Lp(a) Might Not Be Good at Predicting Incident CVD in Low-Risk Women
Only women with total cholesterol above 200 mg/dL in the analysis of more than 34,000 people were at increased risk from heightened Lp(a) levels.
Lipoprotein(a), previously shown to predict atherosclerotic cardiovascular disease and aortic valve disease events, might not be a reliable biomarker for primary prevention in women unless they also have high levels of total cholesterol, suggests a new analysis.
Studies of Lp(a) have been revving up of late, with extended data from FOURIER showing that patients with higher baseline Lp(a) levels may be more likely to derive benefit from PCSK9 inhibition and a mendelian randomization analysis demonstrating that the clinical benefit of lowering Lp(a) is proportional to the absolute reduction.
For this study, published in the July 17, 2018, issue of the Journal of the American College of Cardiology, Nancy Cook, ScD (Brigham and Women’s Hospital, Boston, MA), and colleagues analyzed data from three cohorts: the Women’s Health Study (WHS; n = 24,558), the Women’s Health Initiative (WHI; n = 1,815 cases) observational study, and the JUPITER trial (n = 2,569 women and n = 5,161 men).
Within the WHS group, the researchers found a curvilinear association between increased incident CVD risk and Lp(a) levels of at least 50 mg/dL. This relationship, however, was only observed among women with total cholesterol levels of 220 mg/dL or greater or apolipoprotein B (apoB) levels of at least 100 mg/dL. Similar findings were seen in the WHI cohort, with the risk of CVD increasing for those with Lp(a) levels 75 mg/dL or greater and total cholesterol or apoB levels were above 220 mg/dL or at least 100 mg/dL, respectively.
Results from the women in JUPITER did not match those from WHS and WHI with statistical significance, but a strong association between Lp(a) and CVD risk in the JUPITER men was observed, even in those with total cholesterol levels below 220 mg/dL. The researchers assessed whether hormone replacement therapy in women affected any of their results, concluding that it did not in the WHS and WHI cohorts (no hormone replacement therapy was used in JUPITER).
“In research, people will need to look to see if this can be verified in other data sets and also to understand the biology behind this,” Cook told TCTMD. “In the clinical world, I think they need to be careful and test this [in the] trials of Lp(a) lowering agents. . . . They need to look at the differences in men and women and look at this interaction with the other lipids and make sure they have enough power to do so.”
Additionally, this study “raises some questions about the interactions among these various lipids,” she continued. “There does seem to be an interaction here, in that Lp(a) only seems to be predictive of cardiovascular disease among those with higher lipid levels and higher cholesterol or apoB levels.”
As for why there would be a gender difference, Cook said she isn’t sure just yet. “We thought it might be due to hormone therapy use, but it didn't really seem to explain it. And also, as other investigations have found, we saw that risk was really only increased among those with [Lp(a)] levels say of about 50 [mg/dL] or higher. That's been more consistent,” she explained. “If it's only people with higher levels of cholesterol that have the increased risk, that narrows the field even smaller. So, it might have implications for people who would benefit from Lp(a)-lowering therapies.”
In an editorial accompanying the study, Robert Rosenson, MD (Icahn School of Medicine at Mount Sinai, New York), and Santica Marcovina, PhD, ScD (University of Washington School of Medicine, Seattle), write that the “analysis demonstrates a consistent interaction between Lp(a) cholesterol and incident CVD” in selected women and that “the lack of association between Lp(a) in JUPITER may have been due to the smaller sample size.”
Lp(a), they explain, consists of an LDL-like particle plus the protein apolipoprotein a, or apo(a), which is “characterized by a high level of size heterogeneity.” Prior research has linked smaller apo(a) isoforms to increased rates of coronary heart disease and aortic valve stenosis.
While all three studies in the new analysis calculated Lp(a) concentrations by using a method designed to minimize the impact of apo(a) size heterogeneity, Lp(a) values were reported in mg/dL of total mass in WHS and WHI, but in JUPITER were measured in nmol/L of the Lp(a) protein and then converted. The editorialists point out that “the clinical cut points in the different studies are not directly comparable because a single, arbitrary factor cannot be reliably used to convert values from nmol/L of Lp(a) protein to mg/dL of total Lp(a) mass.”
