OUP user menu

★ Editor's choice ★

Efficacy and safety of evolocumab (AMG 145), a fully human monoclonal antibody to PCSK9, in hyperlipidaemic patients on various background lipid therapies: pooled analysis of 1359 patients in four phase 2 trials

Evan A. Stein, Robert P. Giugliano, Michael J. Koren, Frederick J. Raal, Eli M. Roth, Robert Weiss, David Sullivan, Scott M. Wasserman, Ransi Somaratne, Jae B. Kim, Jingyuan Yang, Thomas Liu, Moetaz Albizem, Rob Scott, Marc S. Sabatine on behalf of for the PROFICIO Investigators
DOI: http://dx.doi.org/10.1093/eurheartj/ehu085 2249-2259 First published online: 4 March 2014


Aims Prior trials with monoclonal antibodies to proprotein convertase subtilizin/kexin type 9 (PCSK9) reported robust low density lipoprotein cholesterol (LDL-C) reductions. However, the ability to detect potentially beneficial changes in other lipoproteins such as lipoprotein (a), triglycerides, high-density lipoprotein cholesterol (HDL-C), and apolipoprotein (Apo) A1, and adverse events (AEs) was limited by sample sizes of individual trials. We report a pooled analysis from four phase 2 studies of evolocumab (AMG 145), a monoclonal antibody to PCSK9.

Methods and results The trials randomized 1359 patients to various doses of subcutaneous evolocumab every 2 weeks (Q2W) or 4 weeks (Q4W), placebo, or ezetimibe for 12 weeks; 1252 patients contributed to efficacy and 1314, to safety analyses. Mean percentage (95% CI) reductions in LDL-C vs. placebo ranged from 40.2% (44.6%, 35.8%) to 59.3% (63.7%, 54.8%) among the evolocumab groups (all P < 0.001). Statistically significant reductions in apolipoprotein B (Apo B), non-high-density lipoprotein cholesterol (non-HDL-C), triglycerides and lipoprotein (a) [Lp(a)], and increases in HDL-C were also observed. Adverse events (AEs) and serious AEs with evolocumab were reported in 56.8 and 2.0% of patients, compared with 49.2% and 1.2% with placebo. Adjudicated cardiac and cerebrovascular events were reported in 0.3 and 0% in the placebo and 0.9 and 0.3% in the evolocumab arms, respectively.

Conclusion In addition to LDL-C reduction, evolocumab, dosed either Q2W or Q4W, demonstrated significant and favourable changes in other atherogenic and anti-atherogenic lipoproteins, and was well tolerated over the 12-week treatment period.

  • Proprotein convertase subtilizin/kexin type 9
  • Low-density lipoprotein cholesterol
  • Randomized controlled trials


One of the cornerstones for reducing cardiovascular disease (CVD) morbidity and mortality is reduction of low-density lipoprotein cholesterol (LDL-C).1,2 Although the great majority of the evidence has been obtained from statin-based randomized outcome trials,3,4 it is also supported by similar reductions in CVD with other agents such as bile acid sequestrants and immediate release nicotinic acid.5,6 However, there is still a large unmet need for additional LDL-C reducing agents because existing therapies may not be well tolerated and not all treated patients can achieve LDL-C goals.7,8

Based on biology and human genetics, proprotein convertase subtilizin/kexin type 9 (PCSK9) provides a new and compelling target for LDL-C lowering therapy.9 Over the last year, early clinical trials with monoclonal antibodies have confirmed this approach with rapid and dramatic LDL-C reductions.1017 Potentially beneficial trends in other lipids and lipoproteins, such as triglycerides, lipoprotein (a) [Lp(a)], and high-density lipoprotein cholesterol (HDL-C), were observed, but the precision of those estimates and the capacity to examine for consistency across clinically important subgroups, such as gender, age, and background lipid therapy, were limited by the sample sizes in each trial.

As part of the Program to Reduce LDL-C and Cardiovascular Outcomes Following Inhibition of PCSK9 in Different Populations (PROFICIO), over 1300 patients from 4 randomized, placebo-controlled phase 2 trials1114 were pooled into a single data set to better assess efficacy, safety, and tolerability of evolocumab (AMG 145).


Data sources

The patient populations, background lipid therapy, and treatment arms of the four phase 2 trials included in this analysis are summarized in Table 1.1114 Patients reviewed and signed an informed consent statement approved by the institutional review board governing each clinical centre. All studies were of 12-week duration, double-blind, placebo-controlled, and randomized patients to subcutaneous (SC) evolocumab at various doses and dose frequencies: 70, 105, and 140 mg every 2 weeks (Q2W); 280, 350, and 420 mg SC every 4 weeks (Q4W); or placebo (Q2W or Q4W, respectively). In two trials, patients were allocated to ezetimibe, either concomitantly with evolocumab or placebo (n = 62) or alone (n = 45).11,14 All lipid and apolipoprotein (Apo) measurements were performed in the same central laboratory, which maintained Centres for Disease Control and Prevention part III of the Lipid Standardization Program.18

