Bococizumab is a humanized monoclonal antibody binding proprotein convertase subtilisin/kexin type 9, which may be a potential therapeutic option for reducing low-density lipoprotein cholesterol (LDL-C) levels in patients with hypercholesterolemia. In this 24-week, multicenter, double-blind, placebo-controlled, dose-ranging study (NCT01592240), subjects with LDL-C levels ≥80 mg/dl on stable statin therapy were randomized to Q14 days subcutaneous placebo or bococizumab 50, 100, or 150 mg or Q28 days subcutaneous placebo or bococizumab 200 or 300 mg. Doses of bococizumab were reduced if LDL-C levels persistently decreased to ≤25 mg/dl. The primary end point was the absolute change in LDL-C levels from baseline to week 12 after placebo or bococizumab administration. Continuation of bococizumab administration through to week 24 enabled the collection of safety data over an extended period. Of the 354 subjects randomized, 351 received treatment (placebo [n = 100] or bococizumab [n = 251]). The most efficacious bococizumab doses were 150 mg Q14 days and 300 mg Q28 days. Compared with placebo, bococizumab 150 mg Q14 days reduced LDL-C at week 12 by 53.4 mg/dl and bococizumab 300 mg Q28 days reduced LDL-C by 44.9 mg/dl; this was despite dose reductions in 32.5% and 34.2% of subjects at week 10 or 8, respectively. Pharmacokinetic/pharmacodynamic model-based simulation assuming no dose reductions predicted that bococizumab would lower LDL-C levels by 72.2 and 55.4 mg/dl, respectively. Adverse events were similar across placebo and bococizumab groups. Few subjects (n = 7; 2%) discontinued treatment because of treatment-related adverse events. In conclusion, bococizumab significantly reduced LDL-C across all doses despite dose reductions in many subjects. Model-based simulations predicted greater LDL-C reduction in the absence of bococizumab dose reduction. The Q14 days regimen is being evaluated in phase 3 clinical trials.
The serine protease proprotein convertase subtilisin/kexin type 9 (PCSK9) regulates cholesterol homeostasis by binding to the epidermal growth factor–like A domain of the low-density lipoprotein receptor (LDLR), inducing its degradation. LDL uptake is reduced, resulting in higher circulating levels of low-density lipoprotein cholesterol (LDL-C). Gain-of-function mutations in PCSK9 increase LDL-C levels and the risk of cardiovascular events. Conversely, loss-of-function PCSK9 mutations decrease plasma LDL-C and cardiovascular risk. Bococizumab (RN316/PF-04950615) is a humanized IgG2Δa monoclonal antibody (mAb) that recognizes and binds to the LDLR-binding domain of PCSK9, thus preventing PCSK9-mediated degradation of LDLR, leading to improved LDL clearance and reduction of serum LDL-C levels. In phase 1 and 2a clinical trials in hypercholesterolemic subjects, bococizumab reduced LDL-C levels by up to ∼75% and was generally well tolerated with few subjects discontinuing treatment because of adverse events (AEs). This dose-ranging phase 2b study evaluated the LDL-C–lowering effect of subcutaneous doses of bococizumab administered every 2 weeks (Q14 days) or monthly (Q28 days). A unique aspect of the trial design was the incorporation of bococizumab dose reductions if persistent LDL-C values ≤25 mg/dl were achieved. To our knowledge, this is the first study of a PCSK9 inhibitor to report the impact on LDL-C reduction when a dose reduction strategy is used in an effort to prevent extremely low levels of LDL-C. This report is the first full publication of bococizumab data from the phase 2 clinical trial program.
Methods
A more detailed description on the inclusion and exclusion criteria, study design, dose selection, drug administration, and laboratory evaluations can be found in the Supplementary Methods .
