new PCSK9 inhibitor pelacarsen withdrawn after trial

 Pelacarsen, the most intensive Lipoprotein (a) antagonist (hereinafter Lp(a)) was just found to be clinically ineffective in the Lp(a) HORIZON Phase III study from Novartis because it "did not meet its primary endpoint of reducing the risk of cardiovascular events, a composite of cardiovascular heath, non-fatal myocardial infarction, non-fatal stroke, and urgent coronary revascularization requiring hospitalization, compared to placebo": https://www.globenewswire.com/news-release/2026/09/04/3356808/0/en/novartis-announces-lp-a-horizon-phase-iii-topline-results-for-pelacarsen-in-patients-with-elevated-lp-a-and-established-cardiovascular-disease-cvd.html

for the New York Times comments: https://www.nytimes.com/2026/09/04/science/heart-drug-fails-novartis-pelacarsen.html?unlocked_article_code=1._FA.allS.qNH-hHuEvKoV&smid=url-share

Details:

-- there are not many details in the Novartis statement, though the plan is to present the data at the annual meeting of the American Heart Association in November, along with a concurrent paper that will be published "in a leading medical journal"

-- in this Lp(a) HORIZON Phase III study, 8,323 people who had high Lp(a) levels and pre-existing heart disease were randomized to injection of the drug vs placebo monthly

-- of all of the upcoming PCSK9 inhibitors, pelacarsen is the one with the most profound lowering of Lp(a) levels by 80%:  https://www.sciencedirect.com/science/article/pii/S1933287425003228

-- the overall population in the Lp(a) HORIZON study was evaluated both for those with baseline Lp(a)>70 mg/dL or Lp(a) >90 mg/dL

Commentary:

-- these unfortunate results could potentially be attributed to several reasons:

    -- is Lp(a) just a surrogate marker for atherosclerotic disease and not really the culprit?

        -- we do know that surrogate markers are often flawed: https://gmodestmedblogs.blogspot.com/2024/06/using-surrogate-markers-for-disease-are.html

        -- we even know several cardiology studies documenting the pitfalls of using surrogate markers instead of clinical endpoints (eg the use of torcetrapib, or one finding that lowering inflammation with a novel anti-inflammatory agent ziltivekimab, an IL-6 inhibitor, that did not improve cardiovascular outcomes: https://gmodestmedblogs.blogspot.com/2026/08/is-new-pcsk9-pill-ready-for-prime-time.html

            -- this last blog refers to another new PCSK9 inhibitor enclicitide, which is a pill that lowers the Lp(a) as has been shown in prior studies, and it was approved by the FDA. this is a quite surprising approval since there are no data on clinical adverse cardiovascular events, and that clinical data will not be available for 2 years. this is contrary to prior FDA approaches and does open the door to a surrogate marker being as important as actual clinical events. see https://gmodestmedblogs.blogspot.com/2026/08/is-new-pcsk9-pill-ready-for-prime-time.html for more details. But on the surface, the FDA seems to have approved for general use a drug that has no clear benefit...

    -- Lp(a) is really not so bad

        -- that possibility has been defused by many studies done over many years. after all, Lp(a) uniquely combines an atherogenic LDL-like particle with kringles, which it turns out are not a breakfast cereal, but are a genetically determined component of Lp(a) in terms of their number present that are structurally similar to plasminogen, competes with them for plasminogen receptors, but are pro-thrombotic (see https://gmodestmedblogs.blogspot.com/2025/06/high-lpa-increases-risk-of-recurrent.html). not surprisingly this combination of a lipid moiety that is both atherogenic and thrombogenic is a particularly bad actor

           -- and high Lp(a) levels are associated with chronic inflammation, a significant precursor to cardiovascular disease

        -- old and current Lp(a) studies confirm that Lp(a) is cardiovascular-toxic with increased atherosclerotic cardiovascular events (ASCVD). the recent 2026 ACC/AHA/etc/etc Dyslipidemia guideline (lipid management ACC guideline2026 in dropbox, or DOI: 10.1161/CIR.0000000000001423) encourages testing Lp(a) levels in all people at least once, and this guideline summarized the following regarding Lp(a) risk (there is a broad conversion for nmol/L to mg/dL of between 2.15 to 2.5 in different studies; the conversion below is in this range, though varies a bit for unclear reasons):