View this table:
Table 1

Characteristics of the four Evolocumab phase 2 studies

StudyaNumber of patientsaRandomizationPatient profile and background RxTreatments and doses
MENDEL14 (monotherapy)b4069 arms, equal allocationNo background anti-lipid therapy
LDL-C ≥2.6 and <4.9 mmol/L (≥100 and <190 mg/dL)
Evolocumab 70 mg, 105 mg, or 140 mg or placebo Q2W
Evolocumab 280 mg, 350 mg, or 420 mg or placebo Q4W
Daily ezetimibe 10 mg
LAPLACE-TIMI 5715 (combination therapy)b6298 arms, equal allocationStatin ± ezetimibe
LDL-C ≥2.2 mmol/L (≥85 mg/dL)
Evolocumab 70 mg, 105 mg, 140 mg, or placebo Q2W
Evolocumab 280 mg, 350 mg, 420 mg, or placebo Q4W
RUTHERFORD16 (heterozygous FH)b1673 arms, equal allocationStatin ± ezetimibe
Heterozygous FH with LDL-C ≥2.6 mmol/L (≥100 mg/dL)
Evolocumab 350 mg, 420 mg, or placebo Q4W
GAUSS17 (statin intolerant)b1575 arms, equal allocationNo/low-dose statin
LDL-C ≥2.6 mmol/L (≥100 mg/dL)
Evolocumab 280 mg, 350 mg, or 420 mg Q4W
Ezetimibe 10 mg + SC placebo Q4W
Ezetimibe 10 mg + evolocumab 420 mg Q4W
  • GAUSS, Goal Achievement After Utilizing an Anti-PCSK9 Antibody in Statin-Intolerant Subjects; LDL-C, low-density lipoprotein cholesterol; LAPLACE-TIMI 57, LDL-C Assessment With PCSK9 Monoclonal Antibody Inhibition Combined With Statin Therapy; MENDEL, Monoclonal Antibody Against PCSK9 to Reduce Elevated LDL-C in Patients Currently Not Receiving Drug Therapy for Easing Lipid Levels; Q2W, every 2 weeks; Q4W, every 4 weeks; PBO, placebo; RUTHERFORD, Reduction of LDL-C With PCSK9 Inhibition in Heterozygous Familial Hypercholesterolaemia Disorder; SC, subcutaneous.

  • aThe number of patients who were randomized and received at least one dose of investigational product.

  • bThe primary endpoint for each trial was percent reduction in LDL-C from baseline at week 12.

Data extraction

A total of 1359 patients were randomized (See Supplementary material online, Figure S1). The pooled efficacy analysis (n = 1252), included data from all patients who received at least one dose of evolocumab (n = 951) or placebo (n = 301) without concomitant allocation to ezetimibe. For the pooled safety analysis (n = 1314), patients who received evolocumab with or without ezetimibe (n = 981) and patients who received placebo with or without ezetimibe (n = 333) were included; patients randomized to ezetimibe alone (n = 45) were excluded.

Endpoint definitions

The primary endpoint in each trial was the percentage change from baseline in LDL-C measured by ultracentrifugation (UC LDL-C) at week 12. Secondary endpoints, all at week 12, included percentage change from baseline in calculated LDL-C by Friedewald formula,19 total cholesterol, HDL-C, non-HDL-C, very-low-density lipoprotein cholesterol (VLDL-C), Apo B, Apo A1, Lp(a), triglycerides, and the ratios of total cholesterol/HDL-C and Apo B/Apo A1. Exploratory analysis included LDL-C, Apo B, HDL-C, and Apo A1 response by gender, age (< and ≥65 years), statin or no-statin background therapy and above or below the mean baseline LDL-C at entry of 3.64 mmol/L (140.6 mg/dL). The incidence of adjudicated events of death (all-cause and cardiovascular); cardiac ischaemic events (myocardial infarction, hospitalization for unstable angina, and coronary revascularization); hospitalization for heart failure; and cerebrovascular events were assessed throughout the trials. All cardiovascular-related events were pre-specified for blinded adjudication by an academically based clinical events committee. Key safety endpoints included the incidence of treatment-emergent adverse events (AEs), treatment-related AEs, safety laboratory values, and the incidence of anti-evolocumab antibodies (binding and if positive, neutralizing).

Statistical analysis

Patients were analysed in the treatment group to which they were randomized. All efficacy endpoints of lipid parameters were analysed using the analysis of covariance model (ANCOVA) within each dosing regimen (Q2W or Q4W) to assess the efficacy of evolocumab dose groups compared with placebo, with last observation carried forward imputation for missing data, a significance level of 0.05, and no adjustment for multiple comparisons. The ANCOVA model includes treatment group and a common stratification factor used in each study (screening LDL values <130 mg/dL or not). Safety analyses were conducted using descriptive statistics. Safety data were reported as observed. All analyses were performed with SAS/STAT®, version 9.2 (SAS Institute, Cary, NC, USA).


Patient demographics, baseline statin and ezetimibe therapy, and lipid levels are shown in Table 2. The mean ± standard deviation (SD) age was 56.4 ± 11.7 years, 56.2% were women, and 85.7% were white. At baseline, a total of 60.3% of patients were taking a statin and 11.8% were taking ezetimibe; the baseline mean ± SD baseline UC LDL-C was 3.64 ± 1.01 mmol/L (141 ± 39 mg/dL).