Enrolled subjects included men and women ≥18 years with hypercholesterolemia, on stable statin therapy (>6 weeks before screening), with a fasting LDL-C ≥80 mg/dl, and triglycerides ≤400 mg/dl. Subjects were excluded if they had a cardiovascular event during the previous 6 months, received treatment with systemic corticosteroids or an mAb during the previous 6 months, had congestive heart failure (New York Heart Association class III or IV), poorly controlled diabetes mellitus or hypertension, or diagnosis of cancer, human immunodeficiency virus, or other serious diseases.
This 24-week, multicenter, randomized, double-blind, placebo-controlled, parallel-group, dose-ranging phase 2b study (NCT01592240) was conducted from July 2012 to May 2013. Subjects were randomized using an Interactive Voice Response System in a 1:1:1:1:1:1:1 ratio to Q14 days subcutaneous placebo or bococizumab 50, 100, or 150 mg or Q28 days subcutaneous placebo or bococizumab 200 or 300 mg. Follow-up was for 6 to 8 weeks after the last dose of study drug. For both the Q14 days and Q28 days regimens, the bococizumab dose was reduced if LDL-C levels decreased to ≤25 mg/dl: the first dose reduction required 2 consecutive LDL-C levels ≤25 mg/dl; all further dose reductions required 1 LDL-C level ≤25 mg/dl. LDL-C data collected at 14 days after the dose provided information on whether dose reduction of bococizumab was required. For the Q14 days regimen, day 43 was the first opportunity for dose reduction. The bococizumab dose reduction sequence for this cohort was 150 mg → 100 mg → 50 mg → 25 mg → placebo or 100 mg → 50 mg → 25 mg → placebo or 50 mg → 25 mg → placebo. For the Q28 days regimen, day 57 was the first opportunity for dose reduction. The bococizumab dose reduction sequence for this cohort was 300 mg → 200 mg → 100 mg → 50 mg → placebo or 200 mg → 100 mg → 50 mg → placebo ( Supplementary Figures 1 and 2 ). The study was conducted under the guidelines of the Declaration of Helsinki and the Good Clinical Practice requirements of the International Conference on Harmonization. Local institutional review boards approved the protocol, and all subjects provided written informed consent. An independent data monitoring committee reviewed all study data.
The primary end point was the absolute change in LDL-C from baseline to week 12 after treatment with placebo or bococizumab, with the primary statistical analysis reporting this change as the placebo-adjusted treatment difference. Secondary end points included percentage change in LDL-C and absolute change and percentage change in non-HDL-C, HDL-C, total cholesterol, triglycerides, apolipoprotein (apo) A-I, apoB, and lipoprotein (a) from baseline at 12 and 24 weeks. The average placebo-adjusted change in LDL-C from weeks 10 to 12 in the Q28 days bococizumab dose groups was assessed as a tertiary end point. Safety end points included the incidence of AEs, serious AEs, laboratory abnormalities, incidence of antidrug antibodies (ADAs), and injection site reactions through to week 24 and up to 8 weeks of post-treatment follow-up. Bococizumab dose groups were compared with their respective placebo group using a mixed-model repeated measures analysis with the dependent variable being change from baseline (in LDL-C for the primary analysis) and including the fixed-effect terms treatment group, study visit time point, baseline value, treatment × study visit time point interaction, and baseline value × study visit time point interaction. An unstructured covariance matrix was used for the within-subject errors. The full analysis set population included all randomized subjects. However, the mixed-model repeated measures used for the statistical analyses of efficacy data only incorporated subjects who had a baseline and at least 1 postbaseline efficacy measurement. The analyses of safety data included all subjects who had received at least 1 dose of study medication. For all efficacy analyses, subjects remained in the dose group to which they were randomized, regardless of any subsequent dose reductions.
Demographic and baseline data were summarized as mean ± SD for continuous variables with the exception of triglyceride, lipoprotein (a), and PCSK9 values, which were not normally distributed and, therefore, presented as median (Q1 and Q3), and n (%) for categorical variables. Total cholesterol and triglyceride levels were assayed using standard enzymatic methods. LDL-C was measured by Friedewald and reflex ultracentrifugation for LDL-C levels ≤25 mg/dl.
