Evolocumab (AMG 145) for primary hypercholesterolemia

Abstract

Evolocumab is a fully human monoclonal IgG2 antibody that inhibits proprotein convertase subtilisin/kexin type 9, a protein that targets LDL receptors for degradation and thereby reduces the liver’s ability to remove LDL-C from the blood. In Phase II and III trials in more than 6000 subjects with primary hypercholesterolemia, evolocumab reduced LDL-C by 50–75% compared with placebo and by 35–45% compared with ezetimibe. Evolocumab reduced the proatherogenic lipid profile, including Lp(a), and modestly increased HDL-C and ApoA1. In subjects with homozygous familial hypercholesterolemia, evolocumab reduced LDL-C by 30%. Safety and tolerability of evolocumab was similar to that of placebo and ezetimibe. The ongoing FOURIER trial, anticipated to report in 2017, will provide definitive evidence on cardiovascular endpoints and additional long-term safety.

Keywords:

• cardiovascular disease
• evolocumab
• hypercholesterolemia
• LDL cholesterol
• PCSK9

The term primary hypercholesterolemia denotes a heterogeneous group of conditions with elevated LDL-C levels. It comprises both autosomal-dominant familial hypercholesterolemia (ADH) with an estimated frequency of 1 per 200–500, and the more common polygenic non-familial hypercholesterolemia, often in combination with adverse environmental factors, such as a hypercholesterolemic diet and obesity. Epidemiological studies have shown a graded relation between the risk of coronary heart disease (CHD) and cholesterol level [1,2], and clinical trials have demonstrated that LDL-C reduction decreases cardiovascular morbidity and mortality [3–6]. International treatment guidelines therefore recommend lowering LDL-C in hypercholesterolemic patients, depending on the individual’s estimated risk for cardiovascular disease, also taking into account other risk factors [7,8].

When added to diet and lifestyle, statins are the current drug of choice for the treatment of hypercholesterolemia. European guidelines from 2011 recommend that LDL-C levels should be below 1.8 mmol/l in high- and very high-risk patients and below 2.5 mmol/l in moderate-risk patients [7]. The American Heart Association/American College of Cardiology guidelines from 2013 recommend treating high-risk patients with high-intensity statins with the aim of lowering LDL-C by at least 50%, and to treat moderate-risk patients with moderate-intensity statins with the aim of lowering LDL-C by 30–50% [8]. Achieving recommended treatment targets can be difficult with current treatment options, especially in familial hypercholesterolemia [9]. In addition, it is estimated that 5–10% of patients are unable to tolerate any statin or an effective dose of a statin mainly because of statin-associated muscle symptoms [10–13].

Serum levels of LDL-C are determined by several factors, including dietary saturated fats, which upregulate cholesterol synthesis in the liver, and uptake of LDL particles in the liver cells by LDL receptors on the surface of hepatocytes. The LDL receptors are recirculated in the hepatocytes, but are eventually degraded. In 2003, the glycoprotein proprotein convertase subtilisin/kexin type 9 (PCSK9) was discovered [14], and shortly after it was reported that ‘gain of function’ mutations in the PCSK9 gene could cause familial hypercholesterolemia [15]. PCSK9, which is synthesized and secreted by the liver, was determined to be involved in the degradation of the LDL receptor, binding to the receptor in serum and targeting it for degradation intracellularly in the lysosome [16,17]. This discovery led to the development of therapeutic monoclonal antibodies that specifically bind to circulating PCSK9, neutralizing the protein and thereby inhibiting degradation of the LDL receptor [18]. Currently, three monoclonal antibodies targeting PCSK9 are in Phase III development, evolocumab, alirocumab and bococizumab. Evolocumab is a fully human IgG2 monoclonal antibody, alirocumab is a fully human IgG1 monoclonal antibody and bococizumab is a humanized monoclonal antibody against PCKS9. In Phase I and Phase II studies, treatment with anti-PCSK9 antibodies were highly effective in reducing LDL-C levels, with few side effects [19–29]. In addition, other approaches to inhibiting PCSK9 are in pre-clinical and Phase I development, such as PCSK9-binding peptides and small proteins (adnectins) as well as antisense and small interfering RNA (siRNA), interfering with PCSK9 mRNA and inhibiting translation [30–33].

Introduction to evolocumab

Evolocumab is a fully human monoclonal IgG2 antibody with high binding affinity for PCSK9 (K_D = 16 pM). As a human IgG2 antibody, the clearance of evolocumab is mediated by both nonlinear and linear pathways. The nonlinear pathway is because of specific binding and complex formation with its target ligand, PCSK9, and therefore predominates at low evolocumab serum concentrations when there is insufficient evolocumab to completely bind all circulating PCSK9. When serum-unbound evolocumab is in excess, the nonlinear pathway is saturated and the majority of clearance is mediated by the typical IgG clearance processes in the reticuloendothelial system. This process results in linear elimination. In both instances, evolocumab is eventually degraded into small peptides and amino acids via these catabolic pathways.