View this table:
Table 2

Patient demographics and lipid parameters at baseline

Q2W (n = 123)Q4W (n = 178)Q4W + EZE QD (n = 32)70 mg Q2W (n = 124)105 mg Q2W (n = 125)140 mg Q2W (n = 123)280 mg Q4W (n = 156)350 mg Q4W (n = 210)420 mg Q4W (n = 213)420 mg Q4W + EZE QD (n = 30)N = 1359
Age, years, mean (SD)57.4 (9.9)54.4 (12.0)62.4 (6.6)56.3 (11.8)54.9 (11.6)58.9 (11.7)57.5 (10.7)55.9 (13.0)56.1 (12.1)62.0 (7.2)56.4 (11.7)
Age <65 years, n (%)95 (77.2)135 (75.8)18 (56.3)90 (72.6)96 (76.8)81 (65.9)111 (71.2)150 (71.4)151 (70.9)17 (56.7)987 (72.6)
Women, n (%)67 (54.5)103 (57.9)18 (56.3)74 (59.7)58 (46.4)81 (65.9)86 (55.1)117 (55.7)111 (52.1)23 (76.7)764 (56.2)
Race, n (%)
 White107 (87.0)158 (88.8)28 (87.5)105 (84.7)104 (83.2)106 (86.2)127 (81.4)185 (88.1)185 (86.9)24 (80.0)1164 (85.7)
 Black14 (11.4)7 (3.9)2 (6.3)13 (10.5)13 (10.4)13 (10.6)18 (11.5)15 (7.1)21 (9.9)1 (3.3)126 (9.3)
 Other2 (1.6)13 (7.3)2 (6.3)6 (4.8)6 (4.8)4 (3.2)11 (7.1)10 (4.8)7 (3.3)5 (16.7)63 (4.6)
Statin use, n (%)78 (63.4)133 (74.7)6 (18.8)78 (62.9)78 (62.4)78 (63.4)84 (53.8)139 (66.2)141 (66.2)4 (13.3)819 (60.3)
Ezetimibe use, n (%)5 (4.1)41 (23.0)0 (0.0)6 (4.8)8 (6.4)7 (5.7)7 (4.5)44 (21.0)42 (19.7)0 (0.0)160 (11.8)
Lipid parameters, mean (SD)
 UC LDL-C,b mmol/L3.39 (0.72)3.64 (0.96)4.73 (0.94)3.34 (0.66)3.43 (0.77)3.31 (0.65)3.68 (1.10)3.74 (1.09)3.74 (1.15)5.03 (1.55)3.64 (1.01)
 Calc LDL-C,c mmol/L3.37 (0.75)3.60 (0.99)4.68 (0.92)3.31 (0.70)3.37 (0.79)3.25 (0.66)3.62 (1.10)3.69 (1.13)3.68 (1.17)5.01 (1.54)3.59 (1.03)
 Total-C, mmol/L5.48 (0.93)5.64 (1.10)7.10 (1.01)5.39 (0.82)5.47 (0.91)5.35 (0.76)5.77 (1.28)5.75 (1.23)5.76 (1.29)7.23 (1.69)5.68 (1.16)
 HDL-C, mmol/L1.42 (0.46)1.36 (0.42)1.57 (0.49)1.41 (0.44)1.39 (0.46)1.41 (0.39)1.42 (0.46)1.34 (0.40)1.34 (0.43)1.55 (0.50)1.39 (0.43)
 Non-HDL-C, mmol/L4.05 (0.85)4.27 (1.10)5.53 (1.04)3.98 (0.78)4.08 (0.88)3.93 (0.69)4.35 (1.23)4.41 (1.21)4.42 (1.33)5.68 (1.57)4.3 (1.13)
 Total-C/HDL-C ratio4.14 (1.15)4.51 (1.65)4.91 (1.55)4.12 (1.20)4.29 (1.41)4.02 (1.07)4.41 (1.46)4.62 (1.64)4.71 (1.91)5.09 (2.42)4.44 (1.55)
 VLDL-C, mmol/L0.65 (0.40)0.64 (0.40)0.8 (0.37)0.64 (0.33)0.65 (0.34)0.63 (0.27)0.67 (0.36)0.67 (0.38)0.68 (0.40)0.66 (0.27)0.66 (0.37)
 Apo B, g/L1.04 (0.18)1.11 (0.25)1.38 (0.22)1.03 (0.17)1.05 (0.20)1.02 (0.17)1.12 (0.27)1.14 (0.28)1.15 (0.29)1.39 (0.33)1.11 (0.25)
 Apo A1, g/L1.57 (0.30)1.51 (0.30)1.71 (0.32)1.56 (0.29)1.54 (0.30)1.57 (0.27)1.58 (0.31)1.52 (0.27)1.51 (0.29)1.63 (0.29)1.55 (0.29)
 Apo B/Apo A1 ratio0.69 (0.17)0.77 (0.27)0.84 (0.23)0.69 (0.18)0.71 (0.20)0.67 (0.16)0.74 (0.22)0.78 (0.26)0.79 (0.28)0.88 (0.33)0.75 (0.24)
 Triglycerides, mmol/L1.51 (0.76)1.50 (0.82)1.85 (0.74)1.46 (0.62)1.55 (0.72)1.50 (0.64)1.58 (0.72)1.60 (0.81)1.61 (0.82)1.48 (0.63)1.55 (0.75)
 Lipoprotein(a), median (IQ range), nmol/L47.0 (16.0, 176.0)42.0 (14.0, 132.0)58.5 (5.5, 140.5)46.0 (14.0, 136.0)40.0 (12.0, 151.0)43.5 (14.0, 141.0)37.0 (12.0, 130.0)36.0 (10.0, 127.0)41.0 (14.5, 152.5)39.0 (14.0, 187.0)40.0 (13.0, 144.0)
 Free PCSK9, ng/mL413.7 (125.2)470.5 (165.6)389.9 (90.9)385.6 (113.0)398 (104.5)411.7 (117.8)402.9 (135.0)459.5 (164.0)468.3 (158.9)379.4 (110.9)427.5 (143.1)
  • Apo A1, apolipoprotein A1; Apo B, apolipoprotein B; calc, calculated; EZE, ezetimibe; HDL-C, high-density lipoprotein cholesterol; IQ, interquartile; LDL-C, low-density lipoprotein cholesterol; PCSK9, proprotein convertase subtilizin/kexin type 9;IQ, interquartile; Q2W, every 2 weeks; Q4W, every 4 weeks; SC, subcutaneous; Total-C, total cholesterol; UC LDL-C, LDL-C measured by preparative ultracentrifugation; VLDL-C, very low-density lipoprotein cholesterol.