Because of protocol-stipulated dose reductions of bococizumab in this study, a population pharmacokinetic/pharmacodynamic (PK/PD) model was developed using data from 7 phase 1 and phase 2 clinical studies to construct a predictor of bococizumab LDL-C lowering in the absence of dose reductions. The population PK/PD data set consisted of 7,574 bococizumab PK observations and 10,177 LDL-C measurements from 674 subjects. The model accounted for dose interruptions or dose reductions implemented in the phase 2 trials and any missed doses that may have occurred during the trials. A 2-compartment PK model with parallel first-order and nonlinear (Michaelis-Menten) elimination was linked to an indirect PD response model describing LDL-C response. Performance of the final PK/PD model was verified with model diagnostics and visual predictive checks and was found to accurately capture the observed pharmacokinetics of bococizumab and LDL-C response in all studies and in this phase 2b study, which included dose reductions triggered by LDL-C levels ≤25 mg/dl. Clinical trial simulations using subject demography data from this study were performed with the final PK/PD model to estimate the expected LDL-C response in this phase 2b study assuming no missed doses or dose reductions.
Results
A total of 354 subjects were randomized and 351 received treatment with either placebo (n = 100) or bococizumab (n = 251) ( Figure 1 ). The 3 subjects who were randomized but not treated did not have postbaseline efficacy measurements and so were excluded from further analysis (1 each from the placebo, 100-mg and 150-mg Q14 days groups). A total of 299 subjects (85% of those treated) completed treatment. The baseline demographics and clinical characteristics of subjects randomized to the various treatment groups were well balanced, including baseline LDL-C levels ( Table 1 ).
| Variable | Q14 days | Q28 days | |||||
|---|---|---|---|---|---|---|---|
| Placebo ∗ (n = 50) | Bococizumab (mg) | Placebo (n = 51) | Bococizumab (mg) | ||||
| 50 (n = 50) | 100 ∗ (n = 52) | 150 ∗ (n = 50) | 200 (n = 50) | 300 (n = 51) | |||
| Age (years) | 61 ± 10 | 59 ± 11 | 62 ± 10 | 61 ± 10 | 58 ± 12 | 60 ± 10 | 60 ± 8 |
| Male | 25 (50%) | 24 (48%) | 26 (50%) | 21 (42%) | 29 (57%) | 19 (38%) | 25 (49%) |
| White | 39 (78%) | 33 (66%) | 37 (71%) | 36 (72%) | 35 (69%) | 39 (78%) | 41 (80%) |
| Black | 9 (18%) | 15 (30%) | 12 (23%) | 10 (20%) | 14 (27%) | 9 (18%) | 8 (16%) |
| Asian and other | 2 (4%) | 2 (4%) | 3 (6%) | 4 (8%) | 2 (4%) | 2 (4%) | 2 (4%) |
| Weight (kg) | 91 ± 22 | 91 ± 24 | 90 ± 21 | 90 ± 17 | 91 ± 21 | 88 ± 24 | 89 ± 19 |
| BMI (kg/m 2 ) | 32 ± 7 | 32 ± 8 | 32 ± 7 | 32 ± 6 | 31 ± 6 | 31 ± 7 | 31 ± 6 |
| LDL-C (mg/dl) | 109 ± 32 † | 108 ± 20 | 113 ± 26 † | 106 ± 18 † | 119 ± 45 | 106 ± 23 | 105 ± 22 |
| TC (mg/dl) | 189 ± 35 † | 186 ± 35 | 195 ± 34 † | 189 ± 25 † | 198 ± 45 | 185 ± 30 | 179 ± 30 |
| TG (mg/dl) | 124 (82, 180) † | 109 (71, 155) | 135 (96, 176) † | 138 (100, 172) † | 109 (80, 169) | 123 (89, 158) | 113 (79, 149) |
| Non–HDL-C (mg/dl) | 137 ± 34 † | 134 ± 36 | 143 ± 31 † | 136 ± 23 † | 146 ± 45 | 133 ± 27 | 130 ± 29 |
