Amgen (NASDAQ: AMGN) was down about 10% at $394.38 on Tuesday afternoon, per Yahoo Finance. On the $213.32bn market cap StockAnalysis reports, that's roughly $23bn of value gone in a session. Amgen didn't report a failure or cut guidance — it announced a Phase 3 win the same morning. The drug that failed belongs to Novartis.

What happened

Novartis said its Phase 3 Lp(a)HORIZON trial of pelacarsen missed its primary endpoint. It ran 8,323 patients with established cardiovascular disease and elevated lipoprotein(a) — Lp(a) of 70 mg/dL or higher, with a higher-risk group at 90 mg/dL and above. The drug lowered Lp(a). It did not beat placebo on the endpoint that decides its future: cardiovascular death, non-fatal heart attack, non-fatal stroke, and urgent coronary revascularisation. Novartis's chief medical officer, Shreeram Aradhye: "These are not the results we hoped for."

The topline landed on 4 September; with US markets shut Monday for Labor Day, Tuesday was the first full session to trade it. Novartis (NVS) fell about 14% to $137.63, Eli Lilly (LLY) about 2%. The S&P 500 tracker was down 0.37% — this repriced one idea, not the market.

Why Amgen got hit

The bit worth learning: Lp(a) is not LDL. LDL is the "bad cholesterol" statins and diet move. Lp(a) is a separate particle, set almost entirely by the genes you're born with — food and exercise barely touch it, and statins don't fix it. It's been tied to heart-attack risk for years, and no approved drug targets it. So the industry built medicines that switch off Lp(a) production. They work spectacularly on the number — which was always the gamble, because moving the marker isn't the same as saving lives. Pelacarsen was the first in the class to test that in a real outcomes trial, and the assumption didn't hold. BMO Capital Markets called it "a meaningful setback for dedicated Lp(a)-lowering therapies."

Amgen's version is olpasiran, which cut Lp(a) by more than 97% in mid-stage work. Its Phase 3 OCEAN(a)-Outcomes trial enrolled 7,297 patients and doesn't read out until around 2028. None of Amgen's own data changed on Tuesday. The bar did: the market will no longer pay for crushing a biomarker, and olpasiran now has to prove it prevents heart attacks.

The business — what today did NOT touch

Amgen did $10.1bn of revenue in Q2 2026, up 10%. Olpasiran contributes none of it.

• Repatha $953m, +37% — the LDL drug. Different target, outcomes already proven, untouched by this.
• EVENITY $714m (+38%), TEZSPIRE $486m (+42%) — bone and severe asthma.
• IMDELLTRA $288m, +115% — the drug that had a good Tuesday.
• Drags: Prolia −32%, Otezla −21%, both losing exclusivity.

The news the tape ignored: Amgen's Phase 3 DeLLphi-305 study of IMDELLTRA with AstraZeneca's IMFINZI hit its primary overall-survival endpoint in 563 patients with first-line extensive-stage small cell lung cancer — the first BiTE therapy to show a survival benefit that early in treatment.

Is it expensive?

Even after the fall, Amgen trades at a trailing P/E near 27 and about 18.6x forward on a $213bn cap, per StockAnalysis — not a bargain multiple for 10% revenue growth. The obvious peer: Novartis at roughly 23x trailing and 16.8x forward, on a larger $295bn cap and $56.7bn of revenue. So even down 10%, Amgen is still the pricier of the two. You're paying for the pipeline, and Tuesday was the market marking one piece of it down. Amgen yields about 2.6% and is still up roughly 38% over 52 weeks.

Who else this touches

• Ionis (IONS) — discovered pelacarsen and licensed it to Novartis. Down about 13% Tuesday.
• Eli Lilly — lepodisiran cuts Lp(a) nearly 94%, but its ACCLAIM-Lp(a) outcomes data isn't due until about 2029.
• Merck — paid around $2bn to Jiangsu Hengrui for oral Lp(a) drug HRS-5346. Same thesis, same new doubt.
• Arrowhead and Royalty Pharma — Arrowhead invented olpasiran and sold the royalty to Royalty Pharma for $250m plus milestones, so 2028 is their number too.

The one thing that flips it

Full Lp(a)HORIZON data goes to a medical congress. If it shows the drug worked in the sickest patients — the 90 mg/dL-plus group — the story stops being "the Lp(a) idea is wrong" and becomes "pelacarsen didn't lower it enough." Olpasiran clears more than 97%, so that reading would make Tuesday's 10% look like a fast trigger finger. If that group looks like everyone else, the whole class has a problem, and 2028 is a long time to wait to find out.

As of 3:30 PM ET, 8 September 2026. Sources: Yahoo Finance, Novartis media release, HCPLive, BioSpace, Amgen Q2 2026 results, Amgen DeLLphi-305 release, StockAnalysis. All prices, levels and percentages are as reported by these sources, not real-time exchange data. For discussion and education only — not investment advice. Verify before acting.