Preclinical and clinical studies strengthen the promise of ANGPTL3 as a therapeutic target
Dewey et al., NEJM May 24, 2017 | Genetic and Pharmacologic Inactivation of ANGPTL3 and Cardiovascular Disease
Graham et al., NEJM May 24, 2017 | Cardiovascular and Metabolic Effects of ANGPTL3 Antisense Oligonucleotides
Genome-wide association studies and exome sequencing have identified associations between genetic variants of angiopoietin-like proteins (ANGPTLs) and lipoprotein and triglyceride levels. ANGPTLs are important regulators of lipoprotein metabolism.
ANGPTL3 is an endogenous inhibitor of lipoprotein lipase (LPL). Loss-of-function (LOF) variants in LPL have been shown to increase the risk of coronary artery disease (CAD), whereas gain-of-function variants decrease this risk [1-3]. Rare LOF variants of ANGPTL3 have been reported to be associated with lower triglyceride (TG) levels and lower LDL-c and HDL-c levels in humans. In ApoE-deficient mice, deletion of Angptl3 reduced development of atherosclerosis . LOF variants in ANGPTL3 have also been associated with beneficial metabolic effects, such as higher LPL and endothelial lipase activity, higher insulin sensitivity and lower levels of free fatty acid in the serum .
Dewey et al examined the relationship between ANGPTL3 LOF variants and CAD in a sample from the large DiscovEHR US health care system, with follow-up studies in four population cohorts. In addition, effects of pharmacologic antagonism of ANGPTL3 with the monoclonal antibody evinacumab (REGN1500) on lipid metabolism and atherosclerosis were evaluated in a mouse model of atherosclerosis, and on lipid levels in human volunteers.
Graham et al evaluated the effect of antisense oligonucleotide (ASOs) directed against Angptl3 mRNA on plasma lipid levels, TG clearance, liver TG content, insulin sensitivity and atherosclerosis in mice, and they conducted a randomized trial with subcutaneous injections of placebo or ASO targeting ANGPTL3 mRNA in 44 humans.
Dewey et al.
- 13 distinct LOF variants in ANGPTL3 were identified in 58335 adult DiscovEHR study participants of European ancestry. Among these variants were 4 premature stop variants, 6 open reading frame-shifting insertion-deletion variants, and 3 splice variants.
- 246 individuals were heterozygous for LOF variants, which corresponds to a total carrier frequency of 1 in 237 participants. No homozygotes or compound heterozygotes were found.
- Carriers of ANGPTL3 LOF variants showed significantly lower TG levels (27% lower than non-carriers), LDL-c (9%) and HDL (4% lower), after adjustment for covariates.
- Of the 13102 persons who developed CAD, 43 carried a LOF variant (cumulative carrier frequency: 0.33%), and 183 of 40430 controls (frequency: 0.45%). An ANGPTL3 LOF was associated with 41% lower adjusted odds of CAD (95%CI: 0.41-0.85, P=0.004). A non-significant, numerically lower odds of myocardial infarction (MI) was noted (carrier frequencies: 0.36% in those with MI, and 0.45% in controls, OR: 0.66, 95%CI: 0.39-1.06, P=0.09).
- In four independent population studies, the association between ANGPTL3 LOF variants and CAD, was evaluated, in a total of 23317 CAD cases and 107166 controls. In these populations, a non-significant, numerically lower odds of CAD was seen among variant carriers (OR: 0.63, 95%CI: 0.39 – 1.02, P=0.06). Combining all data in an inverse-variance-weighted fixed-effects meta-analysis yielded an overall odds ratio of 0.61 (95%CI: 0.45-0.81, P<0.001).
- Treatment of APOE*3Leiden.CETP mice with evinacumab resulted in significantly lower total cholesterol (difference: -52%) and TG (difference: -84%) as compared with the control antibody. Evinacumab was associated with a greater decrease in atherosclerotic lesion size than the control antibody (difference: -39%). Severe type IV and V lesions in evinacumab-treated animals showed a significantly greater decrease in necrotic content (difference: -45%), but no significant differences were seen in macrophage content, collagen content or smooth muscle cell area.
- In healthy human volunteers with mildly to moderately elevated TG levels (150 to 450 mg/dL) or LDL-c (≥100 mg/dL), dose-dependent placebo-adjusted reductions in TG levels (up to 76%) and LDL-c (up to 23%) were observed in those treated with evinacumab.
Graham et al.
- Mouse Angptl3 ASO led to significantly and selectively suppressed hepatic Angptl3 mRNA expression (69-91%) in wild-type C57BL/6 mice, Ldlr-/- mice, Apoc3-/- mice and mice overexpressing human apoCIII.
- Administration of the Angptl3 ASO yielded reductions in levels of TGs (35 to 85%), LDL-c (7 to 64%) and HDL (3 to 23%). Moreover, the TG-content within VLDL, IDL and LDL particles was reduced in some of the mouse models, including in LDLr-deficient mice, indicating that the effects are independent of the LDL receptor pathway.
- A significant reduction in liver TG secretion was seen in C57BL/6 mice who received Angptl3 ASO. Moreover, TG clearance was enhanced, likely independent of whole lipoprotein uptake, because the rate of VLDL clearance did not change.
- Mice overexpressing human apoCIII that were treated with the Angptl3 ASO showed significantly reduced plasma TG and LDL-c levels.
- In mice with diet-induced obesity, Angptl3 ASO treatment resulted in higher insulin sensitivity, as compared with control obese mice, not receiving the ASO. Liver TG accumulation was 81% lower in mice receiving the ASO.
- Ldlr-/- mice treated with Angptl3 ASO showed retarded progression of atherosclerosis, relative to mice treated with the control ASO (by 52% or 37% depending on the weekly dose).
- In the phase I trial, 9 participants were randomized to single-dose administration of ANGPTL3 ASO and 3 to placebo, and 32 to a multiple-dose group (24 active, 8 placebo). Dose-dependent reductions of ANGPTL3 protein were seen. TG levels were reduced from baseline by a maximum of 63.1%, apoCIII by a maximum of 58.8%, non-HDL-c by a maximum of 36.6% and apoB by a maximum of 25.7%. Between the final dose at day 36 and the end of follow-up at day 127, plasma levels slowly returned to normal. No serious adverse events were documented.
Dewey and colleagues reported that in over 400 carriers of heterozygous LOF variants in ANGPTL3, a lower risk of CAD was observed as compared with non-carriers. This observation is in line with the lower lipid levels and slower atherosclerosis progression seen in dyslipidemic mice treated with evinacumab. Moreover, the dose-dependent effects of treatment with evinacumab in humans appear to recapitulate the lipid phenotype of carriers of an ANGPTL3 LOF variant. These findings suggest that the hypolipidemic profile induced by genetic or therapeutic antagonism of ANGPTL3 consisting of reduced levels of HDL-c, LDL-c and TG, is anti-atherogenic.
The study by Graham and colleagues demonstrated that inhibition of ANGPTL3 mRNA resulted in favorable cardiometabolic effects in various mouse models and in healthy humans. The preclinical studies showed that hepatic suppression Angptl3 protein production in mice resulted in lower levels of TG, LDL-c, non-HDL-c and VLDL-c. A lower liver TG content was seen, higher insulin sensitivity and a reduction of atherosclerosis progression. Similar effects were seen in the phase I trial in humans.
These studies provide a rationale for continued development of an ANGPTL3-targeted therapy for persons with elevated levels of TG-rich lipoproteins, in order to lower residual CV risk in patients already taking recommended medical and preventive therapies. Both the antibody-based and the ASO-approach deserve further study.