Plain Language Summary
High cholesterol puts people at risk of heart disease, especially those with diabetes. Physicians set individualized cholesterol treatment goals for each patient. Statins, prescribed to reduce high cholesterol levels, may not lower cholesterol enough in all people. Alirocumab is a medication for lowering cholesterol levels. Alirocumab is intended for use in combination with maximally tolerated statin. Here we compared the effects of alirocumab to 1054 people with diabetes to 2445 people without diabetes.
Our study showed that most people with and without diabetes reached the cholesterol goal of less than 70 mg per deciliter after 24 weeks of treatment with individualized alirocumab doses. Treatment with alirocumab 150 milligrams every 2 weeks resulted in 78% of people with diabetes and 77% of people without diabetes reaching this goal. Treatment with alirocumab 75 mg every 2 weeks (with some individuals getting their dose increased to 150 mg) also produced similar results in people with diabetes (68%) and without diabetes (66%). Eighty percent of people with diabetes and 80% of people without diabetes had adverse reactions with alirocumab treatment. Similar rates of adverse reactions were reported in the corresponding placebo groups with diabetes (82%) and without diabetes (81%). Regardless of diabetes status, the most common adverse reactions among alirocumab-treated people were common cold, chest infection, and injection-site reaction.
In summary, alirocumab provides an additional treatment option for people with and without diabetes who do not to reach their cholesterol goals, even with maximally tolerated statin dose.
Introduction
Cardiovascular disease (CVD) is the most common cause of death in persons with diabetes mellitus (DM) [
1,
2]. Individuals with DM are, on average, at double the risk of atherosclerotic CVD (ASCVD) in comparison to those without DM, and the presence of dyslipidemia in people with type 2 DM further increases the risk of adverse cardiovascular outcomes [
3,
4].
The elevated cardiovascular risk associated with DM is recognized in guidelines [
1,
3,
5,
6], which recommend more intense management strategies for low-density lipoprotein cholesterol (LDL-C)-lowering in individuals with DM than for the general population. Statins are recommended as first-line therapy to reduce LDL-C in DM [
1,
5,
7,
8]. However, many people with DM have persistent lipid abnormalities despite statin treatment [
9,
10]. The 2017 updated American College of Cardiology Expert Consensus Task Force and the 2018 American Diabetes Association standards of care recommend that a proprotein convertase subtilisin/kexin type 9 (PCSK9) inhibitor be considered in individuals with ASCVD and DM whose LDL-C levels are not optimally controlled on high-intensity statin therapy [
6,
8].
The 2017 update of the European Society of Cardiology/European Atherosclerosis Society Task Force also recommends the use of a PCSK9 inhibitor for two categories of individuals with DM, depending on LDL-C levels: (1) individuals with DM and clinical ASCVD whose LDL-C levels are > 100 mg/dL and (2) individuals with DM and familial hypercholesterolemia without clinical ASCVD whose LDL-C levels are > 140 mg/dL, despite maximally tolerated statin and ezetimibe therapies [
11].
The lipid profile of type 2 DM is characterized by reduced high-density lipoprotein cholesterol (HDL-C), high triglycerides, and an increase in the proportion of LDL-C particles that are small and dense, with or without high levels of LDL-C [
12,
13]. In the ODYSSEY DM-INSULIN study (NCT02585778), alirocumab 75 mg every 2 weeks (Q2W; with possible dose increase to 150 mg Q2W) significantly reduced LDL-C levels and other lipids in individuals with type 1 DM (
n = 76) or type 2 DM (
n = 441) treated with insulin [
14]. Furthermore, in individuals with type 2 DM and mixed dyslipidemia, the same alirocumab dosing regimen also resulted in significant reductions in non-HDL-C, LDL-C, and other lipids in ODYSSEY DM-DYSLIPIDEMIA (NCT02642159) [
15]. However, the efficacy and safety of alirocumab 75 mg Q2W (with possible dose adjustment to 150 mg Q2W) and 150 mg Q2W (without dose adjustment) have not been compared in a larger pool of individuals with and without DM treated for longer duration. This subgroup analysis of five placebo-controlled phase 3 studies (LONG TERM [NCT01507831] [
16], HIGH FH [NCT01617655] [
17], COMBO I [NCT01644175] [
18], FH I [NCT01623115] [
19], and FH II [NCT01709500] [
19]) aimed to compare the efficacy and safety of alirocumab in a large group of individuals with and without DM at baseline, with the primary efficacy endpoint being LDL-C reduction from baseline to week 24.
Methods
Trial Participants and Study Designs
The current analysis used patient-level data from study participants according to alirocumab dosing regimen who enrolled for five double-blind, randomized, placebo-controlled ODYSSEY phase 3 studies with 52–78 weeks’ treatment duration. Methods for each of the individual studies have been previously reported [
16‐
19].