| HDL-C (mg/dl) | 52 ± 14 † | 52 ± 16 | 51 ± 14 † | 53 ± 14 † | 53 ± 12 | 52 ± 14 | 49 ± 13 |
| ApoB (mg/dl) | 92 ± 20 † | 89 ± 22 | 97 ± 17 † | 90 ± 16 † | 97 ± 28 | 91 ± 16 | 87 ± 16 |
| ApoA-I (mg/dl) | 151 ± 25 † | 148 ± 24 | 149 ± 25 † | 151 ± 30 † | 150 ± 22 | 151 ± 26 | 145 ± 25 |
| Lp(a) (mg/dl) | 17 (5, 62) † | 25 (8, 64) | 18 (6, 66) † | 22 (7, 86) † | 18 (8, 48) | 15 (5, 60) | 31 (7, 60) |
| PCSK9 (ng/ml) | 305 (260, 344) † | 298 (257, 329) ‡ | 314 (238, 363) ‡ | 339 (308, 400) ‡ | 301 (233, 349) † | 317 (271, 378) † | 299 (252, 364) |
| Baseline statin dose § | |||||||
| Low | 26 (52%) | 25 (50%) | 22 (42%) | 20 (40%) | 28 (55%) | 23 (46%) † | 22 (43%) † |
| High | 24 (48%) | 25 (50%) | 30 (58%) | 30 (60%) | 23 (45%) | 26 (52%) † | 28 (55%) † |
∗ Three subjects (1 each from the placebo, 100-mg, and 150-mg Q14 days groups) were randomized but did not receive treatment.
† Data are missing for 1 subject in each of these groups.
‡ Data are missing for 2 subjects in each of these groups.
§ Low statin dose defined as atorvastatin ≤10 mg, rosuvastatin ≤5 mg, simvastatin ≤20 mg, pravastatin ≤40 mg, lovastatin ≤80 mg, and fluvastatin ≤40 mg daily; high statin dose defined as all higher doses of these statins.
Overall, 16%, 35%, 44%, and 39% of subjects randomized to bococizumab 100 mg Q14 days, 150 mg Q14 days, 200 mg Q28 days, and 300 mg Q28 days, respectively, had their dose reduced at any time during the study. No dose reductions occurred in the bococizumab 50-mg Q14 days group. At the week 10 visit (the Q14 days dosing time point before LDL-C measurement for the primary end point), 16% of available subjects in the bococizumab 100 mg Q14 days and 33% of subjects in the 150-mg Q14 days dose groups had their dose reduced ( Supplementary Figure 3 ). At the week 8 visit (the Q28-days dosing time point before LDL-C measurement for the primary end point), 30% of available subjects in the bococizumab 200 mg Q28 days and 34% of subjects in the 300-mg Q28 days group had their dose reduced ( Supplementary Figure 3 ).
Starting at week 2, Q14 days and Q28 days bococizumab dose regimens significantly reduced LDL-C levels, and this effect was sustained for the duration of the study. At week 12, the primary end point of mean absolute change from baseline in LDL-C was greatest in subjects receiving bococizumab 150 mg Q14 days ( Figure 2 , Table 2 ). The placebo-adjusted mean change from baseline in LDL-C in all subjects, including those with dose reductions, at week 12 ranged from −34.3 to −53.4 mg/dl for those receiving bococizumab Q14 days and from −27.6 to −44.9 mg/dl for those receiving bococizumab Q28 days ( Figure 3 , Table 2 ). This corresponded to placebo-adjusted mean percent changes from baseline of −35.0% to −53.1% for those receiving bococizumab Q14 days and −27.0% to −41.1% for those receiving bococizumab Q28 days ( Table 2 ). The PK/PD model-predicted, placebo-adjusted mean change from baseline in LDL-C at week 12 ranged from −36.4 mg/dl for bococizumab 50 mg Q14 days to −72.2 mg/dl for bococizumab 150 mg Q14 days ( Figure 3 ).
