In the late 1990s, RNA interference looked like biology's improbably neat party trick. Cells could use tiny pieces of RNA to intercept a genetic instruction before it became a protein. If the wrong protein caused disease, the implication was irresistible: send in the right interfering RNA and lower the volume at its source. Then came the maddening part. RNA is fragile. The body clears it. Cell membranes do not politely wave it through. A molecule that worked beautifully in a dish could vanish or wander in a patient.
Alnylam Pharmaceuticals was founded in Cambridge in 2002 to solve that delivery problem and turn the mechanism into a drug class. Its founders included Phillip Sharp, David Bartel, Thomas Tuschl, Paul Schimmel and Phillip Zamore, scientists connected to the foundational work around RNA and gene regulation. The name came from Alnilam, the bright middle star in Orion's belt. The altered spelling was distinctive; the navigation metaphor was apt. For years, the destination was visible while the route remained uncertain.
Delivery was the product
The first workable route used lipid nanoparticles, microscopic fat-based packages that protected small interfering RNA, or siRNA, in the bloodstream and helped carry it into liver cells. That system became ONPATTRO, an intravenous treatment for the nerve damage caused by hereditary transthyretin amyloidosis. Its 2018 approval made it the world's first RNAi therapeutic. Alnylam had needed 16 years to reach day one.
The next delivery trick was smaller and more convenient. Alnylam attached siRNA to GalNAc, a sugar that binds receptors found in abundance on liver cells. Think of it as an address label with unusually good postal service. GalNAc conjugates enabled injections under the skin and longer intervals between doses. GIVLAARI followed for acute hepatic porphyria in 2019. OXLUMO arrived for primary hyperoxaluria type 1 in 2020. AMVUTTRA was approved in 2022 for hereditary ATTR polyneuropathy and expanded in 2025 to cardiomyopathy caused by wild-type or hereditary transthyretin amyloidosis.
The breakthrough was not merely knowing which message to silence. It was learning how to deliver the silencer, predictably, to the right address.
Those medicines treat different conditions, but the operating logic repeats. Find a genetically validated protein made in the liver. Design an siRNA sequence against its messenger RNA. Use familiar delivery chemistry. Then improve stability, specificity and dosing. This repeatability is what makes Alnylam a platform company rather than a collection of unrelated drug bets.
Who buys a molecular mute button?
The patient is the reason for the medicine, but a pharmaceutical customer map is crowded. Neurologists, cardiologists, hepatologists, nephrologists and genetic counselors help identify and treat patients. Hospitals and specialty pharmacies handle administration or dispensing. Distributors move product. Insurers and national health systems decide whether access is affordable. In rare genetic disease, diagnosis itself is a bottleneck, so Alnylam also supports genetic testing, counseling and patient navigation programs.
This matters because ATTR amyloidosis can masquerade as ordinary heart failure or neuropathy. Misfolded transthyretin protein accumulates in tissue, and patients may spend years moving between specialists. A medicine cannot help a person whose disease has not been recognized. Alnylam's collaborations with diagnostic technology company Viz.ai, the American Heart Association and health systems are commercial infrastructure in the broad sense: they aim to find the patient before the disease does more damage.
Alnylam sells four products itself, but it does not insist on owning every launch. Novartis commercializes Leqvio, an Alnylam-originated siRNA that lowers LDL cholesterol, and pays royalties. Sanofi commercializes Qfitlia for hemophilia. Regeneron controls cemdisiran as a monotherapy and has submitted it for generalized myasthenia gravis. Roche shares development of zilebesiran, an experimental treatment designed to reduce liver production of angiotensinogen for sustained blood-pressure control.
2025 net product revenue / $2.987 billion
A business with three engines
Alnylam now operates as a specialty drug company, a research platform and a licensing partner. Direct product sales do most of the work: $2.99 billion in 2025, up 81 percent from the prior year. Collaboration revenue contributed $553 million and royalties another $174 million, producing $3.71 billion in total revenue. The company reported its first full year of GAAP profitability after spending much of its history financing chemistry, trials and commercial infrastructure ahead of sales.
