Episode 158 - October 10, 2025

Summary

• The XBI trading around 104–105 is a 2021 flashback — not 2022, as Brian Skorney guessed. Tess Cameron cited a LifeSci Advisors tally of roughly $81B of year-to-date M&A through September, with many deals over $1B — above full-year 2024, below 2023’s high — and about 30 total deals, already near a typical full-year count with two months left.
• Sam Fazeli’s core thesis for the sector: “pharma companies need pipeline. They hit a brick wall.” Pfizer’s own obesity asset did not work out straight away, so it bought Metsera; Akero’s FGF21 MASH program looked compelling to Novo Nordisk, which agreed to pay up to $5.2B. His bet: within six to twelve months, forgotten FGF21 or adjacent assets — possibly from China — will emerge.
• Akero’s efruxifermin appears strongest on fibrosis efficacy but has higher GI side effects, which matters because the endgame may involve combination with semaglutide. Sam cited 40% fibrosis improvement at 24 weeks rising to 75% at 96 weeks, 39% improvement in F4 patients in SYMMETRY, and MASH resolution three times placebo. On a placebo-adjusted basis, it is “not that much better” than 89bio’s pegozafermin; dosing, tolerability and bone mineral density remain open questions.
• BMS’s $1.5 million Orbital deal highlights the push toward in vivo CAR-T. Yaron Werber called an in vivo, bedside, LNP-driven modality the next CAR-T innovation; Sam noted that BMS already has what he views as the leading anti-CD19 CAR-T, Breyanzi, and that Orbital’s approach could make CAR-T more practical in immune-mediated disease. The traditional process is particularly difficult for broad autoimmune populations, though Sam noted an exception for extreme end-stage disease.
• Ionis is the re-rating story of the year — its stock has more than doubled — and Yaron’s Peter Lynch line captures it: “I’m an overnight success 25 years in the making.” Innovation Day guided to profitability in 2028 and more than $5B in peak sales ($3B from the pipeline and $2B from collaborators), alongside Tryngolza’s 72% triglyceride and 85% acute pancreatitis reductions in sHTG and a surprise in-house siRNA APOC3 program with 90% reduction and pharmacokinetics supporting six-month or annual dosing.
• Lexeo’s regulatory update suggests CBER remains flexible under its new leadership. Brian Skorney said the FDA will retain LVMI as a co-primary endpoint with frataxin expression, apparently may allow Lexeo to pool Phase 1/2 data with the pivotal study planned for the first half of 2026, and indicated that the LVMI analysis window may be under 12 months — patients are crossing the 10% threshold by six months.
• The panel split openly on Peter Marks joining Eli Lilly. Sam said the move was not helpful for the cause, while an unidentified panelist took the other side, arguing that Marks has an “unbelievable skill set” that can benefit patients in industry. Tess said disclosure and recusal can manage conflicts, and that it is worse if regulators cannot consult experts who have industry experience.

Deep dive

1. The XBI at 104 is a 2021 flashback — and the M&A tape is the reason

• Tess Cameron’s opening pop quiz — when did the show last convene with the XBI around 105 — drew a guess of 2022 from Brian Skorney; Tess corrected it to 2021. Sam Fazeli had Bloomberg in front of him. By airtime the index had slipped just under 105, so Tess called it “over 104.” Her framing was that the renewed interest is being “driven by a lot of fundamentally good news.”
• The supporting data point, from a LifeSci Advisors analysis she cited and predating the week’s announcements: roughly $81B of year-to-date M&A through September, with many deals over $1B — a real increase over full-year 2024, though not at 2023’s highs. On deal count, about 30 year to date already matches typical full-year totals from many years since 2020, with two months to go.
• Sam Fazeli wanted the mechanism on record before any single name: “pharma needs pipeline… they hit a brick wall. There are assets out there available and of course we all want them to buy the company that we have on our list or in our portfolios but in the end it’s good for everybody.” Pfizer’s obesity asset did not work out straight away, leading it to buy Metsera; Novo Nordisk bought Akero after finding its FGF21 MASH program compelling. Sam said he did not know whether this was the sixth week in a row with an M&A deal, and mentioned a rumor about Johnson & Johnson and Protagonist as a possible next deal.