The relationship between unbound evolocumab, unbound PCSK9 and LDL-C levels in the serum following a single subcutaneous (SC) dose is illustrated in Figure 1. Absorption from a SC dose of evolocumab gives rise to peak serum concentrations in 3–4 days after administration. Within 4 h of SC administration of evolocumab, there is rapid and maximal suppression of circulating unbound PCSK9 to undetectable levels. This suppression is followed by a dose-dependent reduction in LDL-C that reaches a maximum lowering effect with evolocumab doses ≥140 mg; higher doses extend the duration of lowering. The mean nadir in LDL-C response occurs between 7 and 21 days. PCSK9 continues to be produced and secreted by the liver, contributing to the elimination of evolocumab. The amount of unbound evolocumab eventually declines and unbound PCSK9 reappears in the circulation followed by a return in LDL-C toward baseline because of the PCSK9 role in LDL receptor degradation.

Figure 1. Relationship between unbound evolocumab (AMG 145), unbound PCSK9 and LDL-C following a 140 or 420 mg subcutaneous dose of evolocumab. Reproduced with permission from Amgen.

In clinical studies with repeated SC dosing of evolocumab over 12 weeks, dose-proportional increases in exposure were observed with dose regimens of ≥140 mg. No time-dependent changes were observed in unbound evolocumab concentrations over a period of 52 weeks and changes in unbound PCSK9 and serum lipoproteins were reversible upon discontinuation of evolocumab. No increase in unbound PCSK9 or LDL-C above baseline was observed after cessation of evolocumab, suggesting that compensatory mechanisms to increase production of PCSK9 and LDL-C do not occur during treatment. Population pharmacokinetic analyses suggest that no dose adjustments are necessary for age (18–79 years), gender, race/ethnicity, statin therapy, baseline PCSK9 and body weight.

Evolocumab dose regimens in clinical efficacy studies deliver approximately 80% of the theoretical maximum LDL-C lowering from PCSK9 inhibition by this mechanism. Unlike therapies that are given on a daily basis, convenient biweekly or monthly SC dosing schedules for hypercholesterolemic patients on current anti-PCSK9 monoclonal antibody therapies under investigation, like evolocumab, give rise to u-shaped LDL-C lowering over the dosing interval with deeper u-shapes and more variability observed with lower doses. The magnitude of these changes is dependent on the dose, dose interval and specific monoclonal antibody binding characteristics toward PCSK9. Thus, the average LDL-C lowering over the dosing interval is an important measure of efficacy given the pharmacodynamic profile of anti-PCSK9 antibodies.

Clinical efficacy

Phase II trials

An extensive Phase II program evaluating the safety and efficacy of evolocumab in primary hypercholesterolemia (familial and non-familial) was completed (Table 1) [22–26,28,34]. The studies assessed dosing regimens ranging from 70 to 140 mg SC every 2 weeks (Q2W) and/or 280 to 420 mg SC monthly (QM) in combination with statins with or without ezetimibe and as an adjunct to diet alone in hypercholesterolemic subjects with cardiovascular disease or at cardiovascular risk. In particular, the safety and efficacy of evolocumab was studied in combination with statins in patients with heterozygous familial hypercholesterolemia (HeFH), homozygous familial hypercholesterolemia (HoFH), non-familial hypercholesterolemia and in high-risk Japanese subjects, in statin-intolerance, and as an adjunct to diet (i.e., monotherapy). The 12-week duration Phase II studies were all double-blind, randomized, placebo- and/or ezetimibe-controlled trials except for TESLA Part A, which was a single arm, open-label pilot study to determine whether evolocumab would lower LDL-C in HoFH subjects. The long-term, open-label study, OSLER, allowed patients completing LAPLACE-TIMI 57, RUTHERFORD, GAUSS, MENDEL or YUKAWA to be randomized 2:1 to evolocumab plus standard-of-care (SoC) or SoC alone for 1 year and then receive evolocumab plus SoC for up to four additional years. The open-label extension trial, TAUSSIG [35], facilitated collection of long-term safety and efficacy data in HoFH subjects completing TESLA Part A as well as de novo HoFH and HeFH subjects, including those on apheresis.

Table 1. Phase II trial summaries.

	LAPLACE-TIMI 57		RUTHERFORD	GAUSS	MENDEL		YUKAWA	OSLER		TESLA Part A
No. of subjects	629		167	157	406		307	1324		8
Duration	12 weeks		12 weeks	12 weeks	12 weeks		12 weeks	Up to 5 years		12 weeks
Therapeutic setting	Statin combination (familial and non-familial)		HeFH	Statin-intolerant (familial and non-familial)	Adjunct to diet (familial and non-familial)		Statin combination (familial and non-familial)	Completed Phase II parent study^†		HoFH
Fasting LDL-C	≥85 mg/dl (2.2 mmol/l)		≥100 mg/dl (2.6 mmol/l)	≥100 mg/dl (2.6 mmol/l)	≥100 mg/dl (2.6 mmol/l)		≥115 mg/dl (3.0 mmol/l)	≥85 mg/dl (2.2 mmol/l)		≥130 mg/dl (3.4 mmol/l)
Background therapy	Statin (± ezetimibe)		Statin (± ezetimibe)	Non-ezetimibe lipid-lowering therapy^‡	None		Statin (± ezetimibe)	SoC		Lipid-lowering therapies (e.g., statin ± ezetimibe; no apheresis)
Evolocumab doses	70, 105, 140 mg Q2W; 280, 350, 420 mg QM		350, 420 mg QM	280, 350, 420 mg QM	70, 105, 140 mg Q2W; 280, 350, 420 mg QM		70 and 140 mg Q2W; 280 and 420 mg QM	420 mg QM		420 mg QM
Control	Placebo SC		Placebo SC	Placebo SC + ezetimibe PO QD	Placebo SC or ezetimibe PO QD		Placebo SC	SoC		None
Efficacy	140 mg Q2W	420 mg QM	420 mg QM	420 mg QM	140 mg Q2W	420 mg QM	140 mg Q2W	420 mg QM	420 mg QM	420 mg QM
% change in LDL-C (calc) from baseline to week 12 relative to placebo	–69%	–54%	–59%	NA	–52%	–57%	–72%	–66%	–59%/ –54% compared to baseline^§	–17% compared to baseline
% change in LDL-C (calc) from baseline to week 12 relative to ezetimibe	NA	NA	NA	–38%	–37%^¶	–34%^¶	NA	NA	NA	NA