In summary, all participants, with or without DM, had hypercholesterolemia at study entry and were on maximally tolerated, stable, background statin therapy with or without other lipid-lowering agents. The FH I and FH II studies recruited participants with heterozygous familial hypercholesterolemia (HeFH) and at either very high risk (LDL-C ≥ 70 mg/dL with prior CVD) or high risk (LDL-C ≥ 100 mg/dL but no prior CVD). The HIGH FH study enrolled individuals with HeFH and LDL-C levels ≥ 160 mg/dL. The LONG TERM study included participants who either had HeFH or established coronary heart disease (CHD) or CHD risk equivalents based on the European Systematic Coronary Risk Estimation (SCORE), with baseline LDL-C ≥ 70 mg/dL. The COMBO I study included participants with either established CVD and LDL-C ≥ 70 mg/dL, or CHD risk equivalents and LDL-C ≥ 100 mg/dL, based on the European SCORE, which included DM with other risk factors or chronic kidney disease.
All study protocols were approved by the appropriate institutional review boards, and all participants provided informed, written consent. All trials were performed in accordance with the ethical principles that have their origin in the Declaration of Helsinki and all applicable amendments laid down by the World Medical Assemblies and the International Conference Harmonisation Guidelines for Good Clinical Practice.
Participants were randomized to alirocumab or placebo groups in a 2:1 ratio. Two studies (LONG TERM and HIGH FH) used an alirocumab dose of 150 mg Q2W throughout the trial period. The other three studies (COMBO I, FH I, and FH II) used an initial alirocumab dose of 75 mg Q2W, with an increase to 150 mg Q2W at week 12 if the LDL-C level at week 8 remained ≥ 70 mg/dL. All doses were delivered by subcutaneous injection (alirocumab 75 mg, alirocumab 150 mg, or placebo) using a 1-mL dose volume.
Participants were classified as having DM (type 1 or 2) or not, according to medical history reported by the investigator.
Efficacy Analysis
Efficacy was compared in individuals with and without DM in two pools according to alirocumab dose regimen (pool of 150 mg Q2W studies; pool of 75 mg Q2W with possible increase to 150 mg Q2W, abbreviated in the text to 75/150 mg Q2W).
The main efficacy endpoint for this pooled analysis was the mean percentage change in calculated LDL-C from baseline to week 24 (the primary endpoint of the individual studies). Other efficacy endpoints included percentage change in calculated LDL-C from baseline to week 12 (prior to potential dose increase in trials utilizing the 75/150 mg Q2W dosing regimen), proportion of individuals achieving risk-based LDL-C goals, and the percentage change in other lipid parameters from baseline to weeks 12 and 24. Mean percentage change in LDL-C and other lipid parameters over time is also reported up to week 78.
Changes from baseline for the main LDL-C efficacy endpoint and other lipid values were statistically evaluated with an intention-to-treat (ITT) approach, which included lipid data from all randomized persons regardless of adherence to treatment. The analysis utilized a mixed-effect model with repeated measures to account for missing data, as previously described [
20]. Data on changes over time were presented according to ITT analysis, and on-treatment analysis using a modified ITT approach, which included only lipid data collected while the individual was receiving study treatment.
Other lipid parameters were analyzed either in the same way as the main efficacy endpoint or, in the case of lipoprotein (a) [Lp(a)] and triglyceride (TG) percentage changes, and LDL-C goal achievement proportions, analysis involved a multiple imputation approach then robust regression (for Lp(a) and TG percentage changes) or logistic regression (for LDL-C goal achievement), in the ITT population.
Consistency of treatment effect across subgroups was assessed by providing interaction p values. A further subgroup analysis was performed to compare alirocumab efficacy in persons with and without DM according to HeFH status at week 12 and week 24.
Safety Analysis
Safety data are reported in subgroups of individuals assigned to alirocumab or placebo, regardless of alirocumab dose, according to baseline DM status. All adverse events, regardless of seriousness and irrespective of potential relationship to alirocumab, were recorded by the investigators up to the last visit planned in the protocol. Treatment-emergent adverse events (TEAEs) were defined as events that developed, worsened, or became serious between the first and last dose of study treatment plus 70 days, classified according to the Medical Dictionary for Regulatory Activities. Adverse events of special interest included injection-site reactions, general allergic events, neurocognitive disorders, and adjudicated cardiovascular events. Statistical analysis of the safety population included all randomized individuals who received at least one dose or part of a dose of study drug, and safety data were analyzed by descriptive statistics.
The effect of alirocumab treatment on glycated hemoglobin (HbA1c) and fasting plasma glucose (FPG) was also evaluated according to DM status throughout the studies using descriptive statistics conducted on the safety population.