2. Novo/Akero: strongest fibrosis data, higher GI burden, and the combination problem

• Sam’s read on the up-to-$5.2B deal: efruxifermin has “leading efficacy data” — 40% fibrosis improvement at 24 weeks rising to 75% at 96 weeks, 39% improvement in F4 patients in SYMMETRY, and MASH resolution three times placebo. Its efficacy looks better than Novo Nordisk’s own data on an unadjusted basis; after placebo adjustment, it is “not that much better” than 89bio’s pegozafermin.
• The inverse holds on tolerability: GI side effects appear higher than with the competing FGF21 drugs, although discontinuation rates look similar across them. Sam’s specific worry is the endgame — “these gastrointestinal side effects worry me a little if we’re going to be looking at a combination with semaglutide.” Dosing is another axis: Akero weekly, 89bio twice weekly in a prefilled syringe rather than a lyophilized powder, and a possible once-monthly schedule for GSK’s program. Bone mineral density is “supposed to be a class effect,” but there may be differences among the drugs.
• The thesis vindication he flagged: when GLP-1 drugs started reporting MASH data, investors asked what was left for FGF21 programs. Novo Nordisk — which already has semaglutide and MASH data — just paid up, supporting the view that “this is a disease area where combinations are likely to be needed.”
• Tess added the scoreboard: Akero, 89bio and Boston Pharmaceuticals/GSK have been acquired or partnered, and she struggled to name another independent clinical-stage biotech with a clinical-stage FGF21 program. She also credited Madrigal with showing that the market and unmet need could support a price point capable of producing robust peak sales. Sam bet that forgotten FGF21 or other mechanisms will emerge from stealth or elsewhere within six to twelve months, possibly from China; Tess said she would not bet against him.

3. BMS/Orbital and the in vivo CAR-T land rush

• Yaron Werber’s framing: “the next innovation on CAR-T is really going to be an in vivo, sort of bedside, LNP-driven modality.” The Orbital deal was described in the transcript as $1.5 million and involves a CD19 program using a proprietary mRNA platform for in vivo delivery, alongside similar early work on BCMA. Yaron noted that even J&J is looking at preclinical BCMA assets despite viewing Carvykti as a winner. The attraction is avoiding apheresis, conditioning and processing.
• Sam’s twist on the buyer: BMS has what he views as the leading anti-CD19 CAR-T in Breyanzi, and Orbital’s press release focused on immune and inflammatory disease. For those patients, “you can’t go with the standard CAR-T process” unless targeting absolute end-stage autoimmune disease; an in vivo, off-the-shelf approach could make treatment more practical. Yaron added that the adverse events of the traditional approach are “just nonstarters,” a concern that became clear at ACR about a year earlier.
• Tess’s taxonomy for the space: transient versus integrating. Orbital and Capstan, which had similar upfront acquisition amounts despite being very early — late preclinical or early clinical — sit on the transient side, which “does seem to lend itself” to broad autoimmune populations with severe but nonlethal disease. Interius was this year’s integrating-side deal. All of it, she noted, is pharma buying early pipeline breadth through platforms that could apply across disease areas.