  • aIncludes all randomized and dosed patients from the four clinical trials.

  • bLDL-C measured by preparative ultracentrifugation.

  • cLDL-C calculated by the Friedewald formula.

Efficacy assessments

The percentage changes at week 12 in lipids and apolipoproteins from baseline, with differences relative to placebo, are shown in Table 3. Mean (95% CI) UC LDL-C reductions (all P < 0.001) with evolocumab showed a dose-related response within both the Q2W and Q4W dosing groups which ranged from 40.2% (44.6%, 35.8%) to 59.3% (63.7%, 54.8%), with the greatest decrease seen with the 140 mg Q2W dose. The LDL-C reductions by Friedewald calculation were larger than, but parallel to, those measured by UC for all groups (Figure 1A and B) and demonstrated a greater duration of effect with increasing doses when administered every 4 weeks (Figure 1B). The LDL-C reductions for each dose group by various subgroups are shown in Figure 2 and Supplementary material online, Table S1. In the Q2W dose groups, baseline statin use was associated with statistically significantly greater reductions in LDL-C (P < 0.001) (Figure 2), but this effect was not seen in the Q4W dose groups. Other nominally significant subgroup differences were observed; caution is warranted in the interpretation of these results, which may be due to imbalances in statin use.

View this table:
Table 3

Summary of lipid, lipoprotein and apolipoprotein percentage changes at week 12, treatment difference vs. placeboa

70 mg Q2W (n = 124)105 mg Q2W (n = 125)140 mg Q2W (n = 123)280 mg Q4W (n = 156)350 mg Q4W (n = 210)420 mg Q4W (n = 213)
UC LDL-C−40.20 (−44.64, −35.76)−52.86 (−57.29, −48.42)−59.26 (−63.71, −54.80)−42.55 (−46.53, −38.58)−47.00 (−50.70, −43.31)−52.66 (−56.35, −48.96)
Other lipid parameters
 Calculated LDL-C−42.36 (−46.90, −37.82)−55.32 (−59.85, −50.79)−62.59 (−67.14, −58.03)−44.34 (−48.40, −40.28)−49.54 (−53.31, −45.76)−56.02 (−59.79, −52.25)
 Total-C−25.74 (−28.97, −22.51)−34.43 (−37.65, −31.21)−38.23 (−41.47, −34.99)−28.42 (−31.24, −25.59)−31.43 (−34.05, −28.80)−36.14 (−38.76, −33.51)
 Non-HDL-C−37.24 (−41.31, −33.18)−48.56 (−52.61, −44.50)−55.51 (−59.59, −51.43)−38.90 (−42.51, −35.28)−43.30 (−46.65, −39.94)−49.18 (−52.53, −45.82)
 Total-C/HDL-C ratio−30.50 (−34.08, −26.92)−39.15 (−42.73, −35.58)−44.02 (−47.62, −40.43)−30.45 (−33.72, −27.17)−34.98 (−38.02, −31.94)−39.26 (−42.29, −36.22)
 VLDL-C−24.94 (−37.12, −12.76)−27.84 (−40.09, −15.59)−41.54 (−53.69, −29.39)−21.68 (−31.59, −11.78)−23.27 (−32.49, −14.06)−28.92 (−38.11, −19.72)
 Apo B/Apo A1 ratio−34.77 (−38.69, −30.86)−44.68 (−48.59, −40.77)−52.23 (−56.16, −48.30)−34.35 (−37.96, −30.75)−39.19 (−42.54, −35.84)−44.33 (−47.68, −40.98)
  • All P < 0.001 for evolocumab vs. placebo in the same dose frequency group unless indicated otherwise.

  • Note that one trial (GAUSS17) did not have a true placebo-only group.

  • Apo A1, apolipoprotein A1; Apo B, apolipoprotein B; HDL-C, high-density lipoprotein cholesterol; Lp(a), lipoprotein(a); LDL-C, low-density lipoprotein cholesterol; Q2W, every 2 weeks; Q4W, every 4 weeks; Total-C, total cholesterol; VLDL-C, very low-density lipoprotein cholesterol; UC LDL-C, LDL-C measured by preparative ultracentrifugation.

  • aValues represent least-squares means (95% CI) for treatment differences vs. placebo and were derived from an analysis-of-covariance (ANCOVA) model controlling for treatment and a common stratification factor (screening LDL values <130 mg/dL or not).

Figure 1

(A) Percentage changes from baseline in levels of low density lipoprotein cholesterol (LDL-C) (calculated) for patients treated every 2 weeks (Q2W). (B) Percentage changes in levels of LDL-C (calculated) for patients treated every 4 weeks (Q4W). Error bars represent standard error (SE). Note that visits were not required at all time points in all the studies.