Discussion
Across all studies in the ODYSSEY program, individuals with hypercholesterolemia experienced significant LDL-C reductions from baseline compared with either placebo or ezetimibe following alirocumab treatment with background statin therapy (± other lipid-lowering therapies) [
16‐
19]. Similar reductions were seen with alirocumab treatment in individuals with or without DM, and were maintained for up to 78 weeks.
In the present analysis, alirocumab 150 mg Q2W resulted in 58.5% and 62.1% reductions in LDL-C levels versus placebo at week 24 with no significant difference being observed between those with and those without DM, respectively. Furthermore, LDL-C levels of < 70 mg/dL were achieved by the majority of persons receiving this dose regardless of DM status (76.9–77.7%), with mean achieved LDL-C levels of 50.2–52.3 mg/dL, regardless of DM status.
In the pool of studies where per-protocol dose increase from 75 mg to 150 mg Q2W occurred if LDL-C goals were not reached at week 8, the week 24 LDL-C levels from baseline were changed by − 43.8% (individuals with DM) and − 49.7% (individuals without DM) in the alirocumab groups (placebo, + 0.3% and + 5.1%, respectively), with these changes being consistent regardless of DM status. Possibly as a result of the higher LDL-C levels at baseline, alirocumab dose was increased at week 12 in a higher proportion of subjects without DM, which was associated with a greater magnitude of LDL-C reduction at week 24 in that group (absolute LDL-C change, − 66.3 mg/dL) versus the DM group (absolute LDL-C change, − 58.3 mg/dL); however, similar LDL-C levels (63.1–71.1 mg/dL) were achieved in both groups. In the subgroup without HeFH, the percentage of individuals requiring dose increase was similar regardless of DM status. Therefore, this analysis indicates that the higher proportion of individuals in the subgroup without DM who received alirocumab dose increase versus the subgroup with DM could be explained by more individuals with HeFH with higher baseline LDL-C being included in the subgroup without DM.
At baseline, the differences in prevalence of HeFH, ASCVD, and high-intensity statin use, and baseline LDL-C levels between individuals with and without DM may reflect discrepancies in selection criteria between the studies included in the analysis. Furthermore, consistent with previous studies [
21,
22], fewer individuals with HeFH had DM (12.2–15.5%) compared with those without HeFH (44.4–46.3%). Possibly as a result of this imbalance, fewer individuals in the cohort with DM had a history of ASCVD and received high-intensity statin compared with the cohort without DM. The lower rate of high-intensity statin use in individuals with DM indicates that this high-risk population is undertreated.
Overall, alirocumab treatment was generally well tolerated with no particular differences observed in incidence of TEAEs, serious adverse effects, or deaths compared with placebo in individuals with or without DM. Overall, mean exposure to randomized treatment was similar (78 weeks) in individuals with or without DM regardless of treatment status. This analysis does not allow for conclusive safety observations to be made, in particular for rare adverse events such as adjudicated cardiovascular events. As a consequence, safety results should be considered in the context of the overall ODYSSEY program.
Fewer injection-site reactions were reported in individuals with DM, a difference that may in part be due to a greater familiarity and tolerance associated with glucometer and/or injectable medication use in the DM population (27.8% and 72.2% of persons with DM were receiving antihyperglycemic injectable and non-injectable medication at baseline, respectively). However, as previously reported, administration by injection does not deter persons from self-administration of alirocumab [
23]. The ODYSSEY OUTCOMES study (NCT01663402) will help to further establish if DM status has an influence on the frequency of injection-site reactions [
24].
Mean HbA1c and FPG measurements in individuals with DM were comparable between alirocumab and placebo for up to 78 weeks of treatment. These variables were comparable between alirocumab and placebo in individuals without DM, as previously reported [
25]. On the basis of these results, alirocumab treatment does not appear to affect blood glucose, which is very reassuring given that statins modestly raise the risk of DM [
26], and recent genetic publications have suggested that the PCSK9 pathway might be relevant to glycemia levels [
27,
28]. In a subgroup analysis of 11,031 individuals with DM in FOURIER (median follow-up, 2.2 years), neither HbA1c nor incidence of DM was increased in the evolocumab groups versus the placebo groups [
29]. In contrast, a recent meta-analysis of phase 2/3 randomized PCSK9 inhibitor clinical studies (excluding FOURIER) suggested that PCSK9-mediated lowering of LDL-C does increase risk for DM [
30]. Taking all of these publications into consideration, it is not yet clear whether LDL-C reduction per se or the means of attaining lower LDL-C levels influence DM risk. In the ODYSSEY and FOURIER studies, most individuals received statin therapy, which may mask the glycemic effect of PCSK9 inhibitors [
29,
31]. The results of ODYSSEY OUTCOMES will provide key additional data relevant to this important issue.