^†Subjects who completed one of the five qualifying studies LAPLACE-TIMI 57, RUTHERFORD, GAUSS, MENDEL or YUKAWA were eligible to participate in the extension study.

^‡At screening, low or atypical dose statin permitted: atorvastatin ≤70 mg, simvastatin ≤140 mg, pravastatin ≤140 mg, rosuvastatin ≤35 mg, lovastatin ≤140 mg or fluvastatin ≤280 mg weekly.

^§LDL-C percent change from baseline to week 12/52 (not SOC-adjusted).

^¶LDL-C by ultracentrifugation percent change from baseline to week 12.

HeFH: Heterozygous familial hypercholesterolemia; HoFH: Homozygous familial hypercholesterolemia; NA: Not applicable; PO: Per os (orally); Q2W: Once every 2 weeks; QD: Once daily; QM: Once monthly; SC: Subcutaneously; SoC: Standard of care.

References [22–26,28,34].

In nearly 1700 Phase II subjects with primary non-familial and heterozygous familial hypercholesterolemia, evolocumab 140 mg Q2W and 420 mg QM resulted in significant reductions in LDL-C of approximately 50–70% compared with placebo and approximately 35–40% compared with ezetimibe (Table 1) [22–26,28]. Significant reductions in proatherogenic lipid parameters, including Lp(a), and increases in anti-atherogenic lipid parameters were observed. Safety and tolerability for all dosing regimens evaluated were similar and comparable with placebo or ezetimibe among all of the populations studied. No clinically meaningful differences in efficacy or safety were observed for the 140 mg Q2W and 420 mg QM SC dosing regimens. It was concluded that these two dosing regimens were clinically equivalent and were chosen to be further studied in Phase III. Finally, the effects on LDL-C and other lipid parameters observed at week 12 in OSLER were maintained with long-term evolocumab administration and were reversible upon cessation of treatment with no evidence of rebound [24].

In eight subjects with HoFH in TESLA Part A, a significant reduction in LDL-C of 17% compared with baseline was observed with evolocumab; however, no reduction in LDL-C was observed in two receptor-negative subjects [34]. The safety and tolerability profile of evolocumab in these eight subjects was consistent with that seen in the primary non-familial and heterozygous familial hypercholesterolemia population. Given this efficacy and safety profile in HoFH, 420 mg QM was studied in the double-blind, randomized, placebo-controlled Phase III TESLA Part B trial [36].

Phase III trials

The completed Phase III program along with the ongoing open-label extension study OSLER-2 [37] is outlined in Table 2 [36,38–42]. These studies enrolled over 4000 subjects with primary non-familial and heterozygous familial hypercholesterolemia as well as approximately 50 HoFH subjects. The 12- and 52-week duration randomized, double-blind, placebo- and/or ezetimibe-controlled studies evaluated the safety and efficacy of evolocumab on high- or moderate-intensity statins, other lipid-lowering therapies, and diet alone. Subjects included those with HeFH, HoFH, statin-intolerance and non-familial hypercholesterolemia. Subjects completing the primary non-familial and heterozygous familial hypercholesterolemia Phase III trials were eligible to enroll in the long-term extension study, OSLER-2, where the first year included a SoC control arm. Patients with HoFH completing TESLA Part B were eligible to participate in the one-arm open-label extension study TAUSSIG.

Table 2. Evolocumab Phase III trial summaries.

Study	LAPLACE-2		RUTHERFORD-2		GAUSS-2		MENDEL-2		DESCARTES	OSLER-2 (NCT01854918) [37]	TESLA Part B	TAUSSIG (NCT01624142) [35]
No. of subjects	1896		329		307		614		901	∼3515	49	∼310
Duration	12 weeks		12 weeks		12 weeks		12 weeks		52 weeks	Up to 2 years	12 weeks	Up to 5 years
Therapeutic setting	Statin combination (familial and non-familial)		HeFH		Statin-intolerant (familial and non-familial)		Monotherapy (familial and non-familial)		Range of CV risk (familial and non-familial)	Completed Phase III parent study^† (familial and non-familial)	HoFH	Severe FH (including HoFH with or without apheresis)
Fasting LDL-C	≥80 mg/dl (2.1 mmol/l)		≥100 mg/dl (2.6 mmol/l)		≥ 100 mg/dl (2.6 mmol/l)		≥ 100 mg/dl (2.6 mmol/l)		≥75 mg/dl (1.9 mmol/l)	≥75 mg/dl (1.9 mmol/l)	≥130 mg/dl (3.4 mmol/l)	≥100 mg/dl (2.6 mmol/l)
Background therapy	Statin^‡		Statin (± ezetimibe)		Non-ezetimibe lipid-lowering therapy^§		None		Diet ± atorvastatin ± ezetimibe^¶	SoC	Lipid-lowering therapies (e.g., statin ± ezetimibe; no apheresis)	Lipid-lowering therapy (e.g., statin ± ezetimibe; apheresis permitted)
Evolocumab doses	140 mg Q2W	420 mg QM	140 mg Q2W	420 mg QM	140 mg Q2W	420 mg QM	140 mg Q2W	420 mg QM	420 mg QM	140 mg Q2W/420 mg QM	420 mg QM	420 mg QM or Q2W
Control	Placebo SC and ezetimibe PO QD^#,††		Placebo SC		Ezetimibe PO QD^#		Placebo SC and ezetimibe PO QD^#		Placebo SC	SoC	Placebo SC	NA