Figure 2

Low density lipoprotein cholesterol (LDL-C) percentage changes from baseline relative to the corresponding placebo group in 70, 105, and 140 mg Q2W and 280, 350, and 420 mg Q4W evolocumab treated groups by gender, age (< /≥65 years), statin/no statin, mean baseline LDL-C (ultracentrifugation) (</≥3.64 mmol/L), and baseline triglyceride </≥1.7 mmol/L. Least-squares mean. Error bars represent 95% confidence intervals. *P-values reflect significance of treatment by subgroup interaction within each dose group for all dose groups. UC, ultracentrifugation.

Reductions in calculated LDL-C were approximately 5 percentage points greater than reductions in UC LDL-C at week 12 with all doses of evolocumab (Table 3 and Figure 1A and B). Consistent with the pharmacokinetics of evolocumab, maximal reductions in calculated LDL-C of ∼65–70% were observed 2 weeks following 420 mg Q4W dose administration and the LDL-C reduction remained close to 60% at week 4.

Reductions in total cholesterol, VLDL-C, non-HDL-C, and Apo B paralleled the decreases seen in LDL-C and were all statistically significant (P < 0.001) (Table 3 and Figure 3). Changes in Apo B for each dose group by gender, age, background statin therapy, or LDL-C above or below the mean at baseline are shown in Supplementary material online, Table S2. There was significantly greater (P < 0.001) reduction in Apo B seen with Q2W dosing in those patients on statin therapy but no difference with Q4W dosing (Supplementary material online, Table S2), similar to what was observed for LDL-C.

Figure 3

Percentage changes from baseline in apolipoprotein B, HDL cholesterol, lipoprotein (a), apolipoprotein A1, and triglycerides by dose and dosing frequency of evolocumab vs. placebo at week 12. Least-squares mean. Error bars represent standard error. Values in parentheses represent interquartile ranges. ApoA1, apolipoprotein A1; ApoB, apolipoprotein B; HDL-C, high-density lipoprotein cholesterol. *P < 0.001; P < 0.01; P < 0.05.

Significant reductions (P < 0.001) in triglycerides were seen with all doses and regimens of evolocumab (Figure 3), with mean (95% CI) percentage reductions from baseline relative to placebo ranging from 12.2% (19.1%, 5.2%) to 25.9% (35.0%, 16.8%). Lp(a) levels were significantly (P < 0.001) reduced at week 12 with all evolocumab doses and dosing regimens, with mean (95% CI) percentage reductions from baseline relative to placebo ranging from 15.5% (22.0%, 9.0%) to 31.3% (37.8%, 24.7%).

Treatment differences relative to placebo in HDL-C and its major associated apolipoprotein, Apo A1, are shown in Figure 3. HDL-C was significantly increased with all evolocumab doses and dosing regimens at week 12 (P < 0.05 to P < 0.001), with the largest mean (95% CI) percentage increases from baseline relative to placebo of 8.9% (5.1%, 12.7%) and 5.9% (3.1%, 8.8%) seen with evolocumab 140 mg Q2W and 420 mg Q4W, respectively (P < 0.001). The parallel, but slightly smaller, mean (95% CI) percentage increases from baseline relative to placebo seen in Apo A1 were statistically significant for the evolocumab 140 mg Q2W, 350 mg Q4W, and 420 mg Q4W doses (P-values ranging from <0.001 to <0.05). Percentage changes in HDL-C and Apo A1 for each dose group by gender, age, background statin therapy, or above or below the mean baseline LDL-C are shown in Supplementary material online, Tables S3 and S4. High-density lipoprotein cholesterol showed no difference in response in any of the subgroups, but Apo A1 showed a consistently greater (P ≤ 0.05) increase in patients not on statins at baseline with both 2- and 4-week dosing with evolocumab.

Safety assessments

Adverse events, clinical, and laboratory safety analyses are shown in Table 4. Overall, AEs were reported in 56.8 and 49.2% of patients in the combined evolocumab and combined placebo groups, respectively. Across evolocumab and placebo groups, AEs ranged from 42.3 to 66.7% (Table 4) with nasopharyngitis, upper respiratory infection, headache, diarrhoea, myalgia, and back pain the most commonly reported. Serious AEs (SAEs) were reported in 2.0 and 1.2% of patients in the evolocumab and placebo groups, respectively; none of these was considered to be treatment related by the investigators. Treatment-related AEs were reported in 11.5 and 9.6% of patients in the evolocumab and placebo groups, respectively. Injection-site reactions occurred in 4.1 and 3.3% of patients in the evolocumab and placebo groups, respectively, and muscle-related AEs occurred in 6.0 and 3.9% of patients. One patient receiving evolocumab at 350 mg Q4W was adjudicated to have myopathy.

