Liver Disease 2026: Rezdiffra $364M, 2 PBC Catalysts, Mirum

Everyone is chasing GLP-1s. Meanwhile the liver shelf is quietly producing the numbers that matter: $364 million in a single quarter, a November PDUFA, and a Phase 3 field that just narrowed from two players to one. If you want a disease area where the first mover is actually selling rather than promising, hepatology is it.

MASH: Rezdiffra Is a $1.3 Billion Franchise Hiding in Plain Sight

Madrigal’s Rezdiffra (resmetirom) is the first and still only FDA-approved drug for MASH with moderate to advanced fibrosis, and it is now a real commercial asset. Second-quarter 2026 net sales came in at $364.3 million, up 71% year over year from $212.8 million, with nearly $1.3 billion in trailing twelve-month sales. More than 49,000 patients were on the drug as of June 30, and Madrigal says it has since passed 50,000. That is the boring, compounding ramp that the skeptics said a metabolic liver drug could never generate without an obesity angle.

The stock does not reward it. MDGL closed at $512.16 with an $11.8 billion market cap, and it fell 9.83% on July 30, the day it reported these numbers. The selloff is the tell: a $364 million quarter at 71% growth was not enough for a company the market already values at roughly nine times trailing revenue. The market is pricing Rezdiffra as a $1B-plus drug, which means any deceleration in the growth rate, even from 127% in Q1 to 71% in Q2, gets punished.

The clinical foundation is real. The MAESTRO-NASH trial that supported the 2024 accelerated approval showed MASH resolution in 26% of resmetirom patients versus 14% on placebo (p < 0.001) and fibrosis improvement in 24% versus 14% (p = 0.004). The confirmatory MAESTRO-NASH-OUTCOMES trial is the slow-burn catalyst that converts accelerated approval into full approval. That is the risk to watch: if the outcomes data disappoints, the franchise thesis unwinds.

The competitive picture is where it gets interesting. Akero Therapeutics (AKRO, $54.65, $4.5 billion market cap) is running efruxifermin through the Phase 3 SYNCHRONY program in F2/F3 fibrosis and compensated cirrhosis. And the FGF21 field just got a stamp of approval from big pharma: Roche agreed to buy 89bio for pegozafermin in a deal worth up to $3.5 billion, announced in September 2025, with pegozafermin now advancing through the Phase 3 ENLIGHTEN program. The read-through is that Madrigal at $11.8 billion sits above two pre-revenue challengers valued at $4.5 billion and $3.5 billion respectively. Rezdiffra’s first-mover lead is real, but it is measured in years, not decades. For the full MASH approval story, see Madrigal’s Rezdiffra primer.

PBC: Two Catalysts, Two Different Bets

Primary biliary cholangitis has two near-term events, and they are not the same trade.

GSK’s Lynavoy (linerixibat) won a positive CHMP opinion on July 23 for cholestatic pruritus in PBC, the first itch drug ever recommended for the disease in Europe. The FDA actually beat the EU to it, approving Lynavoy on March 19 based on the GLISTEN Phase 3 trial, which showed a 0.71-point itch score improvement over placebo by Week 2 (p < 0.001) that held through 24 weeks, when 56% of patients hit a clinically meaningful ≥3-point itch reduction versus 43% on placebo. GSK is already selling the asset to Alfasigma, which tells you how strategic a niche rare-disease drug is to a $104 billion company: it is a portfolio footnote. GSK closed at $52.07 with a $104 billion market cap, and Lynavoy does not move that number in either direction. The real story is that the IBAT inhibitor class now has a dedicated pruritus approval, and the full CHMP breakdown is here.

Zydus saroglitazar is the higher-conviction regulatory event and the harder one to trade. The FDA granted Priority Review with a November 27 PDUFA, backed by Phase 3 data showing a biochemical response in 56.7% of treated patients versus 9.8% on placebo, a 46.9 percentage point gap. Saroglitazar is a dual PPAR alpha/gamma agonist, mechanistically distinct from Intercept’s Ocaliva, the only second-line PBC drug, which carries a boxed warning in advanced cirrhosis. The catch is access: Zydus trades on India’s NSE as ZYDU.NS, and there is no US-listed ticker. The catalyst is real; the investable pure play does not exist in US markets. The full Zydus setup is here.