Commenting to TCTMD, Seth Martin, MD (Johns Hopkins Medicine, Baltimore, MD), who was not involved in the study, said the assay his institution uses returns Lp(a) concentrations in nmol/L. “But then to see papers that are reporting Lp(a) in mg/dL, it just makes it a little tricky [as a clinician] to know how to apply it to my patients because I know there's not a reliable conversion factor,” Martin said.
While the different calculations of Lp(a) might not “really impact the results of this particular paper,” he said, “it is definitely something to bear in mind as this field moves forward.
“It's such a basic first principle that if we want to go after Lp(a), we want to be measuring it right and everyone wants to be measuring the same, and if you're a patient and you go see me in Baltimore or you so see someone in New York or you see someone in California, that if you get it measured, you're going to get a similar result because this is truly standardized across labs,” Martin continued. “So I think this is really imperative to get this sorted out very quickly to harmonize the way Lp(a) is being measured across all these different areas from clinical practice to the population studies and the clinical trials.”
Rosenson and Marcovina argue that the guidelines for total cholesterol and apoB set forth in this paper “are of limited usefulness because there are no reliable methods to directly measure cholesterol in Lp(a) nor are there population-based reference values for Lp(a) cholesterol.”
Moreover, the three populations in this study were low risk and not treated with statins, the editorialists note. “Thus, cutoff values in statin-treated patients with clinical atherosclerotic CVD may be different than these thresholds or influenced by other Lp(a) characteristics, such as Lp(a) heterogeneity or functional characteristics.”
Ultimately, Martin said, “this particular study would suggest that in lower-risk women that [Lp(a)] doesn't have a lot of predictive value, but I think that overall when you look at the literature, there's a lot of reason to be optimistic about Lp(a) as a causal target of therapy based on the really elegant mendelian randomization analyses that have been done.”
Also, the results are “really pushing us to take pause about how broadly should this be applied,” he continued. “If someone is a low-risk individual and their standard lipid profile cholesterol or apoB is not super high, . . . it's helping us start to figure where to put the guidepost on the population we define that's likely to benefit from this therapy.” However, he noted that future efforts will likely be based in secondary prevention and not in primary prevention as with this paper.
Cook too said she is optimistic about Lp(a)’s potential in clinical medicine and in future studies. “I do think this is still a promising area of research, and I think there's a lot of biology here to come into play to be discovered, and a lot is known about it as well already,” she said. “There seem to be indications that there is a causal factor related to cardiovascular disease.”
Future studies should include more cohorts looking at gender differences and interactions between CV disease risk, Lp(a), and total cholesterol, she said. “There should be more basic science research into the interactions behind these, and also looking at the therapies. Different therapies affect the lipids differently. Statins actually slightly increase Lp(a), but I understand PCSK9 inhibitors lower Lp(a) as well as other lipids. So those interactions need to be understood clearly.”
Cook NR, Mora S, Ridker PM. Lipoprotein(a) and cardiovascular risk prediction among women. J Am Coll Cardiol. 2018;72:287-296.
Rosenson RS, Marcovina SM. Refining lipoprotein(a) associated cardiovascular risk in women. J Am Coll Cardiol. 2018;72:297-299.
- The WHS (Women’s Health Study) was supported by grants from the National Heart, Lung, and Blood Institute (NHLBI) and the National Cancer Institute. The WHS blood analysis was supported by the Donald W. Reynolds Foundation and the Leducq Foundation.
- The WHI (Women’s Health Initiative) program is funded by the NHLBI, National Institutes of Health, and U.S. Department of Health and Human Services. The WHI blood analysis was supported by the NHLBI.
- The JUPITER (Justification for Use of Statins in Prevention) trial was supported by AstraZeneca, which had no role in the design or conduct of the present study. Quest Diagnostics Nichols Institute performed the Lp(a) measurements in JUPITER at no additional cost to the study. The study was also funded by grants from the NHLBI.
- Cook reports no relevant conflicts of interest.
- Rosenson reports receiving research support from Akcea, Amgen, The Medicines Company, Regeneron, and Sanofi; participating in advisory boards for Akcea, Amgen, CVS Caremark, and Regeneron; and receiving honoraria from Akcea, Kowa, and Pfizer.
- Marcovina reports holding research contracts with Ionis and Amgen and receiving consulting fees from Denka Seiken and MedTest Diagnostics.
- Martin reports serving on scientific advisory boards for Amgen and Sanofi Regeneron for PCSK9 inhibitors as well as for Akcea Therapeutics and Quest Diagnostics.