View this table:
Table 4

Summary of adverse events

 PlaceboEvolocumabAll placebo (n = 333)All evolocumab (n = 981)
Q2W (n = 123)Q4W (n = 178)Q4W + EZE QD (n = 32)70 mg Q2W (n = 124)105 mg Q2W (n = 125)140 mg Q2W (n = 123)280 mg Q4W (n = 156)350 mg Q4W (n = 210)420 mg Q4W (n = 213)420 mg Q4W + EZE QD (n = 30)
AEsa52 (42.3)93 (52.2)19 (59.4)65 (52.4)74 (59.2)69 (56.1)89 (57.1)118 (56.2)122 (57.3)20 (66.7)164 (49.2)557 (56.8)
 Nasopharyngitis9 (7.3)11 (6.2)5 (15.6)11 (8.9)10 (8.0)8 (6.5)11 (7.1)20 (9.5)18 (8.5)3 (10.0)25 (7.5)81 (8.3)
 Upper respiratory tract infection7 (5.7)4 (2.2)0 (0.0)5 (4.0)5 (4.0)6 (4.9)3 (1.9)7 (3.3)11 (5.2)3 (10.0)11 (3.3)40 (4.1)
 Myalgia0 (0.0)3 (1.7)1 (3.1)4 (3.2)2 (1.6)3 (2.4)7 (4.5)7 (3.3)3 (1.4)6 (20.0)4 (1.2)32 (3.3)
 Headache5 (4.1)6 (3.4)0 (0.0)4 (3.2)3 (2.4)6 (4.9)1 (0.6)6 (2.9)6 (2.8)6 (20.0)11 (3.3)32 (3.3)
 Back pain1 (0.8)6 (3.4)0 (0.0)1 (0.8)3 (2.4)4 (3.3)5 (3.2)6 (2.9)10 (4.7)2 (6.7)7 (2.3)29 (3.0)
 Diarrhoea2 (1.6)8 (4.5)1 (3.1)3 (2.4)4 (3.2)4 (3.3)2 (1.3)6 (2.9)8 (3.8)1 (3.3)11 (3.3)28 (2.9)
 Nausea1 (0.8)4 (2.2)1 (3.1)0 (0.0)1 (0.8)6 (4.9)7 (4.5)5 (2.4)7 (3.3)0 (0.0)6 (1.8)26 (2.7)
 Fatigue4 (3.3)1 (0.6)2 (6.3)0 (0.0)2 (1.6)4 (3.3)4 (2.6)4 (1.9)8 (3.8)0 (0.0)7 (2.1)22 (2.2)
 Arthralgia1 (0.8)5 (2.8)1 (3.1)5 (4.0)6 (4.8)2 (1.6)0 (0.0)3 (1.4)3 (1.4)1 (3.3)7 (2.1)20 (2.0)
Treatment-related AEs10 (8.1)15 (8.4)7 (21.9)8 (6.5)16 (12.8)13 (10.6)19 (12.2)27 (12.9)25 (11.7)5 (16.7)32 (9.6)113 (11.5)
AEs leading to discontinuation1 (0.8)2 (1.1)2 (6.3)0 (0.0)0 (0.0)2 (1.6)0 (0.0)2 (1.0)2 (0.9)1 (3.3)5 (1.5)7 (0.7)
 SAEs4 (3.3)0 (0.0)0 (0.0)0 (0.0)2 (1.6)5 (4.1)4 (2.6)4 (1.9)5 (2.3)0 (0.0)4 (1.2)20 (2.0)
Treatment-related SAEs0 (0.0)0 (0.0)0 (0.0)0 (0.0)0 (0.0)0 (0.0)0 (0.0)0 (0.0)0 (0.0)0 (0.0)0 (0.0)0 (0.0)
AEs and laboratory parameters of interest
 Injection-site reaction events3 (2.4)7 (3.9)1 (3.1)2 (1.6)7 (5.6)2 (1.6)9 (5.8)13 (6.2)5 (2.3)2 (6.7)11 (3.3)40 (4.1)
 Muscle-related AEs2 (1.6)7 (3.9)4 (12.5)7 (5.6)5 (4.0)4 (3.3)13 (8.3)11 (5.2)13 (6.1)6 (20.0)13 (3.9)59 (6.0)
 Creatine kinase > 5× ULNb1 (0.8)1 (0.6)1 (3.1)3 (2.4)2 (1.6)1 (0.8)0 (0.0)3 (1.4)5 (2.3)0 (0.0)3 (0.9)14 (1.4)
 ALT or AST > 3 × ULN1 (0.8)0 (0.0)1 (3.1)1 (0.8)0 (0.0)0 (0.0)0 (0.0)2 (1.0)1 (0.5)0 (0.0)2 (0.6)4 (0.4)
Binding antibodies0 (0.0)1 (0.6)0 (0.0)0 (0.0)1 (0.8)0 (0.0)0 (0.0)0 (0.0)0 (0.0)0 (0.0)1 (0.3)1 (0.1)
Neutralizing antibodies0 (0.0)0 (0.0)0 (0.0)0 (0.0)0 (0.0)0 (0.0)0 (0.0)0 (0.0)0 (0.0)0 (0.0)0 (0.0)0 (0.0)
  • AEs, adverse events; ALT, alanine aminotransferase; AST, aspartate aminotransferase; EZE, ezetimibe; Q2W, every 2 weeks; Q4W, every 4 weeks; QD, once daily; SAEs, serious adverse events; ULN, upper limit of normal.

  • aReported in at least 2% of evolocumab recipients, all values represent no. (%).

  • bFour patients in the evolocumab treatment group had creatine kinase >10× ULN, all of which were resolved at follow-up blood test.

A total of 13 patients experienced cardiac (n = 10) or cerebrovascular (n = 3) events as shown in Supplementary material online, Table S5. The majority of cardiac events were revascularization procedures (8 of 10), subsequent in 5 of the patients to an acute cardiac ischaemic event. Of patients experiencing a cardiac event, 0.3% (1/333) was in the placebo arm and 0.9% (9/981) in the pooled evolocumab groups. There were no cerebrovascular events in the placebo group and three (0.3%) events in the pooled evolocumab groups (Supplementary material online, Table S5).

