51.7%) initiated anti-TNF therapy. Among patients receiving anti-TNF therapy, etanercept was the most frequently prescribed agent (12 patients, 38.7%), followed by certolizumab pegol (7 patients, 22.6%), infliximab (six patients, 19.4%), and golimumab (six patients, 19.4%).

Evolution of GC doses over time

The evolution of GC doses during follow-up is presented in Table 2. A progressive reduction in the mean daily GC dose was observed throughout the six-month follow-up period. The mean dose decreased from 8.74 ± 1.93 mg/day at M0 to 7.11 ± 2.28 mg/day at M4, and further to 6.07 ± 2.07 mg/day at M6. Repeated-measures ANOVA demonstrated a statistically significant reduction in mean GC doses over time (F = 53.4, p < 0.001). Post hoc pairwise comparisons using Tukey's post hoc test confirmed significant reductions between M0 and M4 (p < 0.01), between M0 and M6 (p < 0.001), and between M4 and M6 (p < 0.01). These findings indicate a gradual and sustained GC-sparing effect during the first six months following biologic therapy initiation (Table 2).

Variable | M0 | M4 | M6 | Test used | Test statistic | p-value
GC dose (mg/day)1 | 8.74 ± 1.93* | 7.11 ± 2.28* | 6.07 ± 2.07* | Repeated-measures ANOVA | 53.4 | <0.001

At M4, 24 patients (40.0%) achieved a GC dose ≤5 mg/day. At M6, 31 patients (51.7%) achieved a GC dose ≤5 mg/day, including one patient (1.7%) who achieved complete GC withdrawal (0 mg/day), whereas 29 patients (48.3%) remained on a GC dose >5 mg/day (Figure 1).

Figure 1: Distribution of GC doses at M6 following biologic therapy initiationGC: glucocorticoid, M6: month 6

For the comparative analysis presented in Table 3, baseline characteristics were compared between patients who achieved a GC dose ≤5 mg/day at M6 (n = 31) and those who remained on a dose >5 mg/day (n = 29).

Variable | ≤5 mg/day (n = 31) | >5 mg/day (n = 29) | Test used | Test statistic | p-value
Age, years1 | 52.48 ± 10.11 | 52.83 ± 13.19 | Student's t-test | 0.11 | 0.91
Disease duration, years1 | 17.98 ± 8.12 | 16.90 ± 10.45 | Student's t-test | 0.45 | 0.66
Hypertension2 | 5 (16.1) | 5 (17.2) | Chi² continuity correction | 0.00 | 1.00
Osteoporosis2 | 23 (74.2) | 18 (62.1) | Chi² | 1.02 | 0.31
Previous biologic therapy2 | 11 (35.5) | 12 (41.4) | Chi² | 0.22 | 0.64
Fibromyalgia2 | 10 (32.3) | 11 (37.9) | Chi² | 0.21 | 0.64
Anti-CCP positivity2 | 27 (87.1) | 26 (89.7) | Chi² continuity correction | 0.00 | 1.00
Erosive disease2 | 26 (83.9) | 26 (89.7) | Chi² continuity correction | 0.08 | 0.78
DAS28-CRP1 | 4.82 ± 0.77 | 5.36 ± 0.96 | Student's t-test | 2.44 | 0.02
DAS28-ESR1 | 5.26 ± 0.93 | 5.65 ± 0.69 | Student's t-test | 1.84 | 0.07
HAQ1 | 1.06 ± 0.66 | 1.22 ± 0.56 | Student's t-test | 1.02 | 0.31

The two groups were comparable in terms of age and disease duration. Mean disease duration was 17.98 ± 8.12 years in the ≤5 mg/day group and 16.90 ± 10.45 years in the >5 mg/day group (Student's t-test, t = 0.45, p = 0.66). Fibromyalgia was present in 10 (32.3%) patients in the anti-TNF group and 11 (37.9%) patients in the rituximab group, with no significant difference between the two treatment groups (χ² = 0.21, p = 0.64). Baseline functional disability, assessed using the HAQ, was also comparable between the two groups (1.06 ± 0.66 vs. 1.22 ± 0.56; Student's t-test, t = 1.02, p = 0.31). However, baseline DAS28-CRP was significantly lower in patients who achieved a GC dose ≤5 mg/day than in those who remained on a dose >5 mg/day (4.82 ± 0.77 vs. 5.36 ± 0.96; Student's t-test, t = 2.44, p = 0.02) (Table 3).

Comparison of GC doses between treatment groups

GC doses were compared between patients treated with rituximab and those receiving anti-TNF therapy at each time point using Student's t-test. At M0, the mean daily GC dose was 8.36 ± 1.21 mg/day in the rituximab group and 9.10 ± 2.39 mg/day in the anti-TNF group, with no statistically significant difference between groups (t = −1.52, p = 0.13).

At M4, the mean daily GC dose decreased to 6.81 ± 1.88 mg/day in the rituximab group and 7.39 ± 2.61 mg/day in the anti-TNF group, with no statistically significant difference between groups (t = −0.97, p = 0.33). At M6, the mean daily GC dose further decreased to 5.81 ± 1.72 mg/day in the rituximab group and 6.33 ± 2.38 mg/day in the anti-TNF group, again with no statistically significant difference between groups (t = −0.94, p = 0.34). The mean change in GC dose from M0 to M6 (ΔGC dose, M6 − M0) was −2.55 ± 1.70 mg/day in the rituximab group and −2.77 ± 2.40 mg/day in the anti-TNF group, with no statistically significant difference between groups (t = 0.41, p = 0.68).

Overall, both treatment groups showed a reduction in GC exposure over time, with no significant difference in the GC-sparing effect between rituximab and anti-TNF therapy during the six-month follow-up period (Table 4).

Variable | Rituximab group | Anti-TNF group | Test used | Test statistic | p-value
GC dose (mg/day)1M0 | 8.36 ± 1.21 | 9.10 ± 2.39 | Student’s t | -1.52 | 0.13
GC dose (mg/day)1M4 | 6.81 ± 1.88 | 7.39 ± 2.61 | Student’s t | -0.97 | 0.33
GC dose (mg/day)1M6 | 5.81 ± 1.72 | 6.33 ± 2.38 | Student’s t | -0.94 | 0.34
Δ GC dose (M0-M6)1 | -2.55 ± 1.70 | -2.77 ± 2.40 | Student’s t | 0.41 | 0.68