Background/Purpose: To investigate the effect of intravenously administered (IV) golimumab (GLM, 2 mg/kg), an anti-TNFα monoclonal antibody, on sleep problems, total back pain (TBP), and night back pain (NBP) in adult patients with active AS.

Methods: GO-ALIVE is a Phase 3, multicenter, randomized, double-blind, placebo (PBO)-controlled trial. Patients (≥18 years) had a diagnosis of definite AS (modified New York criteria) and BASDAI ≥4, TBP visual analog scale (VAS) ≥4, and CRP ≥0.3mg/dL.  At baseline, 208 patients were randomized to IV GLM 2mg/kg (N=105) at Wk0, 4, and every 8 weeks or placebo (N=103) at Wk0, 4, and 12, with crossover to IV GLM at Wk16 and through Wk52. Sleep problems were assessed using the Medical Outcomes Study Sleep Scale (MOS-SS, range 0-100), a generic instrument designed to assess six dimensions of sleep: Sleep disturbance, Somnolence, Sleep adequacy, Snoring, Awaken short of breath or headache, and Quantity of sleep/optimal sleep during the past 4 weeks. The six dimensions are also used to generate the composite Sleep Problems Index (SPI). An increase in score from baseline represents improvement. TBP and NBP over the past week were assessed using VAS (0-10 cm; 0=no pain, 10=most severe pain). Wk28 results are presented here.  Unadjusted p-values of least square mean differences between treatment groups were based on analysis of covariance controlling for prior anti-TNF therapy.

Results:
Mean changes in MOS-SS Sleep Index and 6 subscales are presented in Table 1. Greater mean improvement (p<0.05) in the MOS-SS sleep index and 4 subscales at Wk8 was observed in GLM compared to PBO, and in the sleep index and 4 subscales at Week 16. Mean improvements from baseline to Weeks 8 and 16 in patient’s assessment of TBP (cm) were greater (p<0.001) in GLM than PBO (-2.70 vs -0.86 and -3.15 vs -1.15, respectively), and after PBO crossed over to GLM, the differences diminished at Wk28 (-3.14 vs -3.34, respectively). Mean improvements at Wk8 and 16 from baseline in patient’s assessment of NBP (cm) were also greater (p<0.001) in GLM than PBO (-3.03 vs -0.87 and -3.44 vs -0.85, respectively), and differences diminished at Wk28 (-3.47 vs -3.42, respectively). Changes from baseline in all subscales of MOS-SS were correlated (Spearman correlations ranging between -0.10 and -0.45) with TBP and NBP at Wk8, 16, and 28 (p values <0.05), with the exception of Snoring and both TBP and NBP at Wk16.  Change in NBP was associated with change in Sleep Problem Index at all 3 time points (p=0.002, p=0.001, and p=0.031, respectively). In the general linear model, most of the association between change in TBP and change in Sleep Problem Index was explained by the association between change in NBP and change in Sleep Problem Index.

Conclusion:
Adult patients with active AS treated with IV GLM showed improvements in sleep problems, TBP, and NBP.  NBP improvement was associated with improvement in sleep problems.  

*At Wk28, PBO has crossed over to GLM