ALN-HSD

ALN-HSD (proposed INN rapirosiran) is a small interfering RNA (siRNA, a short RNA that tells liver cells to destroy a specific gene's messenger RNA) co-developed by Regeneron and Alnylam for metabolic dysfunction-associated steatohepatitis (MASH, the fatty liver disease that progresses to scarring and sometimes cancer). The target is HSD17B13, a liver enzyme where a common natural loss-of-function variant cuts the risk of chronic liver disease by roughly 30% [1]. The Phase 2 trial NCT05519475 is enrolling genetically-selected adults at elevated liver disease risk, using quantitative liver fibrosis (qFibrosis) on biopsy as the primary endpoint [2]. The bet: pharmacologically copy what the protective gene variant does, in patients chosen because their DNA says they need it most. This is a precision-medicine approach landing in a crowded MASH field where Madrigal's resmetirom (Rezdiffra) reached FDA approval in 2024 and where GLP-1 drugs like semaglutide and tirzepatide are showing large weight-and-liver benefits as a side effect of metabolic reprogramming [3]. The Regeneron-Alnylam collaboration is structured as a 50-50 co-development and co-commercialization deal for HSD17B13 [4], so both companies share the upside. The MASH addressable population in the US alone is estimated at roughly 6 to 9 million adults with biopsy-confirmable disease, with a much smaller fraction at the advanced fibrosis stages where therapy is most economically defensible.

ALN-HSD is a novel siRNA, never approved anywhere, sitting in Phase 2 [2]. The trial is recruiting per ClinicalTrials.gov with an enrollment target that has been expanded from the original 120 to approximately 300 patients, and the primary completion date has been pushed from December 2026 to September 2027 [2]. No FDA designations (breakthrough therapy, fast track, orphan drug, RMAT) have been publicly disclosed by the sponsors. The molecule originated at Alnylam Pharmaceuticals, the GalNAc-siRNA pioneer, and moved into a March 2018 Regeneron-Alnylam discovery collaboration specifically for HSD17B13. Per the public terms, Regeneron and Alnylam share development and commercialization on a 50-50 basis, with cost and profit splits aligned to that structure [4]. The exact royalty mechanics, milestone schedule, and which party books revenue post-approval have not been spelled out in public filings. Regeneron leads the MASH program as trial sponsor. Phase 1 data was disclosed in September 2022 [8]. In 44 healthy volunteers, ALN-HSD showed an encouraging safety and tolerability profile, with injection-site reactions the most common treatment-emergent adverse event and no treatment-related serious adverse events. In the first two NASH cohorts (200 mg and 400 mg quarterly subcutaneous dosing, N=20 active and N=4 placebo, six-month exposure) the drug produced robust HSD17B13 knockdown together with numerically lower liver enzymes and a numerically lower NAFLD Activity Score versus placebo. The study was not powered for efficacy, but the result was clean enough to support the Phase 2 design. No clinical hold history is apparent, and no public safety signals have been disclosed in Regeneron's earnings calls or 10-K filings through early 2026 [5].

HSD17B13 is a protein that lives on lipid droplets, the small fat-storage bubbles inside liver cells [6]. It oxidizes various lipid substrates including retinol (vitamin A) and 17-beta-estradiol in lab assays. The big finding came in 2018: a sequencing study of about 46,000 people found that carriers of a specific HSD17B13 variant (rs72613567, an extra T that produces a shortened, non-functional protein) had roughly 30% lower risk of cirrhosis from both alcoholic and non-alcoholic liver disease, plus lower risk of hepatocellular carcinoma [1]. Heterozygous carriers got partial protection. Homozygotes got more. This is human genetic validation, the cleanest kind of biological evidence, not a mouse model with hopeful translation to people. The mechanistic link from the enzyme to fibrosis is the open scientific question. The leading hypothesis is that HSD17B13's enzymatic activity on lipid droplets in inflamed hepatocytes generates bioactive lipid metabolites, or alters lipid droplet composition and turnover, in a way that amplifies inflammatory and fibrogenic signaling from hepatocytes to neighboring stellate cells. Loss-of-function removes this amplifier and dials down the downstream injury response. The precise pathway is still under active investigation, and ALN-HSD is in essence betting on the human genetics first and pathway elucidation second. ALN-HSD copies what the variant does pharmacologically. The siRNA silences HSD17B13 messenger RNA before it can be translated into the protein, so hepatocytes make less of the enzyme. The GalNAc (N-acetylgalactosamine) sugar conjugate is the targeting trick: hepatocytes express the ASGPR receptor that grabs GalNAc-bearing molecules and drags them inside the cell, so the drug concentrates in liver and largely spares the rest of the body. This is the same delivery chemistry behind Alnylam's approved RNAi drugs like patisiran's later GalNAc-conjugated successor vutrisiran, and givosiran, as well as inclisiran (originally Alnylam-discovered, licensed to The Medicines Company and now commercialized by Novartis), which together give the GalNAc platform real safety priors.