^†Subjects completing studies DESCARTES, MENDEL-2, LAPLACE-2, GAUSS-2 and RUTHERFORD-2 were eligible to enroll.

^‡Randomization: rosuvastatin 5 or 40 mg PO QD, atorvastatin 10 or 80 mg PO QD or simvastatin 40 mg PO QD weekly.

^§At screening, low or atypical dose statin permitted: atorvastatin ≤70 mg, simvastatin ≤140 mg, pravastatin ≤140 mg, rosuvastatin ≤35 mg, lovastatin ≤140 mg or fluvastatin ≤280 mg.

^¶Subjects assigned background lipid-lowering therapy based upon their screening LDL-C and National Cholesterol Education Program ATP III risk category: diet alone, diet plus atorvastatin 10 mg PO QD, diet plus atorvastatin 80 mg PO QD or diet plus atorvastatin 80 mg PO QD and ezetimibe 10 mg PO QD.

^#Double-dummy: ezetimibe PO: ezetimibe 10 mg PO QD and placebo SC (Q2W or QM).

^††Ezetimibe control in study 20110115 was only used in subjects who were assigned to atorvastatin 10 or 80 mg PO QD.

CV: Cardiovascular; HeFH: Heterozygous familial hypercholesterolemia; HoFH: Homozygous familial hypercholesterolemia; NA: Not applicable; PO: Per os (orally); Q2W: Once every 2 weeks; QD: Once daily; QM: Once monthly; SC: Subcutaneously; SoC: Standard of care.

References [36,38–42].

Similar to the Phase II results, the Phase III data demonstrate that evolocumab 140 mg Q2W and 420 mg QM resulted in significant reductions in LDL-C of approximately 55–75% compared with placebo (Table 3) [36,38–41] and approximately 35–45% compared with ezetimibe (Table 4). This consistent response was observed regardless of baseline characteristics such as age, race, gender, body mass index, cardiovascular risk, PCSK9 level and statin dose/intensity [40–42]. No clinically meaningful differences between evolocumab doses of 140 mg Q2W and 420 mg QM were observed with respect to reducing LDL-C, total cholesterol, ApoB, non-HDL-C, VLDL-C, triglycerides, Lp(a), total cholesterol/HDL-C and ApoB/ApoA1, and increasing HDL-C and ApoA1 (Tables 3 & 4). As observed in OSLER, the effect of evolocumab on LDL-C reduction and improvements in other lipid parameters were maintained with long-term evolocumab administration [24,39].

Table 3. Effect of evolocumab compared with placebo on lipid parameters in the Phase III trials.

Background	Without statin		With statin ± ezetimibe				Risk-based therapy	HoFH with statin ± ezetimibe
Study	MENDEL-2		LAPLACE-2		RUTHERFORD-2		DESCARTES	TESLA Part B
Treatment	140 mg Q2W	420 mg QM	140 mg Q2W	420 mg QM	140 mg Q2W	420 mg QM	420 mg QM	420 mg QM
Least square mean percent change from baseline versus placebo at week 12
LDL-C calc	–59%	–57%	–73%	–64%	–61%	–60%	–58/59%^†	–32%
Least square mean percent change from baseline versus placebo at mean of weeks 10 and 12
LDL-C calc	–57%	–60%	–72%	–69%	–61%	–66%	NA	–31%^‡
TC	–35%	–37%	–41%	–40%	–42%	–44%	–33%
Non-HDL-C	–49%	–53%	–60%	–60%	–56%	–60%	–50%
ApoB	–47%	–51%	–56%	–56%	–49%	–55%	–44%	–23%^‡
Lp(a)	–25%	–26%	–30%	–27%	–31%	–31%	–22%	–11%^‡
TG	0%	–22%	–17%	–23%	–22%	–17%	–12%	–3.3%^‡
VLDL-C	0%	–22%	–18%	–22%	–23%	–16%	29%
HDL-C	6%	9%	6%	8%	8%	9%	5%	1.3%^‡
ApoA1	3%	5%	3%	5%	7%	5%	3%

^†LDL-C percent change from baseline to week 12/52 relative to placebo; week 12 value is based on preparative ultracentrifugation.

^‡LDL-C percent change from baseline to mean of weeks 6 and 12; week 10 values not available.

References [36,38–41].