4. The IPO window reopens narrowly, and new capital arrives

• Nine IPOs have occurred this year, per Tess — almost all in January or very early February, then nothing, followed by LB Pharmaceuticals. Now MapLight has filed its S-1: it is seeking to improve on Cobenfy, the drug Karuna developed, BMS acquired and is currently launching, and is raising money to fund schizophrenia trials. Tess viewed it as another case of biotechs showing big pharma where important innovation may lie, alongside AbbVie’s acquisition of psychedelics company Gilgamesh.
• Luba Greenwood’s new-fund item was Sante Capital — a name she said she might not be pronouncing correctly — founded by two former Moderna leaders and based in Florida. The $325M fund is focused on development-stage, Phase 1/2 biotech companies with multiple assets rather than single-asset stories. The founders say they have already deployed up to $100M across six startups, including Odyssey, and aim to reduce risk by partnering on technical, operational and scientific work rather than acting as “passive backers.”
• Luba’s personal enthusiasm went to Nelo Therapeutics, which emerged from stealth this week with a $101M Series A to modulate neural circuits controlling the immune system. Her thesis is that targeting “central master-regulator circuits” in brain-body pathways could coordinate multipathway immune modulation with less resistance. The science comes from Zuker’s lab, where animal-model work showed neurons that can dial inflammation up or down. She stressed that the evidence is still very early and said the area is highly underfunded. Backers include Column Group, DCVC, Lux Capital and the Gates Foundation.

5. AI-pharma partnerships: pharma shifts toward partnering

• Luba’s arc was a cycle of enthusiasm and retrenchment: “Initially, everybody wanted AI platforms; then everybody was platformed out. Then pharma decided to build its own AI capabilities. Now they realize that they should stick to what they know and partner with AI companies.”
• The AstraZeneca–Algen Biotechnologies deal is worth $555M, with an undisclosed upfront and a mostly back-loaded structure. It is a multitarget research collaboration in immunology under which AstraZeneca will receive exclusive rights to develop and commercialize resulting therapies, with no equity stake. Algen’s Brain platform combines AI with functional genomics, CRISPR screening and CRISPR modulation to map a causal link between gene regulation and disease progression in human cell types and identify therapeutically actionable genes.
• Her caveat was explicit: “we haven’t yet seen an AI-drug-discovery-enabled drug approved in immunology,” though that could change.
• The second deal is with BenchSci, on whose board Luba sits. Sanofi signed a three-year licensing deal for BenchSci’s ASCEND platform, a “disease-biology AI copilot” that combines scientific literature, Sanofi’s proprietary data and experimental metadata into a biological evidence knowledge map. It can support target triage, experimental design, hypothesis generation and new mechanisms of action.

6. Data week: Arcus impresses, Skye’s CB1 antibody underwhelms, Dyne builds a curve

• Arcus, best known for its Gilead TIGIT partnership, pooled about 121 Phase 1 dose-ranging patients on its HIF-2α inhibitor casdatifan in kidney cancer. Sam’s verdict was “frankly it looks pretty good”: 12.2 months of median PFS versus 5.6 for Merck’s belzutifan, a 31% confirmed response rate versus 20%–30%, and grade 3/4 adverse events of 60% versus 62%. No anemia events led to discontinuation in the casdatifan study; Merck’s discontinuation rate was 6%. Median follow-up was about 15 months. His caveats stand — this is a cross-trial comparison involving a single-arm study, pooled dose groups, and 121 versus 374 patients — and he models a $5B addressable market. The stock moved from a $13.35 close on Friday, October 3, to a peak of $16 and a bit.
• Skye Bioscience’s nimacimab is the CB1 story Sam has long liked: use an antibody that should remain peripherally restricted to avoid the neuropsychiatric problems of brain-penetrant oral drugs. The 26-week Phase 2a monotherapy weight-loss result “underwhelmed,” possibly because the dose was insufficient. Combination with semaglutide produced better efficacy than semaglutide alone, and the possible path forward is to increase the delivered concentration — but Sam’s concern is whether a higher dose could produce enough brain penetration to cause neuropsychiatric side effects. None have appeared at the doses tested so far; extension data are expected in Q1 2026.
• Brian on Dyne’s one-year DM1 Phase 1/2 data at the pivotal dose: improvement continued rather than plateauing — vHOT was better at 12 months than at six, and quantitative muscle-testing total score “basically doubled in effect size” from six to twelve months, producing “a pretty nice curve up and to the right.” Avidity is ahead with a bigger program, but “they’re not that far behind timewise,” and Dyne trades at a significant discount. Pivotal data are expected around the middle of next year, with any eventual comparison necessarily cross-trial.

