Anakinra for the Treatment of Acute Gout Flares

A Randomized, Double-Blind, Placebo-Controlled, Active-Comparator, Non-Inferiority Trial

Carly A. Janssen; Martijn A. H. Oude Voshaar; Harald E. Vonkeman; Tim L. Th. A. Jansen; Matthijs Janssen; Marc R. Kok; Bea Radovits; Caroline van Durme; Hetty Baan; Mart A. F. J. van de Laar


Rheumatology. 2019;58(8):1344-1352. 

In This Article


Patient Characteristics

Eighty-eight patients were enrolled in the study, and all patients (100%) completed the seven-day follow-up. Of these, 43 patients (48.9%) were randomized to a treatment with anakinra and 45 (51.1%) to treatment as usual (Figure 1). Patient baseline characteristics are listed in Table 1. The majority of patients in both groups (>90%) were male and more than half of the patients had monoarticular gout. Of the patients in the treatment as usual arm, 18 received a treatment with colchicine, and 13 and 14 patients received a treatment with naproxen and prednisone, respectively.

Figure 1.

Flow diagram of patient selection

Efficacy Assessment

Pain scores measured on the five-point rating scale decreased to a similar extent in both groups over days 1–4. For the PP population, results of the univariate ANCOVA for the primary outcome showed that the estimated marginal mean difference between treatment arms, −0.132 points on the five-point rating scale, was in favour of anakinra, and that the upper bound of the 95% CI of this difference (−0.44, 0.18) did not surpass the NI margin of 0.4 (Figure 2). In the ITT population, plausible values for missing data of the five-point rating scale pain scores at baseline, day 2, day 3, and day 4 post baseline were generated for 12 (13.6%), 6 (6.8%), 3 (3.4%) and 5 (5.7%) cases, respectively. Here, the primary analysis yielded similar results, with an estimated marginal mean difference of −0.178 (95% CI −0.44, 0.08) (Figure 2). Because in both of these analyses the upper bound of the 95% CI did not surpass the 0.4 NI margin, the null hypothesis that treatment with anakinra is less effective than treatment as usual for treating acute gout flares by at least 0.4 points on the five-point rating scale, was rejected. We concluded that the pain-relieving effect of anakinra in patients with gout was greater than the estimated effect of placebo, and that efficacy of anakinra was non-inferior to treatment as usual.

Figure 2.

The 95% CI for the estimated marginal mean difference in ΔPain between the treatment groups
ΔPain is the mean change in patient-reported pain from baseline to the average of pain scores at days 2–4 on the five-point rating scale. NI: non-inferiority.

Secondary Outcomes

For all the secondary outcomes assessed using the linear mixed effects model, the pattern of change was similar for the anakinra and treatment as usual group over 5 days, with significant improvements over time, but no significant between-group effects, or group-by-time interactions (Figure 3; results for PGA not shown). These findings provide no evidence for a differential effect on any outcome. For PGA, the mean ± standard error (S.E.) scores increased from 4.4 ± 0.34 to 6.7 ± 0.31 in the conventional treatment arm, and from 4.9 ± 0.38 to 7.31 ± 0.32 in the anakinra treatment group. After 2 days, more patients in the anakinra treatment group achieved ≥ 50% decrease in NRS pain scores (odds ratio (OR) 1.41, 95% CI 0.53, 3.73) compared with the treatment as usual arm, although not statistically significant. Also, on days 3–5, OR were in favour for anakinra. However, only on day 3 was this difference statistically significant (Supplementary Table S1, available at Rheumatology online).

Figure 3.

Mean scores on days 1–5 (bars represent one-sided 95% CI) of the secondary outcomes (panel A–D)

Both treatment groups showed a reduction in their CRP values after 7 days from baseline. The mean ± SE reduction in the treatment as usual arm was 13.9 ± 7.1 after 7 days, which was not significantly different to the reduction observed in the anakinra treatment arm of 11.1 ± 7.5.

During the first 7 days following baseline, more patients in the anakinra group (n = 20, 46.5%) compared with the treatment as usual arm (n = 16, 35.6%), took some form of pain-relieving medication, over the counter or prescription-based, aside from their study medication. However, this difference was not statistically significant between the groups.

Safety Assessment

Out of the 88 patients, 36 (40.9%) patients reported an AE during the first 7 days post baseline, of which more than half (n = 21, 58.3%) were patients in the treatment as usual arm (Table 2). Interestingly, three injection site reactions were reported in the treatment as usual group with the placebo injections, compared with none in the anakinra treatment group. One patient in the treatment as usual group reported to have viral gastroenteritis (stomach flu), and one patient in the anakinra group had a respiratory tract infection (severe cold). No SAE were reported during the 7 days following baseline.