Table 4. Effect of evolocumab compared with ezetimibe on lipid parameters in the Phase III trials.

Background	Without statin				With statin ± ezetimibe
Study	MENDEL-2		GAUSS-2		LAPLACE-2
Treatment	140 mg Q2W	420 mg QM	140 mg Q2W	420 mg QM	140 mg Q2W	420 mg QM
Least square mean percent change from baseline versus ezetimibe at week 12
LDL-C calc	–40%	–38%	–39%	–38%	–45%	–42%
Least square mean percent change from baseline versus ezetimibe at mean of weeks 10 and 12
LDL-C calc	–40%	–41%	–38%	–39%	–43%	–46%
TC	–25%	–25%	–24%	–26%	–23%	–25%
Non-HDL-C	–36%	–35%	–32%	–35%	–34%	–39%
ApoB	–34%	–35%	–32%	–35%	–34%	–40%
Lp(a)	–22%	–20%	–24%	–25%	–30%	–33%
TG	–9%	–9%	–3%	–6%	–2%	–8%
VLDL-C	–7%	–10%	–2%	–4%	–1%	–7%
HDL-C	6%	4%	5%	6%	7%	8%
ApoA1	3%	4%	5%	3%	7%	7%

References [40–42].

In 49 subjects with HoFH, a Phase III double-blind, placebo-controlled study (TESLA Part B) demonstrated statistically significant reductions in LDL-C of approximately 30% compared with placebo [36]. Improvements in other lipid parameters (total cholesterol, ApoB and non-HDL-C) were also observed. Early data in the long-term extension study, TAUSSIG, suggest that these changes in lipid parameters were maintained with long-term treatment in HoFH subjects [43]. Furthermore, it appears that in HoFH subjects, who typically have higher baseline PCSK9 levels, a dose of 420 mg Q2W provides additional LDL-C reduction compared with 420 mg QM. This observation will require further validation in the ongoing TAUSSIG study.

Safety & tolerability

The safety and tolerability of evolocumab in the nearly 1700 hyperlipidemic subjects from the seven Phase II trials described in Table 1 were published previously [22–26,28,34,44]. In six Phase III trials that enrolled over 4000 non-familial and familial hyperlipidemic subjects, the safety and tolerability of evolocumab is similar to that of the comparator (placebo and/or ezetimibe) and with that observed in the Phase II program [36,38–42]. The overall incidence of treatment-emergent adverse events (TEAEs) in patients receiving evolocumab was comparable with those receiving control (placebo and/or ezetimibe). Among the 12-week duration studies, the incidence of TEAEs was modestly higher, but balanced, in the statin-intolerance study, GAUSS-2 (ezetimibe 73%; evolocumab 66%) compared with the other 12-week studies (placebo 39, 44 and 49%; ezetimibe 40 and 45%; evolocumab 36, 44 and 56%). In the 52-week DESCARTES trial, the incidence of TEAEs was balanced between evolocumab (75%) and placebo (74%). The incidence of serious AEs was low and balanced ranging from 1–3% on evolocumab to 1–5% on placebo or ezetimibe in the 12-week duration studies. In the 52-week DESCARTES study, rates of serious AEs were 6 and 4% in the evolocumab and placebo arms, respectively. TEAE that led to discontinuation of investigational product were low (≤2%) in the evolocumab and comparator (≤4%) arms except in the statin-intolerance study where the rate was 8% in the evolocumab arm and 13% in the ezetimibe arm.

Evaluation of the most common TEAEs (≥6%) across the trials did not identify safety issues (Table 5) [36,38–42]. TEAEs were generally balanced between evolocumab and comparator. Of note, the incidence of myalgia was balanced and low (<4%) in the 12- and 52-week duration trials except in the double-blind, double-dummy statin-intolerant study GAUSS-2, where the incidence of myalgia on evolocumab was 8% compared with 18% on ezetimibe. A standardized Medical Dictionary for Regulatory Activities (MedDRA) search strategy for TEAEs that could be muscle-related AEs revealed a pattern similar to what was observed with the adverse event of myalgia alone. Although evolocumab is administered as a SC injection, a standardized MedDRA search strategy was performed for possible injection site reactions. Using this search strategy, the incidence of TEAEs that could be injection site reactions ranged from 1 to 8% on control and 1 to 6% on evolocumab. No clinically meaningful difference was observed between control and evolocumab. Given concerns around statins, low LDL-C and neurocognitive dysfunction, a standardized MedDRA search strategy for TEAEs that could be related to neurocognitive dysfunction was performed. These events were rare; no clinically meaningful difference was noted between control and evolocumab.

Table 5. Subject incidence of safety and tolerability of evolocumab in Phase III trials.

