A total of 220 patients were enrolled in the study; 110 were randomised to the MTX+PDN group and 110 to MTX monotherapy. The number of patients lost to the follow-up is shown in Figure . A total of 186 patients with complete clinical and US follow-up data were included in the analyses: 90 patients in the MTX arm and 96 patients in the MTX+PDN arm. Baseline characteristics of patients included in the analyses did not differ from excluded patients.
Flow chart of patients enrolled in the trial. MTX, methotrexate monotherapy; MTX+PDN, methotrexate + prednisone combination therapy.
At baseline, the subjects assigned to the two therapeutic arms did not significantly differ for demographics, disease activity and disability indexes (Table ).
Baseline demographic, clinical, laboratory and ultrasonographic characteristics of patients according to therapeutic arm
Patients in the MTX+PDN group were less likely to adhere to the protocol, with a RR of 0.82 (95% confidence interval (CI) = 0.69, 0.96). The main reasons for discontinuation in the GC group (15 patients) were mild side effects (six patients) and unwillingness to continue GC therapy (nine patients). Among the 90 completers in the MTX monotherapy group, five patients started GC medication because of increased disease activity over time. The maximum MTX dose was 14.3 ± 4.3 mg/day in the combination therapy group and 16.1 ± 3.9 mg/day in the MTX monotherapy group (P = 0.053).
Therapeutic adjustments were made in 101 patients - 57 patients for MTX alone and 44 patients for the MTX+PDN arm - with a trend toward less therapeutic adjustments in the MTX+PDN group (incident rate ratio = 0.77, 95% CI = 0.50, 1.16; P = 0.19). Sixteen patients in the MTX group and nine patients in the MTX+PDN group started anti-TNF therapy before the completion of 12 months of follow-up. At the end of the follow-up, six and 10 patients, respectively, were still in treatment with conventional DMARDs despite not achieving LDA; nine patients denied consent for biologic therapy, three patients had intercurrent adverse events leading to delay in treatment steps, and four patients had disease reactivations at the last visit. Eight patients in the MTX arm and six patients in the MTX+PDN arm underwent intra-articular or peri-articular steroid injections.
At the end of the follow-up period, LDA according to the disease activity score was achieved by 145 out of 186 patients (77.9%), without significant differences between the two treatment arms: 68 patients (75.5%) in the MTX group and 77 patients (80.2%) in the MTX+PDN group (P = 0.44).
On the whole, subjective (VAS score for pain, VAS score for patient's global assessment, VAS score for global health assessment) and objective (swollen and tender joint counts) variables, as well as US variables, significantly improved throughout the follow-up (Figure to ).
Figure 2 Trend over time of measures of clinical subjective and objective variables. Trend over time of measures of clinical subjective and objective variables along the two treatment arms. Subjective: (A) visual analogue scale (VAS) score for pain. Objective: (more ...)
Comparing the two therapeutic arms, the mean VAS score for pain (Figure ) fell more rapidly in the MTX+PDN arm, reaching a significant difference between groups at 2 months (MD = -17.9 (95% CI = -26.4, -9.4), P < 0.001) and at 4 months (MD = -10.8 (95% CI = -19.1, -2.5), P = 0.01). This difference was not significant at 6 and 9 months, while it was weakly significant at the end of the follow-up (MD = -8.8 (95% CI = -17.5, -0.1), P = 0.04). Similarly, the mean swollen joint count (Figure ) fell more rapidly in the MTX+PDN arm, reaching a significant mean difference at 2 months (MD = -2.0 (95% CI = -3.6,-0.3), P = 0.01) and 4 months (MD = -1.9 (95% CI = -3.6, -0.2), P = 0.02). This difference was not significant at 6, 9 and 12 months. No significant differences were observed for the tender joint count (Figure ) throughout the entire follow-up period. Acute-phase reactants (erythrocyte sedimentation rate and C-reactive protein) showed a significantly greater improvement (Figure ) in the MTX+PDN group throughout the entire period of follow-up.
Comparing the two therapeutic arms, the mean DAS28 score (Figure ) fell more rapidly in the MTX+PDN group, while the differences became not statistically significant throughout the follow-up: the DAS28 MD was -0.69 (95% CI = -1.04, -0.34), P < 0.001 at 2 months; -0.76 (95% CI = -1.12, -0.39), P < 0.001 at 4 months; -0.35 (95% CI = -0.72, 0.01), P = 0.06 at 6 months; -0.17 (95% CI = -0.55, 0.20), P = 0.36 at 9 months; and -0.27 (95% CI = -0.67, 0.12), P = 0.18 at 12 months.
GS and PD scores also showed a more rapid response in the MTX+PDN group. The GS score MD was -1.78 (95% CI = -3.66, 0.09), P =0.06 at 6 months and -1.52 (95% CI = -0.66, 3.71), P = 0.17 at 12 months. The PD score MD was -1.72 (95% CI = -3.44, -0.01), P = 0.04 at 6 months and -1.24 (95% CI = -3.21, 0.73), P = 0.21 at 12 months.
The frequency of patients achieving clinical remission (DAS28 < 2.6) was significantly higher in the MTX+PDN group than in the MTX group: 43/96 (44.8%) versus 25/90 (27.8%) (P = 0.02). The relative risk of clinical remission was on average 60% higher in the group of combination therapy when compared with MTX monotherapy (RR = 1.61 (95% CI = 1.08, 2.04)). Globally, SDAI remission was achieved less frequently than DAS28 remission; however, MTX+PDN-treated patients had significantly higher probability of SDAI remission over MTX monotherapy arm at 12 months (30.8% vs. 16%), with a RR of 1.91 (95% CI = 1.08, 3.38; P = 0.01).
Similarly, the probability of reaching PD negativity was significantly higher in the MTX+PDN group (67/96, 69.8%) than in the MTX group (48/90, 53.3%) (P = 0.04), with a RR at 12 months of 1.31 (95% CI = 1.04, 1.64). The combined clinical and US outcome was achieved by 34/96 patients (35.5%) and 14/90 (15.9%) patients in the MTX+PDN and MTX groups, respectively (P = 0.01), with a RR of 2.27 (95% CI = 1.31, 3.95). The abovementioned clinical and imaging outcomes are presented in Figure , shown as the percentages of patients reaching the outcomes at the end of the follow-up period.
Figure 3 Patients' relative frequencies of clinical and/or ultrasonographic outcomes according to treatment group. DAS28R, clinical remission according to the disease activity score among 28 joints; LDA, low disease activity; MTX, methotrexate monotherapy; MTX+PDN, (more ...)
Per-protocol and intention-to-treat analysis results were almost similar: performance on combination therapy was significantly better over MTX monotherapy for both clinical and US outcomes. Specifically, the frequency of patients achieving clinical remission (DAS28 < 2.6) was significantly higher in the MTX+PDN group than in the MTX group with a relative risk (RR) of 1.65 (95% CI = 1.07, 2.54; P = 0.01). Similarly, the probabilities of reaching PD negativity and the combined clinical and US outcome were significantly higher in the MTX+PDN group than in the MTX group: RR = 1.34 (95% CI = 0.99, 1.81; P = 0.05) and RR = 2.49 (95% CI = 1.27, 4.88; P = 0.005), respectively.
Six adverse events related to GC medication were reported in six patients; namely, hyperglycaemia (two patients) and epigastric pain (four patients). Side effects related to MTX medication - such as nausea, fatigue or slight increase in liver test function - leading to discontinuation were found in 10 patients from the MTX group and in six patients in the PDN co-medication group (P = 0.29).