VSA006

VSA006 is an experimental siRNA from Visirna Therapeutics that knocks down HSD17B13, a liver enzyme whose natural genetic absence appears to protect people from cirrhosis and hepatocellular carcinoma [1]. The Phase 2 trial (NCT06322628) is testing the drug in 48 Chinese patients with NASH - the fatty liver disease now called MASLD in updated nomenclature - with a biopsy-confirmed fibrosis improvement endpoint [2]. Visirna is the Hong Kong-based joint venture between Arrowhead Pharmaceuticals and Vivo Capital, formed in 2021 to bring Arrowhead's RNAi assets to Greater China; VSA006 is the Greater China arm of the ARO-HSD program. In November 2021 Arrowhead licensed ARO-HSD to GSK for all territories except Greater China for $120M upfront plus up to $780M in milestones, explicitly retaining Greater China rights through the Visirna JV - that carve-out is the entire reason VSA006 exists as a separately developed asset [3]. The commercial backdrop is tougher than two years ago: Madrigal's resmetirom (Rezdiffra) cleared FDA in March 2024 as the first NASH-specific therapy [4], and Novo Nordisk's semaglutide hit primary endpoints in the Phase 3 ESSENCE trial with 62.9% MASH resolution vs 34.3% placebo, published in NEJM April 2025 [9]. The closest mechanistic analog - Regeneron/Alnylam's ALN-HSD (rapirosiran) - is not dead: the Phase 1 readout showed robust HSD17B13 knockdown plus numerically lower biopsy-derived NAFLD Activity Score, and the Phase 2 NASHGEN-2 trial remains active with a September 2027 completion target [5]. The genetic validation for the target is among the strongest in liver disease.

Novel siRNA, never approved anywhere. The drug is in Phase 2 in mainland China, listed as ACTIVE_NOT_RECRUITING on ClinicalTrials.gov with enrollment closed at 48 patients [2]. The Phase 2 was supported by a prior Phase 1 in healthy volunteers and NASH/suspected NASH patients, per the public ClinicalTrials.gov record for NCT06322628 [2]; specific Phase 1 knockdown and safety numbers for VSA006 itself have not been published in a peer-reviewed venue. No FDA designations apply - this is a National Medical Products Administration (NMPA, China's FDA equivalent) trial, and Visirna has not announced US IND plans for the molecule. The trial is small by Western Phase 2 standards but typical for Chinese proof-of-concept fibrosis studies. Readout timing has not been disclosed publicly; biopsy-based fibrosis trials usually require 48-52 weeks of dosing plus paired biopsy logistics, so a 2027 data window is plausible if dosing wrapped in 2025. The parent compound on the GSK side (GSK4532990, formerly ARO-HSD) is the molecule that matters more for the global thesis. GSK's Phase 2b HORIZON trial (NCT05583344) is testing GSK4532990 in adults with NASH and advanced fibrosis, enrollment of 284 patients is complete (status ACTIVE_NOT_RECRUITING), treatment duration is up to 52 weeks within a 76-week study window, putting a likely readout in late 2026 to 2027 - 12-18 months ahead of VSA006 [10]. A positive or negative HORIZON readout would heavily reshape the VSA006 outlook regardless of the Chinese trial's own results [3,10]. A clearly negative HORIZON readout would essentially kill the program in any market.

HSD17B13 is a small protein that sits on the surface of fat droplets inside liver cells. Its day job involves processing lipid and steroid molecules including estradiol and retinol, though the exact contribution to liver injury is not fully worked out [6]. What is clear is the genetics: in a 2018 Regeneron-led paper in NEJM, people who naturally inherit a broken copy of HSD17B13 had meaningfully lower rates of cirrhosis and hepatocellular carcinoma, including in patients with alcoholic and non-alcoholic fatty liver disease [1]. About one in four people of European ancestry carry at least one protective variant. The genetic story is strong enough that the field has treated HSD17B13 the way it treated PCSK9 a decade ago - a target where loss-of-function carriers tell you that turning the protein off should help patients. VSA006 is a short piece of RNA engineered to destroy the messenger RNA that encodes HSD17B13 before liver cells can make the protein. The drug is delivered via GalNAc conjugation, a sugar tag that liver hepatocytes grab through their asialoglycoprotein receptors - the same delivery technology Alnylam uses for its approved siRNAs. One dose typically silences the target for several months. Two other HSD17B13 siRNAs have already cleared Phase 1 in humans: ALN-HSD showed robust target knockdown plus a numerical drop in NAFLD Activity Score over six months [5], and GSK4532990 (formerly ARO-HSD) showed >90% mRNA knockdown with reductions in ALT and AST in its Phase 1/2a [3]. So the mechanism is clinically derisked at the pharmacology level - what remains unproven is whether silencing for a year or two reverses already-injured livers the way decades-long genetic absence prevents progression.