The rate of aspartate aminotransferase and/or alanine aminotransferase elevations more than three times the upper limit of normal (ULN) at any visit was similar in the evolocumab and placebo groups (0.4 and 0.6%, respectively). Elevated creatine kinase (CK) more than five times ULN at any visit was reported in 1.4 and 0.9% of patients in the evolocumab and placebo groups, respectively. All CK elevations were asymptomatic, single occurrences that resolved spontaneously without discontinuation of investigational product. Of the five patients with CK >10 times ULN who received evolocumab, the elevations were associated with strenuous physical activity; for 4 patients, CK levels were below ULN on follow-up blood testing at the next visit and the 1 patient was adjudicated as having myopathy (as noted above). Anti-evolocumab binding antibodies were observed in one patient on evolocumab and one patient in the placebo group; no neutralizing antibodies were detected.


The PROFICIO analysis represents the largest assessment of a PCSK9 monoclonal antibody phase 2 programme to date, with nearly 1000 patients treated with evolocumab, and allows robust comparisons with a placebo group of over 300 patients. Thus, the ability to determine efficacy in subgroups of evolocumab considerably exceeds that of any prior individual trial. The treatment duration of 12 weeks is almost twice as long as the randomized portion of the phase 2 trials of statins20,21 or other novel lipid-altering agents.22 The patient population included in PROFICIO is also more representative and inclusive of patients in whom the drug is likely to be used in clinical practice than prior phase 2 trials of new lipid agents: 25–30% of patients in PROFICIO were older than 65 years, one-quarter to one-third were at high cardiovascular risk, and two-thirds were on statin background therapy. Despite the background lipid-lowering therapy, the mean ± SD baseline LDL-C level was 3.64 ± 1.01 mmol/L.

The large reductions in LDL-C, total cholesterol, non-HDL-C, and Apo B seen with all doses and dose regimens of evolocumab in this pooled analysis with larger cohorts confirmed the findings in the individual trials. However, it also allowed for greater certainty in the assessment of dose-related reductions and confirmed maximal reductions are achieved with the 140 mg Q2W dosing. In addition, while the maximal LDL-C reductions seen after 2 weeks with 350 and 420 mg dosing were the same, the 420 mg dose maintained a greater LDL-C reduction for a longer period as measured 4 weeks post-dose. The large number of patients per treatment group also allowed for statistically significant changes to be seen in triglycerides, Lp(a), HDL-C, and Apo A1. Despite average baseline triglyceride levels in PROFICIO of slightly <1.70 mmol/L (150 mg/dL), statistically significant reductions of 26 and 16% were seen with the highest doses used in the Q2W and Q4W dosing regimens, respectively. As is well documented with statins, which also upregulate LDL receptor activity, triglyceride reductions are dependent to a large extent on baseline triglycerides.23 Future trials that include patients with higher baseline triglycerides may help better determine the therapeutic potential for evolocumab in more severely hypertriglyceridaemic patients and if a dose-related effect exists as for statins. The large cohorts provide solid and robust confirmation of the significant reduction in Lp(a) with evolocumab (P < 0.001 for each dose vs. placebo) seen in the individual phase 2 trials with evolocumab. In addition, these results also confirm trends observed in smaller studies with another PCSK9 monoclonal antibody, alirocumab.16,17,24

The magnitude of effect of evolocumab on the HDL parameters, HDL-C, and Apo A1 was somewhat unexpected because of the large proportion of patients already on statin therapy, which also raises HDL-C, and with relatively high baseline HDL-C levels [range, 1.34−1.42 mmol/L (51.7–54.7 mg/dL) for the pooled efficacy analysis population]. Thus, the further statistically significant HDL-C increase of ∼9% with 140 mg Q2W and ∼6% with 420 mg Q4W represents an additional potentially favourable effect of therapy. Associated with the HDL-C increases was a uniform increase in Apo A1 with all doses of evolocumab, which reached ∼4% (P < 0.05 to P < 0.001) with the highest doses in each dosing regimen. Thus, in terms of the overall lipid profile, evolocumab appears to provide a superior effect on LDL-C, non-HDL-C, Apo B, Lp(a), and HDL-C than even the highest doses of the most effective statins.25 Whether these changes in lipid biomarkers translate into reductions in CVD events and outcomes will be assessed over the next few years in the 22 500 patient evolocumab trial, Further Cardiovascular Outcomes Research With PCSK9 Inhibition in Subjects With Elevated Risk [FOURIER (NCT01764633)].26

In terms of safety assessment, PROFICIO provides 225 patient-years of exposure to evolocumab and comparison with 70 patient-years of treatment with placebo in these double-blind studies. The overall AE rate was slightly higher with evolocumab (56.8%) than placebo (49.2%), but when individual AEs were compared, no differences or trends in any organ system were seen. There was also a low rate of AEs leading to discontinuation (0.7% with evolocumab and 1.5% with placebo) and no dose- or dose frequency-related trends noted with evolocumab. Although SAEs during the trial were also slightly greater in those treated with evolocumab than placebo (2.0 vs. 1.2%, respectively), there were no SAEs determined to be treatment-related by investigators in either group. The tolerability of SC administration was excellent irrespective of dosing frequency, with injection site reactions reported in 4.1% of evolocumab-treated patients and in 3.3% of placebo-treated patients. This tolerability contrasts significantly with the only approved SC-administered lipid-lowering agent, mipomersen, an Apo B antisense drug given once weekly, with which a high rate (92%) of injection-site reactions was observed.