            -- Lp(a) 430 nmol/L, or 180 mg/dL): 4-fold ASCVD event risk

            -- Lp(a) 350 nmol/L, or 150 mg/dL): 3-fold ASCVD event risk

            -- Lp(a) 250 nmol/L, or 100 mg/dL): 2-fold ASCVD event risk

            -- Lp(a) 125 nmol/L, or 50 mg/dL): 1.4-fold ASCVD event risk

            -- Lp(a) 75-124 nmol/L, or 30-49 mg/dL): 1.2-fold ASCVD event risk

            -- Lp(a) <75 nmol/L, or <30 mg/dL): reference

        -- unfortunately, high Lp(a) levels are really common, affecting about 20% of the global population

    -- the enrollees actually had lots of other cardiovascular risk factors that were suboptimally controlled and overrode the beneficial effect of pelacarsen

        -- when the actual scientific paper is available in 2 months or so, we will get a better sense of this

        -- however, there are lots of cardiovascular risk factors beyond those used in our risk calculators that were likely not included in the risk assessment: https://gmodestmedblogs.blogspot.com/2023/10/update-ascvd-risk-factor-critique.html

            -- for example, any condition (of which there are many) associated with systemic inflammation are risk factors, which includes some medical conditions (rheumatoid arthritis, lupus, etc), chronic infections (hepatidities, syphilis, TB, even well-controlled HIV...), chronic perceived stress (racial discrimination, mental stress, physical stress), depression, air pollution, microplastics, etc. Highlighting our small array of hypertension, lipids, visceral obesity, smoking, and diabetes is in fact tunnel-vision and may distort our assessment of a patient's actual individual risk which may have so many other potential risk factors

    -- we do know that cardiovascular disease is slowly progressive, often beginning in the early teens to late twenties, though this slow development of atherosclerotic coronaries typically does not manifest itself until much later in life, raising a few issues:

            -- we should be providing much earlier intervention than waiting until people have already established cardiovascular disease, especially in light of the attendant major mortality (>50% die from an MI and legions more have disabling morbidity from an MI or stroke)

            -- which means checking lipids and Lp(a) and apoB (apolipoprotein B) levels early in life and being more aggressive in treating when they are high, both by nonpharmacological means (dealing with diet, exercise, overweight), and with meds as needed. for prior blog that discusses apoB see https://gmodestmedblogs.blogspot.com/2024/03/lipoprotein-a-bad-actor.html

            -- and in older people with established cardiovascular disease, it may take more than 7 years as in this study to see a really significant effect since they already have such advanced baseline atherosclerotic disease. 

    -- are there better ways to inhibit Lp(a) than targeting PCSK9 inhibitors?

        -- the Fourier study illuminated the likely role of a PCSK9 inhibitor in lowering Lp(a) levels :

            -- the Fourier study was a really well-designed and conducted study: 25,096 patients were enrolled in this global randomized trial of the PCSK9 inhibitor evolocumab vs placebo: see Lp(a) Fourier study PCSK9 dec events Circ2019 in dropbox, or doi.org/10.1161/CIRCULATIONAHA.118.037184. also prior blog on Fourier with more information: https://gmodestmedblogs.blogspot.com/2025/11/another-pcsk9-study-clinical-benefit.html

                -- these enrollees had a mean age of 63, almost all on statins (only 2/3 on high intensity!!), 80% had prior MI, 20% had prior stroke, 13% had peripheral artery disease

                -- baseline LDL was 93-100 mg/dL (seems inappropriately high, perhaps because so many participants were taking non-high intensity statins...), HDL 45, apoB 85 mg/dL, non-HDL 126 mg/dL, hs-CRP of 3-4

                -- this study found that patients with baseline Lp(a) in the highest quartile (Lp(a) 191-280 nmol/L with median 216) had a higher risk of coronary heart disease death, myocardial infarction, or urgent revascularization, independent of LDL level

                -- by week 48, evolocumab decreased Lp(a) levels by 26.9% (interquartile range of 6.2% to 46.7%); this reduction was greatest in those with the higher baseline Lp(a) level, with P<0.0001, and the absolute reduction in Lp(a) was also highest in those with the highest baseline Lp(a) levels

      -- the benefit of lowering the Lp(a) level at week 12 to a level of 64 nmol/L revealed a pretty steep slope up to about 500 nmol/L

             -- there was a moderate correlation between changes in Lp(a) and LDL levels with the evolocumab

             -- each doubling of the Lp(a) was associated with an 8% increased higher risk of the evaluated cardiovascular events

