The observed number of 24-hour censored observations was higher than expected and the wash-out period for basal insulin treated patients may have been too short to definitively rule out a carry-over effect; however, such an effect, if present, would potentially only affect onset of action and not the primary outcome measure.

Objective: The objective of the study was to evaluate pharmacodynamic (PD) intrasubject variability of a single, s.c. dose of insulin lispro protamine suspension (ILPS) compared with insulin glargine in subjects with type 1 diabetes mellitus and additionally, to compare the intrasubject variability of pharmacokinetic parameters of both insulins. Design: This was a single-center, investigator-blinded and subject-blinded, two-arm, parallel, randomized, four-period study. During the replicate visits, subjects received a single s.c. 0.6 U/kg dose of either ILPS or glargine, and underwent 24-h euglycemic glucose clamps. Results: The intrasubject variabilities of the primary PD parameters, total amount of glucose infused (G tot ) and maximum glucose infusion rate (GIR; R max ), were statistically significantly lower for ILPS when compared with glargine (P!0.0001). Least-square (LS) mean estimates for G tot and R max were 2512.7 mg/kg and 3.740 mg/min per kg respectively for ILPS, and 1291.9 mg/kg and 1.793 mg/min per kg respectively for glargine. The LS mean estimates for G tot and R max were statistically greater (PZ0.0010 and P!0.0001 respectively) for ILPS compared with glargine, suggesting that ILPS had greater 24-h glucose-lowering activity. Glargine demonstrated a flatter GIR-time curve, and ILPS demonstrated a significantly shorter time of maximum GIR (tR max ) and earlier time to half-maximal GIR before tR max and time to half-maximal GIR after tR max . ILPS administration resulted in significantly greater exposure compared with glargine (area under the baseline-corrected serum concentration versus time curve from time 0 to 24 h (AUC 0-24 ): 77 150 vs 53 111 pmol min/l; maximum serum insulin concentration (C max ): 119 vs 68 pmol/l; ILPS versus glargine respectively), but the intrasubject variabilities for AUC and C max were comparable. Conclusion: Although glargine demonstrated a flatter GIR-time profile, the lower PD intrasubject variability of ILPS may provide a more predictable response.