7. Sarepta: sentiment less negative, not positive

• Brian characterized last week’s Sarepta run-up as “people maybe being less negative” rather than a sentiment change, after two deaths from liver toxicity in nonambulatory patients and the summer’s FDA back-and-forth. World Muscle Society posters showed three-year data and younger-patient data with what he reads as a diminished liver-toxicity profile relative to older patients.
• The pivot is prophylactic sirolimus: a single-center experience was presented at WMS, and sirolimus has been part of REGENXBIO’s prophylactic protocol since the initiation of its DMD studies, with similarly diminished liver inflammation. Physicians he spoke with expressed “less of a concern and maybe more of a hope.”
• His balanced close: “a lot of people debate the overall efficacy of the gene therapies in general and how much evidence there is of a true effect,” but “there is generally a belief among the physicians that there is an effect size here, maybe not something that can easily be quantified.” If safety can be secured, Sarepta may get back to a point where it can grow sales again.

8. Ionis re-rated, Lexeo tests CBER, and the panel splits on Peter Marks

• Yaron’s Ionis update covered Spinraza royalties, Wainua taking 40% share from Alnylam within a year through its autoinjector, and Tryngolza — approved in FCS but re-rated on sHTG data showing 72% triglyceride reduction and 85% acute pancreatitis reduction, “literally mic-drop-type data.” Innovation Day added profitability guidance for 2028; peak sales of more than $5B ($3B from the pipeline and $2B from collaborators); and an answer to the standing bear case that “there’s always somebody behind them who’s better with siRNA” — an in-house siRNA APOC3 program with 90% reduction and pharmacokinetics supporting six-month or annual dosing. The siRNA would compete with Ionis’s own monthly ASO Tryngolza. Ionis also unveiled new chemistry in an siRNA platform targeting muscle and moving into HFpEF and heart failure. “This is not your grandfather’s Oldsmobile.”
• Brian on Lexeo’s LX2006 for cardiomyopathy due to Friedreich’s ataxia — a monogenic frataxin replacement therapy showing a dose relationship across three doses in cardiac biopsies — where the real news is regulatory. The FDA is retaining LVMI as a co-primary endpoint alongside frataxin expression, an endpoint previously blessed in Rocket’s Danon program, and apparently may allow pooling Phase 1/2 data with the planned pivotal study starting in the first half of 2026 to lower the sample size. The FDA also indicated that the LVMI measurement window will probably be under 12 months; patients with impaired baseline LVMI are crossing the 10% threshold by six months. Brian’s conclusion was that CBER “isn’t really moving into a much more conservative methodology” and, even if it is not fully carrying on “the Peter Marks tradition,” is “at least offering a high level of flexibility.”
• Then the disagreement. Sam said of Marks joining Eli Lilly that he would not criticize someone who may have been “pushed out” for taking an industry role, but that it was “not going to be helpful” or helpful “for the cause at the minute.” Tess later noted that people who want to use the move as an example of the revolving door would do so. An unidentified panelist took the other side, saying Marks has an “unbelievable skill set” that will be put to use at a valuable company and arguing that leaving medicine for industry was once viewed as sacrilegious before people recognized that “an ecosystem is built of many partners.”
• Tess’s synthesis was the more practical point: conflicts can be managed through disclosure and recusal, and “what’s much worse for our sector is if people aren’t talking to each other.” She asked how much an advisory committee can contribute on a multiple myeloma drug if “you can’t find a multiple myeloma expert who maybe hasn’t been involved in a trial” involving an industry compound. An unidentified speaker then teased that AstraZeneca was apparently about to announce its deal with the administration.