POS0850 A PHASE II RANDOMISED CONTROLLED TRIAL OF ORAL PREDNISOLONE IN EARLY DIFFUSE CUTANEOUS SYSTEMIC SCLEROSIS (PRedSS)

Abstract

BackgroundA highly controversial question is whether or not corticosteroids should be prescribed for patients with early diffuse cutaneous systemic sclerosis (dcSSc). Although the painful and disabling features of early dcSSc (including tight itchy skin, contractures, fatigue) have an inflammatory basis and are likely to respond to corticosteroids, corticosteroids are a risk factor for potentially life-threatening scleroderma renal crisis.ObjectivesOur aim was to examine safety and efficacy of moderate dose prednisolone in patients with early dcSSc. Specific objectives were to evaluate whether moderate dose prednisolone reduced pain and disability, and improved skin score, and whether prednisolone was safe with particular reference to renal functionMethodsPRedSS set out as a Phase II, multicentre, double-blind randomised controlled trial, converted to open-label because of the Covid-19 pandemic. Patients were randomised to receive either moderate dose prednisolone (approximately 0.3 mg/kg) or matching placebo (or no treatment during open-label) for 6 months. The co-primary endpoints were the Health Assessment Questionnaire Disability Index (HAQ-DI) and modified Rodnan skin core (mRSS) at 3 months. Over 20 secondary endpoints included patient reported outcome measures reflecting pain, itch, anxiety and depression, fatigue and helplessness. 72 participants randomised 1:1 were planned and anticipated to yield 60 evaluable, giving over 80% power for each co-primary outcome in ANCOVA analyses [assumptions; HAQ-DI (α = 0.025, δ = -0.6, σ = 0.9, ρ = 0.6), mRSS (α = 0.025, δ = -5.5, σ = 8.2, ρ = 0.6)]. Mixed Models for Repeated Measures (week 6, month 3, month 6) were fitted with covariates trial arm, baseline score, anti-Scl-70 and their interactions with time point. An unstructured covariance matrix was assumed with the primary focus being the trial arm effect at 3 months.ResultsThe study terminated early due to the Covid-19 pandemic and consequently did not meet the recruitment target of 72 patients. Thirty-five patients (Table 1) were randomised (17 to prednisolone and 18 to placebo/control, 25 during the double-blind phase), of whom 34 completed the 3 month assessment. The adjusted mean difference between treatment groups at 3 months in HAQ-DI score was -0.10 (97.5% CI -0.29 to 0.10), p=0.25, and in mRSS -3.90 (97.5% CI -8.83 to 1.03), p=0.070, both favouring prednisolone but not significantly. Patients in the prednisolone group experienced less pain, helplessness and anxiety than control patients at 3 months: mean difference in pain scores -0.49, 95%CI (-0.93 to -0.06), p=0.027, in Hospital Anxiety and Depression (HADS) anxiety scores -2.05, 95%CI (-3.73 to -0.37), p=0.018, and in helplessness scores -1.54, 95%CI (-3.01 to -0.07), p=0.040. There were no renal crises.Table 1.Baseline characteristics of patients by treatment allocationCharacteristicPrednisolone (n=17)Control (n=18)Age (years)52.7 (14.0)55.3 (12.7)Female n (%)10 (59)9 (50)Duration of skin thickening (years)1.6 (0.8)1.7 (0.8)Anti-topoisomerase-1 n (%)5 (29)6 (33)Anti-RNA polymerase III n (%)6 (35)8 (44)HAQ-DI1.6 (0.8)1.7 (0.7)mRSS18.8 (7.9)23.5 (8.6)Values are mean (standard deviation) unless stated otherwiseConclusionPRedSS exemplified the challenges of running a clinical trial of an investigational medicinal product potentially associated with increased infection risk during the Covid-19 pandemic. Because PRedSS was terminated prior to target recruitment, it was underpowered, and any conclusions have to be extremely cautious. Although PRedSS suggested some benefit from moderate dose prednisolone, the small sample indicates the need for a further randomised trial.