NCT06322628 is a Phase 2 study in Chinese NASH patients, n=48, status ACTIVE_NOT_RECRUITING [2]. The primary endpoint is the percentage of patients achieving at least one stage of histological fibrosis improvement with no worsening of NASH - this matches the FDA's accepted biopsy framework for NASH accelerated approval and gives the readout regulatory weight beyond China's NMPA (National Medical Products Administration). Forty-eight patients is small. For comparison, Madrigal's MAESTRO-NASH (the trial that supported resmetirom approval) randomized over 950 patients [7], and GSK's HORIZON Phase 2b enrolled 284 [10]. A 48-patient design is fine for showing a signal but cannot be the registration trial in any major market. Sponsor is Visirna Therapeutics HK; the trial's full specifics on comparator arm structure, dosing regimen, biopsy timing, and secondary endpoints (ALT, liver fat by MRI-PDFF, fibrosis biomarkers) are not all publicly disclosed. ALT (alanine aminotransferase, a blood marker of liver cell damage - easier to measure than a biopsy but considered a weaker efficacy signal for approval purposes) is the usual companion readout in NASH/MASH siRNA programs. The trial is reasonable for proof-of-concept but the binary win condition is essentially: did the histology move in roughly the right direction at this small N?

The model gives this drug a 4% chance of eventually being approved. That number starts from the historical approval rate for Phase 2 drugs in this area, around 23%, then adjusts based on ten specific facts about the trial and its sponsor. It gets a boost from having more secondary endpoints than usual, but is pulled down by unusually heavy blinding, the sponsor's thin or weak approval record, and earlier-phase results that were weak or limited. The remaining facts were close to average for this stage, so they didn't shift the estimate much.

The efficacy risk remains the dominant one, but it is more nuanced than the simple 'analog already failed' framing some commentary has implied. ALN-HSD's Phase 1 actually showed both target knockdown and numerically lower NAFLD Activity Score (NAS, a composite of steatosis, ballooning, and lobular inflammation) versus placebo over six months in 20 NASH patients [5]. The genetic case predicts long-term protection from cirrhosis, not necessarily reversal of established disease, and VSA006 is being tested in already-injured livers - a different question than the one the genetics actually answers. ALT (alanine aminotransferase, a blood marker of liver cell damage - easier to measure than a biopsy but considered a weaker efficacy signal for approval purposes) reductions alone will not be sufficient. The histology endpoint is what matters for the regulatory case and the investment case. Safety risk is the smallest concern. GalNAc-conjugated siRNAs have been well-tolerated across Alnylam's approved drugs and the broader class, and HSD17B13 LOF carriers do not show major safety issues, so on-target toxicity should be low. Watch for changes in serum lipids and any sex-hormone-related signals given the enzyme's in vitro activity on estradiol and retinol metabolism [6]. Execution risk comes from the small trial size. Forty-eight patients in a single-country study with limited disclosed design specifics leaves the data harder to interpret [2]. NASH histology reading has known inter-reader variability; central reading is standard but small N magnifies any noise. Commercial risk is real even with a successful trial. Resmetirom is launched and ramping in the US [4], and semaglutide produced a positive Phase 3 readout in ESSENCE with 62.9% MASH resolution vs 34.3% placebo [9]. Even if VSA006 works, Chinese MASH pricing, reimbursement, and patient flow differ markedly from the US, and Visirna has no commercial infrastructure. The realistic path to revenue runs through NMPA approval plus either Visirna building a Chinese commercial arm or a multinational partnership.

Watch, do not bet. The science is real - HSD17B13 is one of the most genetically validated targets in liver disease - and the closest analog (ALN-HSD/rapirosiran) is alive and produced a Phase 1 histology signal, not the quiet failure earlier secondary-source commentary suggested [5]. VSA006's Phase 2 readout is informative only if it produces something convincingly better than ALN-HSD's Phase 1 directional move, on histology and not just on ALT. The smarter trigger to watch is GSK's Phase 2b HORIZON trial of the parent compound (GSK4532990, formerly ARO-HSD): same chemistry, same target, 284 patients, ACTIVE_NOT_RECRUITING, NCT05583344, likely readout late 2026 to 2027 [10]. If GSK announces a positive Phase 2b fibrosis signal at AASLD or a major liver meeting, that revalidates the entire HSD17B13 thesis and Visirna's Chinese asset goes from noise to a meaningful regional play. If HORIZON reads out flat or negative, VSA006 is effectively dead regardless of its own data. Commercial geography matters. China has roughly 290-340 million adults with NAFLD/MASLD (pooled prevalence ~29-31% of Chinese adults across recent surveys) [11], and the MASH subset is large and growing - a meaningful regional market if Visirna can clear NMPA registration. The realistic gate is the NMPA-aligned key trial design (the current 48-patient study is not it) plus Visirna building or partnering for commercial infrastructure. Patent-wise, Arrowhead holds the foundational HSD17B13 siRNA IP that flows to Visirna under the JV structure, and GalNAc delivery composition-of-matter remains protected through standard 20-year terms. Visirna itself is a vehicle, not a story; the JV exists to monetize Arrowhead's pipeline in Greater China and does not yet have a clinical track record of its own. Check back at AASLD 2026 or the next GSK R&D update for the HORIZON signal, which leads VSA006 by 12-18 months.