Biliary Atresia: Bylvay Failed, and the Field Is Mirum’s to Lose

The liver event of the summer was a failure. On July 24, Ipsen announced that Bylvay (odevixibat) missed the primary endpoint in the Phase 3 BOLD trial, the first global Phase 3 ever run in biliary atresia. Native liver survival at 104 weeks did not separate from placebo across 254 patients in 19 countries. Bylvay is an IBAT inhibitor, the same mechanism as Lynavoy, and its failure was not a mechanism failure. It was an endpoint failure: a two-year survival endpoint in a pediatric disease where roughly half of patients need a transplant anyway. The BOLD miss is covered here.

That failure handed the field to Mirum. Mirum’s EXPAND trial (NCT06553768) is testing LIVMARLI (maralixibat), its own approved IBAT inhibitor, in cholestatic pruritus from biliary atresia and other rare cholestatic liver diseases. The design is different: it measures pruritus severity at Week 20 instead of survival at Week 104, it uses a symptom endpoint the FDA has already accepted for LIVMARLI in Alagille syndrome and PFIC, and it reads out in Q4 2026. Mirum closed at $98.08 with a $5.98 billion market cap after reporting $176 million in Q2 product sales, up 37.5%, and raising full-year guidance to $680 to 700 million.

The setup is binary and asymmetric, at roughly 8.5 times forward revenue. If EXPAND hits, Mirum files an sNDA in early 2027 and becomes the first company with a drug for a disease that currently has zero approved therapies. If it misses, the stock finds support around $80 because the LIVMARLI base business still supports a mid-$5 billion valuation. The full EXPAND primer, with trial design and thresholds, is here, and the orphan pricing math that makes a first-in-class biliary atresia drug valuable is here.

The Structural Point: Two Pricing Regimes in One Shelf

Hepatology is the one place where orphan-drug pricing meets a mass-market total addressable market. MASH is a roughly 5 million patient American market priced like a chronic metabolic drug. PBC and biliary atresia are rare diseases priced like orphans. The two regimes produce two very different investment profiles: Madrigal is a revenue-multiple story, and Mirum is a binary catalyst story. You cannot own them the same way. This is the same distinction that runs through the cardiometabolic catalyst roundup, where MASH appears as just one of five catalysts rather than the whole shelf.

Risks

For Madrigal, the specific risk is the MAESTRO-NASH-OUTCOMES confirmatory readout, plus the Roche and Akero programs that could compress Rezdiffra’s first-mover window. For Mirum, the volixibat NDA delay to H1 2027 is the overhang that caps the multiple even before EXPAND reads out. For Zydus, the risk is commercial, not regulatory: convincing US hepatologists to switch from a known FXR agonist to a dual PPAR agonist from a company with no US commercial footprint.

Verdict

Ranked by risk-reward:

1. Mirum. The only name where a single Q4 catalyst can move the stock 15% up or 15% down, with an empty competitive field behind it. Do not size a full position; a 1 to 2 percent allocation ahead of EXPAND is the right framing.
2. Madrigal. The best business in the group, but at roughly nine times trailing revenue the market has already paid for the ramp. The July 30 selloff on good numbers tells you the multiple, not the fundamentals, is the risk. Hold, do not chase.
3. Zydus. The highest-conviction PDUFA, a 46.9 point delta, with no US-listed vehicle. Watch it, but if you cannot access the NSE, you are a spectator.
4. GSK. Lynavoy is a rounding error for a $104 billion company, and it is being sold to Alfasigma anyway. Skip unless you already own GSK for other reasons.

The liver shelf is where the numbers are, but the returns are concentrated in one binary event and one fully-priced ramp. Pick your spot.