Study	LAPLACE-2			RUTHERFORD-2		GAUSS-2		MENDEL-2			DESCARTES		TESLA Part B
Treatment	Pbo (n = 558) n (%)	Eze (n = 221) n (%)	EvoMab (n = 1117) n (%)	Pbo (n = 109) n (%)	EvoMab (n = 220) n (%)	Eze (n = 102) n (%)	EvoMab (n = 205) n (%)	Pbo (n = 154) n (%)	Eze (n = 154) n (%)	EvoMab (n = 306) n (%)	Pbo (n = 302) n (%)	EvoMab (n = 599) n (%)	Pbo (n = 16) n (%)	EvoMab (n = 33) n (%)
Treatment-emergent adverse events
Any	219 (39)	89 (40)	406 (36)	53 (49)	124 (56)	74 (73)	135 (66)	68 (44)	70 (45)	134 (44)	224 (74)	448 (75)	10 (63)	12 (36)
Serious	13 (2)	2 (1)	23 (2)	5 (5)	7 (3)	4 (4)	6 (3)	1 (1)	1 (1)	4 (1)	13 (4)	33 (6)	0	0
Leading to discontinuation of investigational product	12 (2)	4 (2)	21 (2)	0	0	13 (13)	17 (8)	6 (4)	5 (3)	7 (2)	3 (1)	13 (2)	0	0
Common (≥6%) treatment-emergent adverse events
Myalgia	11 (2)	4 (2)	12 (1)	0	6 (3)	18 (18)	16 (8)	3 (2)	3 (2)	3 (1)	9 (3)	24 (4)	0	0
Nausea	8 (1)	1 (<1)	8 (1)	1 (1)	8 (4)	7 (7)	9 (4)	1 (1)	3 (2)	8 (3)	10 (3)	20 (3)	2 (13)	0
Nasopharyngitis	8 (1)	4 (2)	15 (1)	5 (5)	19 (9)	3 (3)	7 (3)	3 (2)	6 (4)	6 (2)	29 (10)	63 (11)	0	2 (6)
Fatigue	5 (1)	2 (1)	16 (1)	1 (1)	0	10 (10)	9 (4)	1 (1)	3 (2)	4 (1)	9 (3)	13 (2)	1 (6)	0
Upper respiratory tract infection	6 (1)0	3 (1)	15 (1)	3 (3)	7 (3)	0	3 (1)	4 (3)	5 (3)	5 (2)	19 (6)	56 (9)	1 (6)	3 (9)
Headache	15 (3)	5 (2)	19 (2)	4 (4)	9 (4)	9 (9)	16 (8)	4 (3)	5 (3)	10 (3)	11 (4)	24 (4)	1 (6)	0
Influenza	2 (<1)	2 (1)	3 (<1)	0	7 (3)	3 (3)	1 (<1)	1 (1)	2 (1)	2 (1)	19 (6)	45 (8)	0	3 (9)
Pain in extremity	7 (1)	3 (1)	17 (2)	3 (3)	5 (2)	1 (1)	14 (7)	1 (1)	1 (1)	3 (1)	13 (4)	13 (2)	0	0
Muscle spasms	6 (1)	6 (3)	17 (2)	2 (2)	4 (2)	4 (4)	13 (6)	1 (1)	1 (1)	2 (1)	8 (3)	14 (2)	0	0
Back pain	14 (3)	7 (3)	20 (2)	1 (1)	9 (4)	3 (3)	9 (4)	3 (2)	1 (1)	5 (2)	17 (6)	37 (6)	1 (6)	0
Gastroenteritis	3 (1)	0	4 (<1)	1 (1)	0	0	2 (1)	2 (1)	2 (1)	1 (<1)	6 (2)	18 (3)	0	2 (6)
Events associated with lipid-lowering therapies
Muscle-related adverse events^†	17 (3)	8 (4)	23 (2)	1 (1)	10 (5)	23 (23)	25 (12)	6 (4)	5 (3)	8 (3)	20 (7)	55 (9)	0	0
Events associated with injectable protein therapies
Injection site reaction^‡	8 (1)	2 (1)	15 (1)	4 (4)	13 (6)	8 (8)	6 (3)	8 (5)	7 (5)	15 (5)	15 (5)	34 (6)	1 (6)	o
Positively adjudicated clinical cardiovascular events^§
All	2 (<1)	0	5 (<1)	0	3 (1)	0	0	0	0	0	2 (1)	6 (1)	0	0
Laboratory assessments
CK > 5x ULN^¶	2 (<1)	0	1 (<1)	2 (2)	0	3 (3)	2 (1)	2 (1)	0	2 (1)	1 (<1)	7 (1)	1 (6)	1 (3)
AST or ALT > 3x ULN^¶	6 (1)	3 (1)	4 (<1)	0	0	0	0	5 (3)	1 (1)	3 (1)	3 (1)	5 (1)	1 (6)	2 (6)
Anti-evolocumab binding antibodies^#	NA	NA	0 (0)	NA	0	NA	0	NA	NA	0	0	1 (<1)	0	0
Anti-evolocumab neutralizing antibodies	NA	NA	0	NA	0	NA	0	NA	NA	0	0	0	0	0

^†Defined by MedDRA search terms that include myositis, myalgia, musculoskeletal pain, muscular weakness, increased plasma creatinine and blood CK increased.

^‡Defined by MedDRA search terms that include injection site bruising, erythema, reaction, hematoma, hemorrhage, inflammation, rash, swelling, irritation and pain.

^§Death, myocardial infarction, hospitalization for unstable angina, coronary revascularization, stroke, transient ischemic attack or congestive heart failure requiring hospital admission.

^¶At any post baseline visit.

^#New antidrug antibody at any postbaseline visit.

AST: Aspartate aminotransferase; ALT: Alanine aminotransferase; CK: Creatine kinase; NA: Not applicable; NR: Not reported; ULN: Upper limit of normal.

References [36,38–42].