NCT05519475 is a Phase 2 trial run by Regeneron, recruiting adults at genetic risk for MASH progression with an updated enrollment target of approximately 300 patients and a primary completion date of September 2027 [2]. The primary endpoint is change in quantitative liver fibrosis (qFibrosis), an AI-driven histology score read off liver biopsy slides that measures the structural extent of scarring on a continuous scale. The choice is interesting and risky. Traditional MASH endpoints used by FDA for accelerated approval are categorical: MASH resolution without worsening of fibrosis, or fibrosis improvement by at least one stage on the NASH CRN scoring system. qFibrosis is more sensitive to subtle change but does not yet have an FDA-blessed regulatory link to clinical outcomes. For a Phase 2 proof-of-mechanism trial this is fine, even smart. For Phase 3, the program will need to either bridge qFibrosis to a recognized endpoint or convince regulators that the quantitative score predicts cirrhosis-related events. Patient selection is the design's strongest feature. Enrolling on genetic risk concentrates the trial population in people whose biology says the drug should work, which is the right way to test a precision RNAi. The Phase 1 NASH cohort already showed numerical signals on liver enzymes and NAFLD Activity Score at the 200 mg and 400 mg quarterly doses, which informed Phase 2 dose selection [8]. The expanded sample size of roughly 300 (up from the original 120) supports a biomarker readout with more statistical headroom and reflects the kind of size needed if any meaningful subgroup analyses are planned. A placebo-controlled design is the standard assumption; no active comparator arm is publicly described.

Our model estimates a 7% chance this drug is eventually approved. It starts from the historical base rate for Phase 2 drugs in this area (about 23%), then adjusts using ten facts about the trial and sponsor. What moves the number most: it is helped by more secondary endpoints than usual and the sponsor's strong record of getting drugs approved; it is held back by heavier-than-usual blinding and weak or limited earlier-phase results. The other facts land near average for this stage, so they leave the estimate roughly where the base rate put it.

Efficacy risk is real. MASH trials have killed dozens of programs over the past decade, including obeticholic acid, selonsertib, cenicriviroc, and elafibranor in its first NASH attempt. qFibrosis as the primary endpoint is novel and not yet FDA-validated for accelerated approval, so a positive Phase 2 may still leave Phase 3 endpoint design uncertain. A statistically significant qFibrosis change that does not map to NASH CRN fibrosis stage improvement would put the program in regulatory limbo. Safety risk is moderate. HSD17B13 oxidizes retinol and 17-beta-estradiol in vitro [6], so chronic hepatocyte-restricted knockdown could in principle disturb vitamin A or estrogen handling in liver, though Phase 1 surfaced no such signal [8] and the genetic loss-of-function carriers in the population show no obvious vitamin A or estrogen phenotype. GalNAc-siRNA platforms have accumulated a generally clean safety record across Alnylam's approved drugs. Revusiran was discontinued years ago for a mortality signal in advanced cardiac ATTR patients, a finding attributed to the disease population and the target biology rather than the GalNAc delivery platform itself, which has since supported multiple successful approvals. Execution risk: biopsy-based screening is slow to enroll. A trial registered in 2022 that is still recruiting in 2026, with the primary completion date now pushed to September 2027 [2], suggests pace below initial plan, although that is not unusual for MASH and the enrollment target was also expanded. Commercial risk is the underrated one. Resmetirom (Rezdiffra) reached MASH approval in 2024 and is the new bar [3]. Semaglutide and tirzepatide, already reimbursed, are showing meaningful MASH benefits as a consequence of weight loss. The US MASH addressable population is large, estimated at roughly 6 to 9 million adults with biopsy-confirmable disease and 1.5 to 3 million at advanced fibrosis stages [9], but payer willingness to fund a genetically-restricted siRNA priced like other GalNAc drugs (north of $300K per year) will demand either dramatic efficacy or a clear story about preventing transplant and liver cancer.

Worth watching. The HSD17B13 story is one of the strongest human-genetic validations in metabolic disease, in the same lineage as PCSK9 for LDL cholesterol. That alone justifies tracking the program. The Phase 1 data already shows the drug hits the target hard and is well tolerated [8], which removes the most basic risk of biological failure and leaves the question squarely on whether HSD17B13 knockdown translates into structural liver improvement in symptomatic patients. The signal to watch for at the Phase 2 readout: qFibrosis improvement plus ALT and AST normalization plus a clean safety profile. If that hits, Regeneron will register a Phase 3 and ALN-HSD becomes a real asset rather than a curiosity. The signal that says ignore it: a flat qFibrosis result, improvement that does not track with liver enzymes, or any vitamin A or estradiol-related adverse events. Any of those would suggest the protective biology of the genetic variant does not translate cleanly to pharmacological knockdown in already-sick patients. Next concrete catalysts: Q3 2026 Regeneron earnings (typically late October) for any commentary on enrollment pace or interim data, and any further changes to the NCT05519475 primary completion date on ClinicalTrials.gov. A date extension would signal continued enrollment lag; a date hold or acceleration would be a signal on pace. Regeneron reported about $14.3B in annualized revenue per the 2024 10-K filed February 2025 [5], so the company can fund this program for years without strain. Because the Regeneron-Alnylam structure is 50-50 [4], any positive readout flows to both companies' valuations, and Alnylam's smaller market cap makes it the more used way to express a view on this specific asset. The strategic question is whether a genetics-selected MASH siRNA fits alongside Rezdiffra and the GLP-1s as a third pillar of MASH care, or whether it ends up as a niche product for a narrow genetic subset. The Phase 2 readout will start to answer that.