The hepatic function and muscle-related enzyme measurements are reassuring as neither showed any signals of toxicity in this large, but short duration, study analysis. The inclusion in PROFICIO of a trial in which the entry criteria required documented evidence of muscle-related AEs on prior statin therapy and patients who were randomized to evolocumab or ezetimibe (i.e. no placebo-only arm) provided a large number of patients ‘sensitized’ to this side effect. Thus, the relatively low reported rate of muscle-related AEs with evolocumab, which was not significantly different from placebo, provides further evidence of the importance of the drug for statin-averse patients. A total of 0.3% of placebo-treated patients experienced a cardiac event compared with 0.9% treated with evolocumab. However, the phase 2 programme was not designed to evaluate cardiovascular outcomes and the aforementioned proportions represent a total of only 13 patients experiencing a cardiac or cerebrovascular event. Nonetheless, the imbalance indicates the need for continued close, independent monitoring for cardiac and cerebrovascular events during the phase 3 programme. Reassuringly, genetic association studies have demonstrated lower rates of CVD events in patients carrying loss-of-function PCSK9 variants.27 The definitive evaluation of pharmacologic PCSK9 inhibition on cardiovascular events will have to wait for the results from the ongoing FOURIER trial.26


PROFICIO examined two different patient populations across the four studies, patients with and without background statin therapy. Despite the large patient population, it has limitations specifically for assessing safety. The trials were all of 12 weeks' duration and thus were unable to determine the safety of a drug that will likely be administered for life after patients commence therapy; results from an ongoing open-label extension study will provide additional safety data. In addition, short-term trials with highly motivated and monitored patients in which the drug is administered in a research clinic may not be an optimal setting to fully evaluate the acceptance of, and long-term compliance with, a lipid-lowering agent that requires SC administration. These issues will be explored further in the phase 3 programme.


The PROFICIO analysis confirms the large and rapid reductions achievable in Apo B and its related lipoproteins, LDL-C, VLDL-C, and non-HDL-C seen in individual trials with evolocumab. The pooled analysis further provides new and additional data on the prolonged effect of the 420 mg Q4W dose on LDL-C and the robust decreases in triglycerides and Lp(a). Our results also substantiate the statistically significant, if moderate, increases in the anti-atherogenic lipoprotein HDL-C and Apo A1. The large pooled safety analysis of 981 patients treated with evolocumab, albeit for only 12 weeks, did not reveal any new or unexpected safety signals. These results support large-scale phase 3 trials of evolocumab, including a cardiovascular outcomes trial.


This work was supported by Amgen, Inc.

Conflict of interest: E.A.S. has received consulting fees from Amgen, BMS/Adnexus Therapeutics, Genentech, Regeneron, and Sanofi related to PCSK9 inhibitors and his institution has received research funding related to PCSK9 clinical trials from Amgen, Alnylam, BMS/Adnexus Therapeutics, Genentech, Sanofi, and Regeneron. M.J.K. is an employee of Jacksonville Centre for Clinical Research, which has received research grants from Amgen. F.J.R. has received consulting fees from Amgen and Sanofi related to PCSK9 inhibitors, and his institution has received research funding related to PCSK9 inhibitor clinical trials from Amgen and Sanofi. E.M.R. has received fees as a speaker for Merck, AstraZeneca, and Amarin, and as a consultant to Regeneron and Sanofi. R.W. has no conflicts of interest to disclose.

D.S. has received research funding from Amgen, Abbott Products, AstraZeneca, Merck, Sharp and Dohme, and Sanofi Aventis; he has also received funding for educational programs from Abbott Products, AstraZeneca, Merck, Sharp, and Dohme, Pfizer Australia, and Roche and travel support from Merck, Sharp, and Dohme. He served on advisory boards for Abbott Products, Merck, Sharp, and Dohme, and Pfizer Australia. S.M.W., R.S., J.B.K., J.Y., T.L., M.A., and R.S. are employees of Amgen Inc. and have received Amgen stock/stock options. R.P.G. and M.S.S. have received research grant support through Brigham and Women's Hospital for the LAPLACE-TIMI 57 trial from Amgen, and have served as consultants for Amgen. R.P.G. has received honoraria for consulting from Regeneron and Sanofi and also received research grant support through Brigham and Women's Hospital, and honoraria for consulting and CME lectures from Daiichi-Sankyo and Merck. M.S.S. has also received research grant support through Brigham and Women's Hospital from AstraZeneca, AstraZeneca/Bristol-Myers Squibb Alliance, Bristol-Myers Squibb/Sanofi-Aventis Joint Venture, Daiichi-Sankyo, Eisai, Genzyme, GlaxoSmithKline, Intarcia, Merck, Sanofi-Aventis, and Takeda, and received honoraria for consulting from Aegerion, Amgen, AstraZeneca/Bristol-Myers Squibb Alliance, Bristol-Myers Squibb, GlaxoSmithKline, Intarcia, Merck, Pfizer, Sanofi-Aventis, and Vertex.


We thank Miranda Tradewell, PhD, of Complete Healthcare Communications, Inc. (whose work was funded by Amgen Inc.) and Meera Kodukulla, PhD, of Amgen Inc. for editorial support.


View Abstract