Laboratory analyses did not identify safety issues. Elevation of creatine kinase (CK) and increases in transaminases were infrequent and balanced between evolocumab and control. In the Phase III evolocumab program, there were no cases of anti-evolocumab neutralizing antibodies and the development of anti-evolocumab binding antibodies was rare. Among the 2480 subjects receiving evolocumab in Phase III, only one subject tested positive for the development of new anti-evolocumab binding antibodies. The absence of neutralizing antibodies to evolocumab and the rare (<1%) observation of binding antibodies is consistent with what was observed in the Phase I and II evolocumab trials.

Cardiovascular events, composed of death by any cause, cardiovascular death, myocardial infarction, hospitalization for unstable angina, coronary revascularization, stroke, hospitalization for heart failure or transient ischemic attack were adjudicated. These events were infrequent. There was no significant imbalance in events between control and evolocumab [45].

Conclusion

The accumulated data with evolocumab presented here in nearly 6000 subjects with primary non-familial and heterozygous familial hypercholesterolemia demonstrates that evolocumab consistently reduces LDL-C by 50–75% compared with placebo and by 35–45% compared with ezetimibe. This result is consistent across the populations studied which ranged from subjects with atherosclerotic cardiovascular disease on high-intensity statin therapy with ezetimibe to subjects with cardiovascular risk being managed by diet alone. Evolocumab reduced the proatherogenic lipid profile, including a 20–33% reduction in Lp(a), and resulted in modest increases in HDL-C and ApoA1. In receptor-defective HoFH subjects, evolocumab reduced LDL-C by approximately 30% compared with placebo; favorable, clinically meaningful changes were noted in the lipid profile in HoFH subjects, although the magnitude of these changes was less marked than in non-HoFH subjects. One patient with two receptor negative mutations and one patient with autosomal recessive HoFH did not respond to evolocumab treatment. This observation appears to be consistent with the mechanism of action, which requires some degree of LDL-receptor functionality.

Based on the accumulated safety data in the Phase II and Phase III program, the safety profile of evolocumab appears similar to comparator. TEAE were balanced between evolocumab and control with the overall incidence of TEAE leading to discontinuation of investigational product and serious AEs being low and balanced. A careful evaluation of muscle-related AEs, abnormalities in CK and transaminases and anti-evolocumab antibodies was unremarkable. Although the number of cardiovascular events was modest, there does not appear to be evidence of harm and the first examination of long-term data in OSLER was encouraging [24]. Definitive evidence will be provided from the ongoing FOURIER trial [46].

Expert commentary

PCSK9-inhibition with monoclonal antibodies is a new, powerful tool to reduce LDL-C in high-risk patients. The changes in the lipid profile compare favorably to what is achieved with high-intensity statin therapy with the exception of PCSK9-mediated reduction of Lp(a). Importantly, the impact of LDL-C reduction by PCSK9 inhibition on cardiovascular outcomes is unknown and currently under investigation in large clinical trials.

Current data supports the cardiovascular benefits of marked LDL-C reduction. There are observational data that proatherogenic non-LDL-C lipids may contribute to the pathogenesis of atherosclerosis and clinical manifestations of cardiovascular disease. There has been discussion of the impact of remnant particles as well as Lp(a) on cardiovascular outcomes. With regards to remnant particles, current data with evolocumab indicate that remnant particles will be reduced as well. There are some genetic data that suggest that other proatherogenic lipids, particularly Lp(a), may contribute to manifestations of atherosclerosis, particularly as it relates to aortic stenosis. The impact of PCSK9 is most marked on LDL-C reduction and it is clear that LDL-C is a major modifiable cardiovascular risk factor. Thus, it will be challenging to dissect the proven cardiovascular benefits of marked LDL-C reduction from the other potentially beneficial effects on lipids, such as the reduction in Lp(a) of approximately 20–30% and the reduction in triglycerides of up to approximately 20%. Studies specifically recruiting hypertriglyceridemic patients have not been performed, but the drug could possibly also become a treatment option in difficult-to-treat hypertriglyceridemic patients.

To date, the safety and tolerability in the trials show adverse event profiles comparable between evolocumab and comparators and low rates of discontinuation and serious AEs. Careful evaluation of the safety database has not revealed evidence of muscle-related side effects, neurocognitive or glycemic impairment, issues with anti-evolocumab antibodies or laboratory evidence of muscle or liver injury. Nevertheless, evolocumab and other anti-PCSK9 antibodies allow for achievement of LDL-C levels not previously attainable. Although there are data beyond 1 year, the long-term effects of low and very low levels of LDL-C are unknown. PCSK9 is also expressed in the nervous system, intestines and pancreas and long-term PCSK9 inhibition could possibly affect these organ systems.