             -- the overall conclusion was that evolocumab led to a reduced risk of the gamut of cardiovascular outcomes by 23% in those with a baseline Lp(a) greater than the median but only 7% if less than the median, reinforcing that Lp(a) reduction is a pivotal component to cardiovascular risk reduction in those with high Lp(a) levels

                -- ie, the PCSK9 inhibitor evolocumab was clinically effective in lowering the Lp(a), more so in those with high Lp(a) levels

so, the real question comes up: why is a potent PCSK9 medication such as pelacarsen ineffective in preventing adverse cardiovascular events??  After all, of the 8 new PCSK9 inhibitors being developed, it is by far the most potent, with an 80% reduction in Lp(a) levels. so, why is this really potent Lp(a) reducer not associated with decreasing cardiovascular outcomes reflecting cardiovascular morbidity and mortality in this study?

    -- a blog that includes brief summaries of the 8 PCSK9 agents being developed: https://gmodestmedblogs.blogspot.com/2025/11/another-pcsk9-study-clinical-benefit.html , which also argues that the new meds need to have documented clinical efficacy (and criticizes the FDA for approving enlicitide and having it available based solely on lowering Lp(a) levels and some other lipids)

    -- my guess is that the issue is the upside-down approach to medicine in the US. though we have a great focus and ability to help those with established severe cardiovascular diseases, with cardiac transplants, left ventricular assist devices, etc, we are fundamentally lacking in preventive medicine:

        -- we do know that >50% of patients with heart attacks die, largely related to most of them occuring outside of a hospital and particularly outside of an intensive care unit

        -- we do know that the atherosclerotic process begins at a young age, progressively (but slowly) getting worse, and leads to bad cardiovascular outcomes decades later. and those patients who survive the initial cardiovascular event may well have subsequent life-altering morbidities after their heart attack or stroke

        -- any rational/humane health care system would prioritize initial prevention of disease instead of our quite expensive and often inadequate system that focuses on the bad medical events that occur downstream

        -- so, the current acknowledgement of the importance of testing everyone for their Lp(a) levels is a real step forward

            -- but, i personally have very many patients with very high Lp(a) levels, several >500 mg/dL!!! in the probably 100 people i have tested

            -- however, our system does not allow me to treat them: i saw a 40yo with an Lp(a) of >500 mg/dL. her insurance will not cover the quite expensive PCSK9 inhibitors that exist; and if they did cover the evolocumab, she would not have the money to cover the deductible and copays necessary even though she understands the quite likely severe consequences of inadequate prevention. And another of my patients who was very high-risk after having a stroke could not afford these medications because of these financial obstacles, explicitly stating that the health care system was designed to benefit the affluent

    -- my guess from the above is that this study with pelacarsen would have had a very different outcome if the study design were different: a study of asymptomatic younger patients with high Lp(a) levels randomized to the PCSK9 inhibitor at a young age for cardiovascular disease prevention, a particularly important goal since cardiovascular disease is still the number one killer in our society (a number which does not reflect the additional huge potential morbidity of a prior cardiovascular event). PCSK9 inhibitors are certainly necessary for those with high Lp(a) levels given that about 90% of Lp(a) levels are genetically determined. But, of course, we also need to reinforce the importance of a healthy diet, exercise, maintaining a healthy body weight, and decreasing visceral fat. after all, Lp(a) does have some non-genetic associations: https://pmc.ncbi.nlm.nih.gov/articles/PMC7400957/ reviews the 10% of non-genetic factors that affect Lp(a) levels, including lifestyle issues

        -- though this study would take a long time, it would be possible to have periodic non-invasive assessments of differences in the development of increasing atherosclerotic coronary artery disease.

    -- and we cannot do well with a public health system, as we have now, that elevates red meat and tallow (ie animal fat) to the same level as healthy fruit and vegetables!!!  And public health leaders who propagate a pervasive anti-vax ideology that is leads many to not getting measles vaccines and then denying that the 2 recent deaths in unvaccinated kids were not from measles.......,.

geoff

-----------------------------------

If you would like to be on the regular email list for upcoming blogs, please contact me at gmodest@bidmc.harvard.edu

to get access to all of the blogs:  go to http://gmodestmedblogs.blogspot.com/ to see the blogs in reverse chronological order

to get access to the dropbox: https://www.dropbox.com/scl/fo/pcf4vymtpo77f49wd2ljr/ANyJqB40CuBd-DMEr67XDoI?rlkey=5dsalmuhf8ekd0svsgvn93xm1&dl=0

or you can just click on the magnifying glass on top right, then type in a name in the search box and get all the blogs with that name in them