Lipoprotein A Testing (Lp(a)): Who, When, and Next Steps

Order lipoprotein a testing once for every adult, read the result in the lab's own units, and intensify standard risk-factor control when it runs high. As of October 10, 2026, no Lp(a)-lowering drug has shown an outcome benefit.

Who should get a lipoprotein(a) test?

Test every adult once, ideally with the first lipid panel. The 2026 ACC/AHA dyslipidemia guideline gives universal one-time measurement a Class 1, Level B-NR recommendation, and expert commentary calls it the first US guideline to say so.

The 2022 EAS panel says repeat testing generally adds no prediction. Exceptions include kidney or liver disease and acute infections.

The 2026 document is the newest US guideline I found (searched October 10, 2026), so its positions take precedence over 2018 for US practice.

How do you read the result?

Apply the reporting lab's own cutoffs to its own units. The 2026 guideline's management section defines elevated Lp(a) as 125 nmol/L or higher, or 50 mg/dL or higher, matching the 2018 guideline.

• Treat the cutoff as a gradient. The EAS panel says the 50 mg/dL line needs reconsideration. About 100 mg/dL (roughly 250 nmol/L) approximately doubles ASCVD risk, and levels above 180 mg/dL (430 nmol/L) carry lifetime risk comparable to untreated heterozygous familial hypercholesterolaemia.
• Never convert units. EAS recommends no standard factor, and a 2020 review states that all conversion factors depend on isoform size and that nmol/L is the more appropriate unit. It adds that no commercial assay is inherently isoform insensitive.
• Flag the assay when comparing sites. The 2026 guideline recommends assays insensitive to apo(a) isoforms and traceable to reference standards.

What do you do with an elevated result?

Intensify control of every other risk factor and test relatives. No Lp(a)-specific therapy has outcome evidence, so the number works as a risk enhancer.

The 2026 ACC/AHA guideline recommends these steps:

• Control modifiable risk factors optimally and early (Class 1).
• In clinical ASCVD above LDL-C and non-HDL-C goals on maximally tolerated statin, add a PCSK9 monoclonal antibody with proven cardiovascular benefit (Class 1, B-R).
• Test first-degree relatives when a patient has FH, premature ASCVD, or high Lp(a) (Class 1). Lp(a) is inherited in an autosomal dominant pattern.

Keep statins going through modest Lp(a) increases, and skip niacin: the EAS panel says statins should not be stopped and niacin is not recommended for Lp(a) lowering.

Which Lp(a)-lowering drugs have outcome data?

None. Pelacarsen is the only agent with a completed phase 3 outcomes trial, and it missed its primary endpoint.

• Pelacarsen: Novartis announced on September 4, 2026 that Lp(a)HORIZON (Lp(a) of 70 mg/dL or higher, established cardiovascular disease) missed its primary composite endpoint. The 90 mg/dL or higher subgroup was not reported, and full data are pending. TCTMD reported that no Lp(a) drug in development is FDA-approved.
• Olpasiran: OCEAN(a)-Outcomes is reported to read out in 2028 (estimated primary completion March 31, 2028). A 2024 report expecting results at the end of 2026 is outdated.
• Lepodisiran: ACCLAIM-Lp(a) is active and no longer recruiting, with primary completion listed for March 2029 (registry checked October 10, 2026).
• Zerlasiran and oral muvalaplin: phase 2 Lp(a)-lowering data only. No outcome results exist.

Where does this fall short?

No trial shows that lowering Lp(a) reduces events, and the universal-testing recommendation rests on nonrandomized evidence (Level B-NR). The 2026 guideline does not define "high Lp(a)" in its cascade section, full Lp(a)HORIZON data are pending, and phase 2 populations were mostly white and male. I read the 2026 guideline through a secondary summary because the journal pages blocked access, so confirm class and wording in the guideline text before citing it.