Concerns emerged related to an increased risk of diabetes and neurocognitive dysfunction associated with statin therapy. Thus, theoretical safety issues for long-term PCSK9 inhibition relating to cellular membrane integrity, levels of fat-soluble vitamins, steroid hormones, diabetes and possible neurological side effects need to be addressed. In the 52-week results from the OSLER open-label study, overall TEAEs, serious AEs and elevations of CK and aminotransferases were not appreciably greater in those who achieved low (<1.3 mmol/l) and very low (<0.65 mmol/l) LDL-C levels, compared with those with LDL-C ≥1.3 mmol/l [24]. Analyses of OSLER and DESCARTES have not identified safety issues with fat-soluble vitamins, steroid hormone biosynthesis, glycemic parameters or neurocognitive function. These observations regarding vitamin and steroid hormones are not unexpected because intracellular cholesterol originates primarily via intracellular synthesis and is not reliant on circulating LDL-C. Heterozygous PCSK9 loss-of-function mutations, observed in approximately 1–4% of the population, have been associated with a reduced risk of CHD, whereas increased morbidity for other conditions, such as cognitive performance, has not been observed [47,48]. Furthermore, a woman with loss-of-function PCSK9 mutations in both genes (compound heterozygote), and LDL-C of approximately 0.4 mmol/l is healthy, fertile and without evidence of adverse consequences of life-long very low LDL-C [49]. While these observations are encouraging, larger and long exposure to PCSK9 inhibition and achievement of very low LDL-C levels will enhance our confidence in clinical and observational data.

Anti-PCSK9 antibodies are injectable protein therapeutics and as such carry the risk of anti-drug antibody development, allergic reactions and injection site reactions. The incidence of anti-evolocumab binding antibodies is rare and there were no neutralizing antibodies. The rare occurrence of anti-evolocumab binding antibodies was not associated with allergic reactions or impaired pharmacokinetics or pharmacodynamics in the clinical program. Clinical trial data show that the incidence of injection site reactions was low and similar to placebo. The majority of injection site reactions were mild and did not impact adherence to therapy.

Five-year view

Large cardiovascular outcomes trials with PCSK9 inhibitors are being conducted and results are anticipated as early as 2017. These studies will be instrumental in defining the place for PCSK9 inhibitors in the treatment of hypercholesterolemia. Until then, if approved and available as anticipated in 2015–2016, these drugs will probably be reserved for use in difficult-to-treat patients, that is, the substantial number of ADH patients not responding sufficiently to treatment with statins and ezetimibe [50–53]. Patients with ADH and CHD should be treated to LDL-C levels below 1.8 mmol/l (70 mg/dl), which is extremely difficult to achieve in these high-cholesterol patients. If approved, the drug will probably also be used in secondary prevention in other difficult-to-treat primary hypercholesterolemia patients, such as those with recurrent cardiovascular events, as well as in the 5–10% of patients who are unable to tolerate statins. While much has been made of the challenge of defining statin-intolerant patients, the approach employed in clinical trials to date has been pragmatic based on what clinicians do when encountering patients with statin-associated muscle symptoms. Physicians switch to another statin and usually try several statins before endeavoring to find a low or atypical dose regimen that a patient can tolerate. To date, these trials enrolled subjects with a mean LDL-C of approximately 4.9 mmol/l (190 mg/dl) and significant cardiovascular risk [42,54]. PCSK9 inhibitors may represent an impactful solution for these patients.

Evolocumab and similar PCSK9 inhibitors open the possibility to lower LDL-C to levels well beyond current treatment targets. The ongoing endpoint studies should provide data to address questions about the level of low LDL-C associated with clinical benefit and the long-term safety of low LDL-C. If these trials show a substantial benefit in cardiovascular disease reduction and there are no major long-term safety issues, these drugs could turn out to be the next lipid-lowering drug of choice after statins.

Cost of the PCSK9-inhibitors have not been determined, but could be anticipated to be on par with recombinant monoclonal antibodies for use in other therapeutic areas. Competition between pharmaceutical companies and focus on cost–effectiveness will hopefully reduce the price. Widespread use is not medically warranted until results from the large outcomes studies become available. If substantial benefits in CVD reduction are achieved, treatment with PCSK9 inhibitors in a wider group of patients could prove to be cost-effective.

Given the excitement around the PCSK9 target and its biology, the alternative mechanisms to PCSK9 inhibition will continue to be explored.

Key issues

• Proprotein convertase subtilisin/kexin type 9 (PCSK9) is a circulating protein secreted by the liver that regulates LDL-C uptake by binding to the LDL-C receptor, targeting the receptor for degradation in the lysosomes.

• Evolocumab is a fully human monoclonal IgG2 antibody with high binding affinity for PCSK9 that binds PCSK9 and prevents lysosomal degradation of the LDL-C receptor.

• Evolocumab reduced LDL-C by 50–75% compared with placebo and by 35–45% compared with ezetimibe in Phase II and III studies.

• Safety and tolerability of evolocumab in trials of up to 52 weeks duration were similar to that of placebo and ezetimibe.

• Results from FOURIER, a large cardiovascular endpoint trial, are anticipated no later than 2017.

• If cardiovascular endpoints are favorable and no long-term safety and tolerance issues emerge, anti-PCSK9 antibodies could revolutionize lipid-lowering treatment in difficult-to-treat hypercholesterolemic patients, allowing patients to attain LDL-C levels not achievable with conventional treatment.

Financial & competing interests disclosure

G Langslet has received moderate advisory board fees from Amgen, Sanofi and Janssen Pharmaceuticals. M Emery is an employee of Amgen and thus receives salary, stock and benefits from Amgen. S Wasserman is an employee of Amgen and thus receives salary, stock and benefits from Amgen. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed. Meera Kodukulla was involved in the editorial and formatting of this document. Mary Elliott, Thomas Liu and Jingyuan Yang were involved in the statistical quality control of this document.