Lp(a) and LPA Genetic Risk Score Both Predict Incident ASCVD: UK Biobank
The risk score doesn’t add much beyond Lp(a) levels but may be used as a proxy when genetic testing is more prevalent.
Adding a measurement of lipoprotein(a) or even a genetic risk score for LPA, the gene that encodes apolipoprotein(a), can modestly improve the prediction of atherosclerotic cardiovascular disease (ASCVD) risk when used on top of traditional risk calculators, according to new data.
The LPA genetic risk score, which is comprised of 43 single nucleotide polymorphisms (SNPs), did not lead to additional prognostic information when combined with measured lipoprotein(a), but investigators say the genetic risk score can be used to fill out the picture of a patient’s risk if a direct measure of lipoprotein(a) isn’t available.
“It turns out both predict cardiovascular disease,” senior investigator Pradeep Natarajan, MD (Massachusetts General Hospital, Boston), told TCTMD. “Surprisingly, they both predict risk similarly, and that’s probably because Lp(a) is so highly heritable and we have a genetic test that has fairly high diagnostic yield for Lp(a). Most of the Lp(a) variation is explained by genetics.” While both Lp(a) and the LPA genetic score predicted ASCVD risk to a similar extent, that’s not the case for other biomarkers, he said, noting there are other strong nongenetic determinants of LDL cholesterol that also influence cardiovascular disease, for example.
The unique aspect of this study, which was presented yesterday at the virtual European Atherosclerosis Society Congress 2020 and published simultaneously in JAMA Cardiology, is that the genetic test score can capture risk to the same extent as a direct measure of Lp(a), the researchers say. While that might be an interesting scientific finding, is there a clinical implication given that doctors can simply order a Lp(a) test without having to worry about the patient’s genetics? The answer, it turns out, is “yes and no,” according to Natarajan.
“If Lp(a) levels are known, there’s probably no need for a genetic test because there was no additional prognostic utility with the LPA genetic risk score,” he said. “But if you look at the way preventive cardiology is moving and how widespread genetic testing is, in the not-too-distant future—and it happens a little bit today—people who’ve never had their lipids checked will be showing up in the office with a full genetic profile largely [obtained] through direct-to-consumer genetic testing or research testing if they have access to the data. One could say that from that simple baseline information, you get the prediction information you’d need for Lp(a).”
At present, elevated Lp(a) is defined as ≥ 50 mg/dL or 125 ≥ nmol/L and considered a “risk-enhancing” factor that might help guide treatment in the 2018 American College of Cardiology/American Heart Association (ACC/AHA) cholesterol guidelines. The European Atherosclerosis Society/European Society of Cardiology takes a much broader screening approach, however, and recommended that all adults over age 40 be screened for Lp(a) to identify those with elevated levels (class IIa, level of evidence C).
Lp(a) vs LPA Genetic Risk Score
In the latest study, Natarajan said they were motivated to reevaluate Lp(a)’s role in risk prediction in a much larger contemporary dataset. Using data from the UK Biobank, which is a prospective, observational study of predominantly white/European adults aged 40 to 69 years, the researchers also looked at the comparative predictive utility of directly measured Lp(a) and the LPA genetic risk score—alone and combined—with ASCVD risk. The overall study population included 374,099 people, of whom 204,355 were female. During a median follow-up of 11.1 years, 5.1% of participants developed incident ASCVD (PAD, CAD, MI, ischemic stroke, and cardiovascular death).
Overall, the LPA genetic risk score was associated with measured levels of Lp(a) in all subjects regardless of race/ethnicity. For individuals of white/European ancestry without prevalent ASCVD who were not using cholesterol-lowering medication, a 120-nmol/L increase in either the measured Lp(a) levels or based on the LPA genetic risk score was associated with increased risk of ASCVD. The hazard ratios per the 120-nmol/L increase in measured Lp(a) and the LPA risk score were 1.26 (95% CI 1.23-1.28) and 1.29 (95% CI 1.26-1.33), respectively. The association between the LPA genetic score and risk of incident ASCVD was attenuated when the models adjusted for measured Lp(a) levels.
