Will Less Aggressive Lp(a) Targets Do the Trick? Time and Clinical Trials Will Tell
A new analysis hints that trials should enroll patients with Lp(a) levels of > 100 mg/dL and more modest cuts will be meaningful.
Early calculations hoping to lay the groundwork for the lipoprotein(a) targets to be used in clinical trials may have been overestimated, a new mendelian randomization analysis suggests.
These numbers suggest that drugs that achieve less aggressive Lp(a) targets may yet prove clinically effective in reducing the risk of coronary heart disease.
A similar population-based study released last year by Stephen Burgess, PhD (University of Cambridge, England), and colleagues concluded that an absolute reduction of 101.5 mg/dL in Lp(a) would be required to have roughly the same 20% relative reduction in clinical event risk as a 38.67 mg/dL drop in LDL cholesterol.
But the findings published online today in JAMA Cardiology suggest that Lp(a) might only need to be lowered by 65.7 mg/dL to achieve the same comparative effect.
While the methods used in 2018 study were roughly the same as was done in the current genetic analysis, senior study author Florian Kronenberg, MD (University of Innsbruck, Austria), told TCTMD “the only thing which was a little bit suspicious was the extremely high Lp(a) levels they had in some of the cohorts.” One cohort had mean and median levels of the lipid marker upwards of three times as high as would be expected in a Caucasian population, something typically only observed in patients with nephrotic syndrome, he specified.
With the development of antisense oligonucleotides (ASO)—which have shown to reduce Lp(a) substantially—and plans for a phase III trial in the works, Kronenberg said studies like this one are imperative for the field to move forward. “The big question is always which patients should be included, how high should the Lp(a) be, and how strong should it be lowered—the same questions as they were in former times when statins became available. Because when you are doing such a trial, this trial costs a lot of money, and if you are planning the trials in the wrong way, then you have lost the money and maybe you have also caused some harm to the patients,” he said.
However, Sotirios Tsimikas, MD (University of San Diego, CA), an employee of Ionis Pharmaceuticals, which owns the subsidiary company that developed the ASO product, told TCTMD the results from both of these mendelian randomization analyses are “frankly irrelevant” given that they study primary prevention and the outcomes trials will only be including patients who have already experienced cardiovascular events and want to prevent a second one. “The patients who have already had one event are at much higher risk for having a second one, so nobody has actually studied the analysis to kind of estimate the size of a trial or how much Lp(a)-lowering we'll need to do,” he explained.
Tsimikas pointed out that the question is not whether or not Lp(a) is a risk factor but rather if lowering it will have a clinical impact. “Anything over 30 mg/dL is potentially a problem, and if you could do a trial over a lifetime, you probably don't need to lower the Lp(a) that much—you probably could just lower it 10 mg/dL—but you don't have the luxury of doing that in a trial,” he said. “You only have 3 to 5 years to kind of get an answer, so the drug has to be a lot more potent to reduce it.”
Given that ASOs have been shown to reduce Lp(a) up to 80%, he indicated there is no cause for concern, noting that “the drugs are already more potent than what these guys are suggesting [is needed] in their studies.” Tsimikas, who is involved in the design of the phase III trial, said: “We're going to be able to test the hypothesis, because we do have the right tools now with these new antisense oligonucleotides to do exactly what these papers are suggesting. . . . If we had a drug that only lowered Lp(a) [by] 40%, it might be a bigger issue.”
Counting the SNPs
For the study, Claudia Lamina, PhD (University of Innsbruck), Kronenberg, and colleagues included 13,781 individuals from the Lp(a)-GWAS-Consortium from five primarily population-based studies as well as 20,793 coronary heart disease patients and 27,540 controls. The median Lp(a) values in all studies ranged from 11-12 mg/dL.
The researchers restricted their analysis to single-nucleotide polymorphisms (SNPs) with a minor allele frequency of at least 1%, hence only including 27 of the 43 SNPs used by the Burgess et al analysis. After applying their methodology to the data set used by Burgess et al and confirming their previous findings, Lamina and colleagues reported that “possible differences in our further calculations are not explained by these missing 16 SNPs.”