| Variable | Q14 days | Q28 days | |||||
|---|---|---|---|---|---|---|---|
| Placebo (n = 47) | Bococizumab (mg) | Placebo (n = 46–47) | Bococizumab (mg) | ||||
| 50 (n = 43–44) | 100 (n = 42–44) | 150 (n = 46) | 200 (n = 48) | 300 (n = 50) | |||
| LDL-C | |||||||
| Mean (SD) change, mg/dl | –2.8 (29.2) | –35.4 (26.6) | –52.3 (31.3) | –54.2 (27.0) | –1.3 (37.2) | –21.3 (28.0) | –38.3 (41.3) |
| Mean (95% CI) placebo-adjusted change, mg/dl | — | –34.3 (–45.1, –23.5) | –45.1 (–55.9, –34.2) | –53.4 (–64.1, –42.7) | — | –27.6 (–40.5, –14.7) | –44.9 (–57.7, –32.1) |
| P -value | — | <0.001 | <0.001 | <0.001 | — | <0.001 | <0.001 |
| Mean (SD) percent change, % | 0.6 (25.5) | –33.6 (23.3) | –44.9 (23.4) | –52.0 (24.7) | 3.3 (25.0) | –19.5 (26.6) | –33.3 (35.2) |
| Mean (95% CI) placebo- adjusted percent change, % | — | –35.0 (–44.9, –25.1) | –42.3 (–52.3, –32.3) | –53.1 (–63.0, –43.3) | — | –27.0 (–38.3, –15.7) | –41.1 (–52.3, –29.9) |
| P -value | — | <0.001 | <0.001 | <0.001 | — | <0.001 | <0.001 |
| Total cholesterol | |||||||
| Mean (SD) change, mg/dl | –5.6 (29.6) | –35.8 (24.7) | –52.7 (32.7) | –58.6 (32.9) | –0.4 (40.0) | –19.5 (27.4) | –37.6 (43.6) |
| Mean (95% CI) placebo-adjusted change, mg/dl | — | –29.7 (–41.8, –17.6) | –42.5 (–54.5, –30.4) | –55.6 (–67.6, –43.6) | — | –24.8 (–38.9, –10.7) | –45.6 (–59.8, –31.4) |
| P -value | — | <0.001 | <0.001 | <0.001 | — | <0.001 | <0.001 |
| Mean (SD) percent change, % | –2.4 (14.4) | –19.0 (12.3) | –26.5 (14.9) | –31.6 (18.0) | 1.2 (16.6) | –10.5 (15.3) | –19.4 (22.1) |
| Mean (95% CI) placebo- adjusted percent change, % | — | –15.7 (–22.1, –9.3) | –22.2 (–28.5, –15.8) | –30.1 (–36.5, –23.8) | — | –13.8 (–20.9, –6.7) | –23.8 (–30.9, –16.7) |
| P -value | — | <0.001 | <0.001 | <0.001 | — | <0.001 | <0.001 |
| HDL-C | |||||||
| Mean (SD) change, mg/dl | 0.1 (7.0) | 1.7 (7.1) | 2.1 (7.5) | 1.0 (7.5) | –1.0 (7.4) | 3.4 (9.2) | 3.1 (7.7) |
| Mean (95% CI) placebo-adjusted change, mg/dl | — | 1.6 (–1.4, 4.6) | 1.7 (–1.2, 4.7) | 0.5 (–2.4, 3.5) | — | 4.5 (1.2, 7.7) | 3.9 (0.6, 7.1) |
| P -value | — | 0.281 | 0.251 | 0.724 | — | 0.007 | 0.019 |
| Mean (SD) percent change, % | 0.8 (14.6) | 3.9 (15.0) | 3.9 (13.2) | 2.7 (12.7) | –0.4 (14.0) | 7.1 (17.1) | 6.9 (16.9) |
| Mean (95% CI) placebo- adjusted percent change, % | — | 3.4 (–2.4, 9.2) | 2.6 (–3.1, 8.4) | 1.4 (–4.4, 7.1) | — | 7.7 (1.3, 14.0) | 6.5 (0.2, 12.8) |
| P -value | — | 0.246 | 0.371 | 0.643 | — | 0.018 | 0.043 |
| Triglycerides † | |||||||
