Physician’s Weekly spoke with first author Nicholas A. Marston, MD, an Associate Physician in the Cardiovascular Division at Brigham and Women’s Hospital (Boston, USA), an Instructor of Medicine at Harvard Medical School, and an Investigator at the TIMI Study Group.
Patients treated with PCSK9 inhibitors had reduced risk for venous thromboembolism (VTE), especially if they had elevated lipoprotein(a). However, individuals with loss-of-function mutations of the LDL receptor gene (LDLR) and/or a high score on a polygenic model, retain risk of increased cardiovascular risk despite intensive lipid-lowering therapy.
Presenting at the American Heart Association Scientific Sessions 2020, Dr Nicholas Marston pointed out that although there seems to be no correlation between baseline LDL and VTE risk reduction, patients with elevated Lp(a) at baseline treated with PCSK9 inhibitors experienced Lp(a) lowering in addition to decreased risk for VTE, but patients with lower baseline Lp(a) did not [1, 2].
In a 2-trial meta-analysis of the FOURIER trial (NCT01764633, evolocumab, n=27,564) and the ODYSSEY OUTCOMES trial (NCT01663402, alirocumab, n=18,924), subjects receiving a PCSK9 inhibitor had a 31% risk reduction for VTE compared with those in the placebo arm (HR = 0.69; 95% CI 0.53-0.90; P=0.007).
In a post-hoc analysis of FOURIER, PCSK9 inhibition had no effect on VTE within the first year of treatment for atherosclerotic CVD (HR 0.96; 95% CI 0.57-1.62). Beyond that first year, however, patients experienced a 46% risk reduction (HR 0.54; 95% CI 0.33-0.88; P=0.014). The overall risk reduction was 29% (HR 0.71; 95% CI 0.50-1.00; P=0.05).
Among patients with elevated baseline Lp(a) levels, PCSK9 inhibition reduced Lp(a) by 33 nmol/L and the risk for VTE by 48% (HR 0.52; 95% CI 0.30-0.89). Among patients with lower baseline Lp(a), researchers found a Lp(a) reduction of just 7 nmol/L and no effect of evolocumab on VTE risk (P for interaction = 0.087; P for heterogeneity=0.037). This interaction led researchers to determine that when Lp(a) was modeled as a continuous variable, there was a mediating effect of baseline Lp(a) concentration on the magnitude of VTE risk reduction (P=0.04).
After an exploratory genetic analysis of the FOURIER trial, researchers found that a polygenic risk score was able to identify patients at increased genetic risk for VTE and who derived greater relative (P for interaction = 0.04) and absolute VTE reduction (P for heterogeneity = 0.009) from evolocumab compared with those without high genetic risk.
Within the FOURIER cohort, there were 111 (0.8%) patients identified with an LDLR loss of function mutation, representing an approximately 4-fold enrichment compared to the general population. These patients had mean baseline LDL-C levels 62 mg/dl higher than those without an identified LDLR mutation (160 vs. 98 mg/dl, P<0.0001) and a 70% increased risk of CV events (HR 1.70; 95% CI 1.04-2.79; P=0.03). When polygenic risk was combined with LDLR mutation status there was a significant gradient of risk identified (P=0.001).The effects were additive; patients without LDLR mutation or high polygenic risk had a 2.5-year Kaplan-Meier event rate of 8.6%; those with either a LDLR mutation or high polygenic risk score had an intermediate 10.0-12.9% event rate; but those with both a LDLR mutation and a high polygenic risk rate had a 17.1% event rate.
Physician’s Weekly talked with Dr Marston to gain perspective on these new findings:
“The FOURIER trial (NCT01764633) which studied the PCSK9 inhibitor evolocumab in over 27,000 patients and, while that trial was positive for reducing MI, stroke, and CV death as a composite endpoint, it did not look at venous thromboembolism as a prespecified endpoint. One prior trial in particular, JUPITER (NCT00239681), had looked at rosuvastatin and showed that it did reduce the VTE and the mechanism was not resolved [3]. The findings from JUPITER gave us the idea and hypothesis to interrogate whether the same was true for PCSK9 inhibition. Comparing VTE risk after PCSK9 inhibition versus statins also allows us to potentially tease out what the mechanism is. Both statins and PCSK9 inhibitors lower LDL but statins are also anti-inflammatory, PCSK9 inhibitors are not, and while PCSK9 inhibitors lower LP(a), statins do not. These differences provide a nice model for testing it but the first step was just to see if in fact, it did reduce venous thromboembolism and we found that it did, in FOURIER, by about 30% relative risk reduction. Fortunately, the ODYSSEY OUTCOMES trial (NCT01663402), the other PCSK9 inhibitor trial, which looked at patients after MI to see if it reduced recurrent events, they had a similar finding that they had published in an abstract so we were able to confirm our findings in FOURIER and combine it with the data from ODYSSEY in kind of a two trial meta-analysis. That gave us more confidence that the signal was true, that PCSK9 inhibitors do reduce VTE by about 31%, a pretty impressive reduction. That was the main finding and I think the most unique piece of this.”