Running their analysis on the current data set, they estimate that reducing Lp(a) by 10 mg/dL would be associated with 8.8% lifetime and 3.7% short-term lower risk of CHD. In other words, Lp(a) would need to be reduced by 65.7 mg/dL (95% CI 46.3-88.3) to have the same effect on CHD risk as a 38.67 mg/dL reduction in LDL-cholesterol—corresponding to a 45% lifetime and 22% short-term reduction in CHD risk.
Kronenberg said these results mean that future trials of Lp(a)-lowering should only include patients with a baseline level of 100 mg/dL and above, at least to start. Data from the FOURIER trial showed that people with Lp(a) levels above the median—but "not skyrocketing high Lp(a) levels”—see a stronger reduction in cardiovascular outcomes with evolocumab (Repatha; Amgen) compared with those below the median. This “means for me that even our estimate of 66 mg/dL might be really on the very, very safe side,” he said. “I would expect that even a lowering of Lp(a) by 50 mg/dL will show some positive effects, but it's very important, especially for the first trials, to start really with a selected group of patients with really high Lp(a) levels.”
‘A Key Starting Point’
In an accompanying editorial, George Thanassoulis, MD, MSc (McGill University Heath Center and Research Institute, Montreal, Canada), praises the role of this kind of research. “This natural (ie, mendelian) randomization sets up the perfect conditions to evaluate the causal role of Lp(a). Indeed, because Lp(a) is randomly assigned, we can identify the perfect counterfactuals to these patients: individuals who . . . inherited very few Lp(a)-increasing variants,” Thanassoulis says.
The estimates put forth in this study provide “a key starting point to plan future trials,” he writes. “Knowing that novel Lp(a)-lowering agents have been shown to lower Lp(a) by up to 90%, we can determine that the mean Lp(a) level of enrolled patients will need to be greater than 75 mg/dL; assuming similar event rates as prior statin trials, such RCTs would need to last roughly 5 years and enroll a similar number of patients as prior statin trials in secondary prevention.”
There are some caveats, however. First, Thanassoulis argues that the confidence intervals put forth in the study indicate a margin of error that could affect exactly how much Lp(a)-lowering would be required and future studies may need to enroll patients with “much higher” baseline levels. Additionally, “these estimates depend on whether Lp(a) has a similar chronic cumulative effect as LDL-C,” he writes, noting that Lp(a) is thought to be thrombogenic “and if such an acute pathophysiological effect truly exists, then the degree and duration of Lp(a)-lowering needed for benefit would be less than the estimated amount.”
Thanassoulis also agrees with Tsimikas in that the degree of Lp(a)-lowering needed for secondary prevention may in fact be different than what has been shown in primary prevention cohorts.
“Notwithstanding these limitations, it is nonetheless remarkable that we can validate novel therapeutic targets such as Lp(a) and also resolve key trial parameters prior to starting any RCT, based on such genetic studies,” Thanassoulis concludes. “By leveraging the natural randomization of Lp(a) variants, we have developed a robust evidence base that strongly supports Lp(a) as a clinically relevant therapeutic target in cardiovascular disease. However, the ultimate test of the Lp(a) hypothesis has yet to be conducted: a true randomized clinical trial of Lp(a)-lowering. Our patients with high Lp(a) are waiting.”
Lamina C, Kronenberg F, et al. Estimation of the required lipoprotein(a)-lowering therapeutic effect size for reduction in coronary heart disease outcomes: a Mendelian randomization analysis. JAMA Cardiology. 2019;Epub ahead of print.
Thanassoulis G. Using genetics to plan future randomized trials of lipoprotein(a) lowering—how much reduction, for how long, and in whom? JAMA Cardiology. 2019;Epub ahead of print.
- Kronenberg reports receiving personal fees from Kaneka and Amgen outside the published work.
- Lamina reports no relevant conflicts of interest.
- Tsimikas reports being an employee of Ionis Pharmaceuticals and holding patents related to oxidized phospholipids.
- Thanassoulis reports serving on advisory boards for Amgen, Sanofi-Regeneron, and Ionis and on speaker bureaus for Amgen and Sanofi; receiving grant funding from Ionis and Servier; receiving support from grants from the Canadian Institutes of Health, Heart and Stroke Foundation of Canada, and the National Institutes of Health/National Heart Lung and Blood Institute; and receiving salary support as a clinical research scholar from the Fonds de Recherche Québec–Santé.