Overall efficacy and safety findings from these placebo-controlled studies were consistent with findings from studies comparing alirocumab with ezetimibe and employing background statin therapy [
32,
33]. A recently published study of the efficacy and safety of alirocumab in individuals with type 2 DM and mixed dyslipidemia (defined as non-HDL-C ≥ 100 mg/dL; TGs ≥ 150 mg/dL and < 500 mg/dL) showed similar responses for non-HDL-C (37.7% reduction) and TGs (13.0% reduction) at week 24 compared to the larger group of individuals with type 1 and 2 DM in the present analysis [
15]. In the DM-INSULIN phase 3b study, alirocumab demonstrated similar results in reducing LDL-C and other lipids in insulin-treated type 1 DM and type 2 DM individuals [
14]; all individuals were at high cardiovascular risk and received maximally tolerated statin therapy.
This analysis was limited by the non-randomized nature of DM status in subgroups, which could introduce bias to the analyses. Nevertheless, this placebo-controlled analysis of up to 78 weeks adds to the body of evidence on PCSK9 inhibitor use in people with DM [
34].
Acknowledgements
The authors would like to thank study individuals and investigators, and the following persons from sponsors for their contributions to data collection and analysis, assistance with statistical analysis, or critical review of the manuscript: Regeneron Pharmaceuticals Inc.: Eva-Lynne Greene, Carol Hudson, Rita Samuel, and Desmond Thompson; Sanofi: Corinne Hanotin, Michael Howard, L. Veronica Lee, Guillaume Lecorps, Laurence Merlet, Christelle Lorenzato, and Jonas Mandel.
Disclosures
Henry N. Ginsberg has received other research support from Genzyme (Sanofi), Merck, Sanofi, Regeneron Pharmaceuticals Inc., and Amgen; and is a consultant and on the advisory board for Amarin, Amgen, AstraZeneca, Bristol Myers Squibb, GlaxoSmithKline, IONIS, Janssen, Kowa, Merck, Novartis, Sanofi, Regeneron Pharmaceuticals, Inc., and Pfizer. Michel Farnier has received research support from and participated in a speakers’ bureau for Amgen, Merck, and Sanofi; received honoraria from Abbott/Mylan, Akcea/Ionis, Eli Lilly, and Pfizer; and acted as a consultant/advisory panel member for Amgen, AstraZeneca, Roche, Kowa, Merck, Pfizer, Sanofi, and Servier. Jennifer G. Robinson has received research grants from Amarin, Amgen, AstraZeneca, Daiichi-Sankyo, Eli Lilly, Esai, Genentech/F. Hoffman La Roche, GlaxoSmith Kline, Merck, Pfizer, Regeneron Pharmaceuticals, Inc./Sanofi, and Zinfandel/Takeda; and is a consultant and on the advisory board for Akcea/Ionis, Amgen, Eli Lilly, F. Hoffman LaRoche, Merck, Pfizer, and Sanofi/Regeneron Pharmaceuticals, Inc. Christopher P. Cannon has received research grants from Accumetrics, Arisaph, AstraZeneca, Boehringer Ingelheim, GlaxoSmithKline, Janssen, Merck, Takeda, Amgen, BMS, American College of Cardiology, and Sanofi/Regeneron Pharmaceuticals, Inc; and is a consultant and on the advisory board for Boehringer Ingelheim CSL Behring, Essentialis, GlaxoSmithKline, Merck, Kowa, Takeda, BMS, Pfizer, Sanofi, Regeneron Pharmaceuticals, Inc., and Lipimedix. Alexia Letierce is an employee and a stockholder of Sanofi. Maja Bujas-Bobanovic is an employee and a stockholder of Sanofi. Marie T. Baccara-Dinet is an employee and a stockholder of Sanofi. Michael J. Louie is an employee and stockholder of Regeneron Pharmaceuticals, Inc. Naveed Sattar has received honoraria from Merck; is a consultant and on the advisory board for Sanofi, Amgen, Eli Lilly, Boehringer Ingelheim, AstraZeneca, and Merck; and has been involved in the conduct of the GAUSS3 clinical trial of PCSK9 inhibition in patients with reported statin intolerance, sponsored by Amgen. Helen M. Colhoun reports grants, personal fees, and non-financial support from Sanofi and Regeneron Pharmaceuticals, Inc., during the conduct of the study; grants, personal fees and non-financial support from Eli Lilly & Company, grants and other from Roche Pharmaceuticals, grants from Pfizer Inc., Boehringer Ingelheim, and AstraZeneca LP, and shareholding in Bayer and Roche.