Was it LDL or Lp(a)?
“I think it will still require some confirmation but we looked at cross quantiles of LDL as well as Lp(a) and we found that when we looked at cross quantiles of LDL, we really did not see a gradient of VTE reduction. However, when we looked at cross categories of baseline Lp(a) we did see more VTE reduction. Just to give you an example: when we looked by baseline Lp(a) in those who had below the median baseline Lp(a), those patients did not derive any VTE reduction with PCSK9 inhibitor. However, patients with high baseline Lp(a) levels, they had much higher Lp(a) reduction and that is where we saw the VTE reduction. In both of those groups, low baseline Lp(a) and high baseline Lp(a), they had the same reduction in LDL cholesterol. It allowed us to hypothesize that this is more likely related to Lp(a) than it is to LDL. Again, I think hypotheses generally need to be confirmed, but it is certainly interesting, especially now that we have specific therapies for targeting Lp(a). It will be very interesting to see whether VTE risk is reduced in upcoming trials testing Lp(a)-lowering approaches, not necessarily as a primary endpoint but as a pre-specified secondary endpoint. That would confirm the hypothesis that Lp(a) is involved in VTE.”
Clinical implications?
“The absolute event rate of VTE in the population is quite low, so while our relative risk reduction is large, the absolute risk reductions were more modest. The strategy of treating everyone with PCSK9 inhibitors to try to reduce VTE is probably too expensive at this point and would not be the approach, but as PCSK9 inhibition is used increasingly more frequently to reduce myocardial infarction, ischemic stroke, and other cardiovascular diseases, I think that a reduced VTE burden is going to be a nice secondary benefit in patients who are receiving PCSK9 inhibitors. I think that that is useful and clinically relevant.”
How do you interpret the polygenic risk?
“We published a GWAS analysis looking at the VTE and the UK biobank and found around 200 single nucleotide polymorphisms (SNPs) that are associated with VTE. If you combined those SNPs, each of them with a small effect but adding them together into a polygenic risk score when summed up can give you a larger effect [4]. There was a 2-fold increase risk in VTE in individuals with a high polygenic risk score; how can we apply polygenic risk scores into the clinical setting and make them more than just a research tool is one of my personal interests. We wanted to see if there was pharmocogenomic interaction so that the patients who were going to have the greatest benefit for VTE reduction with PCSK9 inhibitor, are they those who have the highest genetic risk? I mentioned the absolute risk of VTE is quite low, but if you can identify a population that has a higher absolute risk, then they may derive greater benefit. That was what we tested and that was what we found. The patients with the highest genetic risk in the top one third had 55% relative risk reduction in VTE, so that would really be the group you would want to target if you were to try to use this in a preventive strategy, rather than just treating everyone. Potentially, you would target patients with genetics But this needs to be confirmed, again, this is an exploratory analysis. I think it is interesting and exciting to think about how genetics could potentially provide us with more precision therapies.”
References
- Marston NA, et al. Cardiovascular Outcomes in Patients With Established Atherosclerosis and LDLR Loss of Function: Results From the FOURIER Trial. Circulation, Abstract 13588, AHA Scientific Sessions. 2020;142:A13588.
- Marston NA, et al. Association Between Triglyceride Lowering and Reduction of Cardiovascular Risk Across Multiple Lipid-Lowering Therapeutic Classes: A Systematic Review and Meta-Regression Analysis of Randomized Controlled Trials. Circulation. 2019 Oct 15;140(16):1308-1317.
- Glynn RJ, et al. A randomized trial of rosuvastatin in the prevention of venous thromboembolism. N Engl J Med. 2009 Apr 30;360(18):1851-61.
- Marston NA, et al. Predicting Benefit From Evolocumab Therapy in Patients With Atherosclerotic Disease Using a Genetic Risk Score: Results From the FOURIER Trial. Circulation. 2020 Feb 25;141(8):616-623.