| Median (Q1, Q3) change, mg/dl | –18.0 (–52.0, 13.0) | –12.0 (–40.0, 9.0) | –13.0 (–43.5, 10.0) | –21.0 (–43.0, 9.0) | 5.0 (–23.0, 32.0) | –9.0 (–35.0, 11.5) | –14.5 (–51.0, 13.0) |
| Median (Q1, Q3) percent change, % | –14.5 (–32.5, 11.8) | –14.1 (–28.5, 7.8) | –14.8 (–35.6, 10.7) | –18.6 (–34.5, 5.0) | 3.7 (–11.3, 31.1) | –7.6 (–24.2, 14.7) | –13.8 (–32.7, 11.3) |
| Non–HDL-C | |||||||
| Mean (SD) change, mg/dl | –5.7 (26.9) | –37.5 (22.4) | –55.0 (32.8) | –59.6 (33.2) | 0.6 (38.0) | –22.9 (29.1) | –40.7 (44.9) |
| Mean (95% CI) placebo-adjusted change, mg/dl | — | –31.1 (–42.6, –19.6) | –44.1 (–55.6, –32.6) | –55.6 (–67.0, –44.2) | — | –29.3 (–43.5, –15.1) | –48.6 (–62.7, –34.4) |
| P -value | — | <0.001 | <0.001 | <0.001 | — | <0.001 | <0.001 |
| Mean (SD) percent change, % | –2.3 (20.2) | –28.3 (15.8) | –38.5 (21.4) | –44.9 (24.3) | 2.8 (20.3) | –17.3 (22.3) | –28.7 (30.9) |
| Mean (95% CI) placebo- adjusted percent change, % | — | –24.6 (–33.3, –15.9) | –33.2 (–41.9, –24.6) | –42.8 (–51.4, –34.2) | — | –22.6 (–32.4, –12.8) | –34.7 (–44.5, –24.9) |
| P -value | — | <0.001 | <0.001 | <0.001 | — | <0.001 | <0.001 |
∗ These values include subjects who had their bococizumab dose reduced in the weeks prior to week 12.
† The mixed model repeated measures analysis was not undertaken for triglycerides.
Before bococizumab dose reductions (up to week 6 for Q14 days and week 8 for Q28 days), placebo-adjusted mean reductions in LDL-C were greater than those observed at week 12 ( Supplementary Figure 4 ). No dose reductions occurred in the 50-mg Q14 days group, and reductions in LDL-C were similar at weeks 8, 10, and 12 (−34.8, −33.3, and −34.3 mg/dl, respectively; Supplementary Figure 4 ). For the 100- and 150-mg Q14 days groups, reductions in LDL-C were maximal at week 6 (−59.9 mg/dl) and 8 (−66.9 mg/dl), respectively ( Supplementary Figure 4 ). For the Q28 days regimen, changes were greatest at week 4 for both the 200-mg (−30.9 mg/dl) and the 300-mg groups (−54.9 mg/dl) ( Supplementary Figure 4 ). Reductions in LDL-C were maintained between doses in subjects who received bococizumab Q14 days but not in those receiving bococizumab Q28 days ( Figure 4 ). The individual subject LDL-C responses for subjects receiving bococizumab 150 mg Q14 days who had their dose reduced but did not miss a dose are depicted in Supplementary Figure 5 , together with the percentage of these subjects on each bococizumab dose at each dosing visit, highlighting the effect of bococizumab dose reduction on LDL-C levels.