Partnerships widen the aperture and share risk. Roche paid $310 million upfront in 2023 for the zilebesiran alliance; the agreement can reach up to $2.8 billion including potential milestones. Alnylam and Roche share U.S. profits and losses, while Roche has rights outside the country and Alnylam can receive royalties there. It is a practical division of labor: Alnylam supplies RNAi depth, while Roche supplies reach in a common disease treated across huge, fragmented markets.
The direct model offers more upside but demands manufacturing, medical affairs, distribution and payer negotiation. The partner model gives away some economics but can place a medicine inside a much larger global machine. Alnylam uses both, choosing where independence matters and where scale matters more.
That choice is backed by an unusually broad internal bench for a focused biotech. At the end of 2025, Alnylam employed about 2,500 people across research, clinical development, medical affairs, commercial operations and manufacturing. It added roughly 270 employees during that year and gives equity compensation to every employee. Its plant in Norton, Massachusetts is part laboratory, part industrial argument: RNA medicines cannot scale on clever sequences alone. In 2025, the company announced a $250 million expansion that will add enzymatic ligation capacity, a manufacturing method intended to assemble siRNA with less waste than older chemical processes. FDA acceptance of Alnylam's siRELIS platform into its Emerging Technology Program offered an early regulatory vote of interest, not a product approval. The culture that emerges from the public record is less startup sprint than scientific relay race. Discovery hands a candidate to development, development to manufacturing, and manufacturing to a commercial organization that must help specialists find a small, dispersed patient population.
In July 2026, Alnylam cut its full-year TTR product revenue forecast to $4.2-$4.5 billion even as AMVUTTRA crossed $1 billion in quarterly sales. Management said pent-up second-line demand had normalized. A powerful platform does not repeal launch curves, competition or reimbursement friction.
Why the field is getting crowded
Alnylam's most durable advantage is evidence. Six approved medicines originated on its platform. It has more than two decades of sequence, safety, delivery and manufacturing data, including the failed experiments that rarely appear in publications. Arrowhead Pharmaceuticals and Silence Therapeutics also develop RNAi medicines, while Ionis Pharmaceuticals uses a related antisense approach to alter RNA. In individual diseases, competitors range from established pills and antibodies to one-time gene-editing programs.
RNAi sits in a useful middle ground. It can produce deep, long-lasting protein reduction without permanently changing DNA. If trouble appears, the effect should diminish as the medicine clears and cells make new RNA. The tradeoff is repeat dosing. It also remains much easier to reach some tissues than others. The liver is Alnylam's home turf because its biology welcomes GalNAc-tagged molecules. Reaching the brain, muscle or other organs is a different delivery problem.
That is why the pipeline matters beyond its individual drug names. Mivelsiran targets amyloid precursor protein in the central nervous system and is in Phase 2 studies for cerebral amyloid angiopathy and Down syndrome-associated Alzheimer's disease. ALN-HTT02 targets Huntington's disease. ALN-6400 is being studied in bleeding disorders. Early programs pursue obesity. An AI collaboration with Inceptive combines foundation models with Alnylam's proprietary RNAi record, hoping to turn 20 years of molecular memory into faster designs.
The move from rare to routine
Rare diseases were a rational proving ground: severe unmet need, relatively clear genetics and specialist communities that could identify eligible patients. Common disease changes the equation. Zilebesiran may offer months of blood-pressure reduction from one injection, potentially addressing the very ordinary problem that people forget daily pills. But it would enter a world full of cheap generics, entrenched guidelines and enormous safety expectations. Convenience must translate into outcomes and acceptable economics.
Alnylam's market position is therefore unusual. It is no longer an early biotech with a promising mechanism, yet it is not a diversified pharmaceutical conglomerate. With roughly 2,500 employees at the end of 2025, internal manufacturing in Massachusetts, four directly sold medicines and major-pharma alliances, it occupies the narrow space between platform pioneer and scaled specialist.
The company has already answered the question that shadowed RNAi for a decade: can gene silencing become medicine? The answer sits in infusion rooms, specialty pharmacies and quarterly revenue statements. Its next question is less romantic and more consequential. Can the same molecular machinery produce repeatable products for diseases that affect millions, while keeping access, safety and cost in balance?
Alnylam's history suggests an answer will not arrive as a single eureka moment. It will arrive as chemistry, delivery, trials, factories, diagnosis and reimbursement lining up. The star gave the company its name. The unglamorous work of navigation built the business.