Method: SLE patients treated with HCQ for at least 6 months, with stable disease and a SELENA-SLEDAI ≤12, and without neither known retinopathy nor suspected non adherence, were included in this randomized, double-blind, placebo-controlled, multicenter (n=37) trial and had a centralised [HCQ] dosage. Patients with [HCQ] ranging between 100 and 750 ng/ml were eligible for randomisation between no daily dosage change (group 1) and increased HCQ dosage to obtain a [HCQ] ≥1000 ng/ml (group 2). All randomised patients received 4 tablets/day (including HCQ and placebo), and were followed-up at 1, 3, 5, and 7 months.
Result: 573 subjects were included. At day 0, the median SLEDAI was 2.02 ± 2.4. The median [HCQ] was 857 [0-3316] ng/ml. There was an inverse correlation between [HCQ] and disease activity measured by SELENA-SLEDAI (p=0.023).
10 patients were excluded for major non-adherence ([HCQ]<100 ng/ml), 354 had early study discontinuation ([HCQ]>750 ng/ml), and 209 were eligible for randomisation (100< [HCQ] <750 ng/ml). 33 patients were not randomised due to patient refusal (n=11), contra-indications (ophthalmological n=8, and renal insufficiency n=1), SLE flares between inclusion and randomization (n=4), non-adherence recognition (n=4), pregnancy (n=2) or other causes (n=3). 5 dropped out just after randomization. Then, 171 patients were analyzed.
Median [HCQ] significantly increased between inclusion and randomisation (548 [102-749] to 613 [35-2005]; P<0,001), meaning before any therapeutic action.
Median [HCQ] were similar between group 1 (n = 84) and group 2 (n = 87) at randomization (623 [193-1610] versus 591 [35-2005]; p = 0.2), and were significantly higher in group 2 at M1 (1250 [9-3253] vs 733 [185-1700]; p<0.001), M3 (1362 [44-4005] vs 661 [239-1797]; p <0.001), M5 (1428 [33-3282] vs 718 [194-2013]; p <0.001) and M7 (1271 [13-3311] vs 665 [80-1578]; p <0.001).
The number of SLE flares was similar in both groups (26.2% vs 27.6%; p = 0.83).
Many patients in group 1 had [HCQ] in the therapeutic range and conversely, patients in group 2 had low [HCQ]. When we mixed both groups, the comparison of patients with [HCQ] < 1000 and those with [HCQ] ≥ 1000 during the 7 months of follow-up showed a rate of SLE flares of 18/53 (34%) and 8/36 (22%) respectively (p=0.23).
Conclusion: We confirm that low [HCQ] is associated with higher SLE activity. We failed to demonstrate that the adaptation of HCQ dosage reduces SLE flares in patients with low [HCQ]. However, the knowledge of low [HCQ] and the inclusion in the study improved treatment adherence, allowing many patients to reach higher [HCQ]. Patients with higher [HCQ] during the study had a trend toward a reduction of SLE flares although not statistically significant.
Disclosure: N. Costedoat-Chalumeau, None; L. Galicier, None; O. Aumaitre, None; C. Francès, None; V. Le Guern, None; F. Lioté, None; A. Smail, None; N. Limal, None; L. Perard, None; H. Desmurs-Clavel, None; B. Asli, None; C. Grandpeix, None; O. Pourrat, None; F. Ackermann, None; T. Papo, None; B. Brihaye, None; O. Fain, None; J. Stirnemann, None; M. Jallouli, None; J. Cohen, None; M. L. Tanguy, None; J. Hulot, None; L. Musset, None; Z. Amoura, None; .. the investigators of the PLUS study, None.