Method: APPLE is a randomized, placebo-controlled, multicenter trial randomizing 221 subjects with pSLE (age 10-21 years) to 36 months of atorvastatin or placebo. CIMT was measured by standard protocol at 0, 12, 24 and 36 months. Fasting lipids and hscrp were measured centrally. Subgroups were defined as LDL < or >/= 110 mg/dL, hscrp < or >/= 1.5 mg/dL, age < or >/= 15.5 yrs, SLE duration < or ³ 24 months, adherence < or >/= 75%. Prepubertal status was defined by self-reported Tanner stage < 4. An additional combined subgroup of post-pubertal subjects with hscrp ³ 1.5 mg/dL vs. others was defined.
The statistical approach replicated primary APPLE efficacy analysis comparing CIMT progression rates between treatment groups based on test of interaction between treatment group and time in a longitudinal linear mixed effects model. Similar models were used to assess non-CIMT outcomes. Indicator variables were included for subgroup level as well as two- and three-way interactions between subgroup, treatment group and time between subgroup level, treatment group and time. All statistical analyses were 2-sided, and the level of significance for all analyses was 0.05.
Result: Differences in treatment effects between hscrp subgroups were identified for mean-mean common (p=0.029) and mean-mean near wall (p=0.014). In both cases, the reduction in CIMT progression associated with atorvastatin assignment was more pronounced in the elevated hscrp subgroup. Similar differences were identified by pubertal status for mean-mean bifurcation (p=0.019) mean-max near wall (p=0.027) and mean-mean near wall (p=0.021), and by adherence for mean-mean far wall (p=0.035) and mean-mean near wall (p=0.028), where reduction in CIMT progression with atorvastatin assignment was more pronounced among post-pubertal or more adherent participants. No significant differences in treatment effects were observed between subgroups defined by LDL, age or SLE duration. The combined pubertal and high hscrp subgroup atorvastatin group had lower CIMT progression rates in five of 12 CIMT segments: mean-mean common (p=0.005), mean-mean (p=0.008), mean-mean bifurcation (p=0.023), mean-max near wall (p=0.029), and mean-mean near wall (p=0.002).
Conclusion: These exploratory results need to be confirmed and interpreted cautiously as multiple comparisons were performed. APPLE secondary analyses showed a trend of lower CIMT progression rates in the atorvastatin-treated pubertal and elevated hscrp subgroups, particularly in the combined pubertal subjects with high hscrp, suggesting that this subgroup may benefit from statin therapy.
Disclosure: S. P. Ardoin, None; L. E. Schanberg, Pfizer Inc, 9 ; C. I. Sandborg, None; H. Barnhart, None; E. Yow, None; G. Evans, None; K. Mieszkalski, None; N. T. Ilowite, Abbott Immunology Pharmaceuticals, 8, Genentech and Biogen IDEC Inc., 8, Regeneron, 2, Novartis Pharmaceutical Corporation, 5, Centocor, Inc., 5 ; E. von Scheven, None; B. A. Eberhard, None; L. F. Imundo, None; D. M. Levy, None; Y. Kimura, None; E. D. Silverman, None; S. L. Bowyer, None; M. G. Punaro, None; N. G. Singer, None; D. D. Sherry, None; D. K. McCurdy, None; M. Klein-Gitelman, None; C. A. Wallace, Amgen,Pfier, Centocor, Novartis, BMS, 2 ; R. M. Silver, None; L. Wagner-Weiner, None; G. Higgins, None; H. Brunner, None.