Antimalarials in Incomplete Lupus Erythematosus - Simulation Study of an Ongoing Clinical Trial

Open Access
Feger, Danielle Marie
Graduate Program:
Public Health Sciences
Master of Science
Document Type:
Master Thesis
Date of Defense:
October 26, 2017
Committee Members:
  • Vernon Chinchilli, Thesis Advisor
  • Duanping Liao, Committee Member
  • Nancy Olsen, Committee Member
  • power calculation
  • sample size
  • clinical trial
  • lupus
  • systemic lupus erythematosus
  • incomplete lupus
  • loss to follow-up
Incomplete lupus erythematosus (ILE) is believed to be a preclinical form of systemic lupus erythematosus (SLE)1, a chronic and potentially debilitating autoimmune disease. Early treatment is necessary to limit the potential for irreversible organ damage2,3. Hydroxychloroquine (HCQ), an antimalarial drug, is widely used in the treatment of both diseases1,4. However, it is unknown whether HCQ slows the progression of ILE to systemic lupus. The Study of Antimalarials in Incomplete Lupus Erythematosus (SMILE), a 96-week randomized, double-blind, placebo-controlled clinical trial, recently was initiated in autumn 2017. This trial proposes to quantify the effectiveness of HCQ to slow or halt the progression of incomplete lupus into systemic lupus, as defined by the Systemic Lupus International Collaborating Clinics (SLICC) criteria. The purpose of this thesis is to calculate statistical power and necessary sample size for the SMILE trial. Of primary interest is the risk of accumulating one or more SLICC criteria; secondary outcomes of interest include the risk of progressing from ILE to SLE, as well as changes in 52 soluble mediators and 13 autoantibodies over the course of the study. Original estimates for expected risk between treatment and placebo groups were provided by the principal investigators, and additional simulations to account for variance in risk estimates, inflation of sample size, and increased loss to follow-up were performed. Forty percent of patients receiving placebo and 20% of patients receiving hydroxychloroquine are expected to accumulate at least 1 SLICC criteria throughout the course of the study. With a total of 180 subjects, there is 90% statistical power to detect a statistically significant difference with a two-sided, 0.05 level test between treatment and placebo groups. Total patient recruitment may be inflated to 240 subjects to allow for up to 25% loss to follow-up while maintaining 90% statistical power. Statistical power to detect a significant difference between treatment and placebo groups with a two-sided, 0.05 level test is decreased when the risk ratio between groups is decreased, and increased when the proportion of patients accumulating SLICC criteria increased. Overall, the SMILE trial is well-powered to detect a significant difference in the risk of accumulating one or more SLICC criteria between treatment and placebo groups with a target sample size of 180 subjects needed to complete the trial. Statistically significant differences in secondary outcomes also are detectable at this sample size. However, previous clinical trials of lupus patients have shown relatively large loss to follow-up in this population. Therefore an enlarged target sample size of 240 subjects is appropriate for the SMILE trial.