When researchers added measured Lp(a) levels and the LPA genetic risk score individually to the QRISK3 risk-prediction calculator, there were modest improvements in the predicted risk of incident ASCVD. In both instances, the area under the receiver operating characteristic curve increased from 0.640 to 0.642. Similar results were observed when Lp(a) and the LPA genetic risk score were added to the ACC/AHA pooled cohort equation for the assessment of ASCVD risk.
Børge Nordestgaard, MD, DMSc (Copenhagen University Hospital, Denmark), who wasn’t involved in the study, noted that the European guidelines are the first to ever recommend measuring Lp(a) in everybody once in a lifetime. And while there are no US Food and Drug Administration-approved therapies for reducing the atherogenic lipoprotein, he said, elevated levels do provide clinical information.
“If you have very high Lp(a), whether it’s in primary or secondary prevention, it tells you they have much higher risk than [a patient] with the same risk factors but with lower Lp(a),” he told TCTMD. “So the advice would be to do whatever you can to bring down LDL cholesterol as much as possible, and to try to reduce triglyceride levels as well, and then [do] all the usual stuff such as not smoking, lowering weight, and eating sensible.”
Although the European guidelines recommend screening more broadly than in the US, Nordestgaard said elevated Lp(a) is still considered a “risk-enhancing” factor that identifies patients in which “physicians should do even more than they would otherwise to lower cardiovascular risk.”
Gazing Into the Crystal Ball
For Natarajan, Lp(a) as a risk-enhancing factor is good guidance at present rather than useful in broad, population-based screening. For individuals with high Lp(a), the biomarker has “pretty good” positive predictive value, he said, although it’s not as strong a predictor as LDL cholesterol, diabetes, hypertension, and some other traditional risk factors. “It makes more sense as a risk-enhancing factor, which is meant to up-classify risk,” said Natarajan. “That’s going to be most helpful where you think a statin might be helpful, but you’re not sure, so you’re looking for other factors that’ll help up-classification.”
When genetic testing becomes more widely available, more individuals with an elevated index of Lp(a) will be identified. Natarajan cautioned that they can’t make any inferences about the effects of potential treatments. For example, will a patient with directly measured elevated Lp(a) respond differently to Lp(a)-lowering therapy than a patient with a high index of Lp(a) based on the LPA genetic risk score? It’s unknown at the moment, he said, but possible given that some might have more enriched LPA alleles than others.
In terms of the future, Natarajan thinks their research could have implications for clinical research. Instead of relying on physicians to measure Lp(a) before referring patients to a study, a clinical trialist could reach out to a research consortium or a private company to screen patients based on the LPA risk score. “That could be a pretty efficient trial design for Lp(a)-related therapies,” he said.
Nordestgaard said he’s hopeful that the randomized trials aimed at lowering Lp(a) will translate into a reduction in ASCVD events, especially since some drugs, such as a promising antisense oligonucleotide, have been shown to lower Lp(a) levels by as much as 80%. “I’m optimistic, but of course I’m also realistic, so I’m not going to say anything until I see the final results published,” said Nordestgaard.
Novartis recently licensed the rights to the antisense oligonucleotide previously known as AKCEA-APO(a)-LRx and that agent, now called TQJ230, is currently being tested in the randomized, phase 3 Lp(a)HORIZON study. The PCSK9 inhibitors also lower Lp(a) by roughly 25% but none of the randomized trials focused on the cardiovascular impact of reducing Lp(a) alone, said Nordestgaard.
Trinder M, Uddin MM, Finneran P, et al. Clinical utility of lipoprotein(a) and LPA genetic risk score in risk prediction of incident atherosclerotic cardiovascular disease. JAMA Cardiol. 2020;Epub ahead of print.
- Natarajan reports grants from Amgen, Boston Scientific, grants/personal fees from Apple, personal fees from Novartis and Blackstone Life Sciences, and support from Vertex.
- Nordestgaard reports consulting for Akcea, Amgen, Sanofi, Regeneron, and AstraZeneca.