Summary

Novo Nordisk’s Amycretin set a new speed benchmark with 22% weight loss at 36 weeks among patients who reached 20 mg and remained on treatment. That rivals what tirzepatide delivers over roughly 18 months, but the result was as-treated and Novo disclosed neither discontinuations nor meaningful tolerability data beyond saying it was consistent with incretin therapies. Brian Skorney’s hinge question: “Is it 5% discontinuation, or is it 15% discontinuation?”—because 15% would materially weaken both efficacy and safety.
Sam Fazeli sees obesity efficacy approaching a ceiling where tolerability matters more than another percentage point of weight loss. Amycretin looked slightly better than Lilly’s retatrutide at the same point, with a cited placebo-relative result of 24%, yet “23%, 24%, 25%—I don’t care anymore.” Its single-molecule GLP-1/amylin design may nevertheless be a positive read-through for Zealand’s amylin program, whose shares rose roughly 4%.
The XBI’s roughly 5% post-JPMorgan rise looked technical rather than like a durable return of biotech fund flows. Tim Opler heard that one or two pressured funds were covering shorts, while investors remained reluctant to wall-cross for financings. Selectivity is stark: Ascentage completed an IPO at disappointing pricing, Disc Medicine upsized its follow-on, and private investors sounded more constructive than public-market hedge funds.
Dealmaking indicators point to an active 2025 even if the ITCI acquisition remains exceptional in size. Stifel’s M&A and licensing pipeline was at a “historic high,” with multiple billion-dollar-plus transactions in process; deals carrying more than $75 million upfront rose from six in 2023 to 17 in 2024, approaching 2020’s peak of 20. Chinese assets represented about 9% of agreements, including 19 fourth-quarter deals, while oncology retained a 25-30% share.
The new administration’s freezes at NIH, FDA and other health agencies produced a genuine split over whether this is a short transition pause or a strategically dangerous disruption. Luba Greenwood expected a few weeks of reprioritization followed by a more pro-business environment; Tim warned that the government was putting “sand in the gears” just as China accelerates scientific investment. Sam highlighted the unusual cancellation of meetings, travel and reviews alongside the WHO withdrawal and roughly $700 million of U.S. support over the prior year placed in question.
Carvykti’s $334 million quarter confirmed multiple myeloma as both a major commercial market and a coming competitive stress test. The therapy has overall-survival data, more than 90% response rates and growing treatment entrenchment, while Arcellx and Gilead may compete through a cleaner neurological safety profile rather than efficacy. Off-the-shelf bispecifics add a second threat because community centers can administer them more readily.
The $500 billion Oracle–OpenAI–SoftBank AI initiative lifted biotech narratives, but the panel separated useful capital formation from promises that understate biology’s difficulty. Larry Ellison’s cancer-vaccine comments helped Moderna far more than BioNTech even though he named neither company; Sam’s bottom line was that the stock reaction “makes no difference to the data readout.” The credible near-term opportunity is narrower: large-scale healthcare data, imaging, diagnostics and workflow automation.
Agentic AI is moving from content generation toward autonomous execution just as Biogen moves in the opposite organizational direction—shrinking internal discovery and leaning harder on external assets. Luba Greenwood cited research, trial, manufacturing and commercial agents, plus a forecast that 40% of the workforce could be automated over 20 years; Chris Garabedian stressed accountability when systems diagnose or prescribe. Biogen’s unspecified research cuts, meanwhile, acknowledge that celebrated preclinical science did not replenish its “melting ice cube” MS franchise. Sam also cited a January 20 Nature Medicine analysis reporting 40 positive versus 18 negative GLP-1 events, with rheumatoid arthritis among the negative findings, while the roles of weight loss and GLP-1 agonism remained unresolved.

Deep dive

1. Amycretin raised the efficacy bar while withholding the data that decide its value

Brian described Novo’s once-weekly subcutaneous Amycretin as a single molecule targeting both GLP-1 and amylin receptors. Patients escalated every 12 weeks from 1.25 mg to 20 mg, and those who reached the top dose and stayed on treatment lost 22% of body weight by week 36—the fastest weekly pace he had seen.
The comparisons made the headline compelling: tirzepatide takes roughly a year and a half to reach similar weight loss, while earlier phase-one programs from Amgen and Viking delivered around 14% at six months. “In some respects, this does set a new bar.”
Brian’s objection was methodological: Novo presented an as-treated analysis excluding discontinuations and supplied no detailed safety table. A true intent-to-treat analysis or conservative missing-data imputation would probably reduce the 22%; “if it was 15% discontinuation, that would be really bad,” whereas 2% would support the result.
Sam sharpened the contradiction: “This was a safety and tolerability study, and we didn’t get any safety and tolerability comments.” He considered the efficacy only slightly ahead of retatrutide and argued that the market is “scratching or hitting the ceiling”; patients now need comparable loss without nausea, vomiting and treatment abandonment.

2. Amycretin’s mechanism matters more than one extra point of weight loss

Sam contrasted Amycretin’s monomolecular GLP-1/amylin design with CagriSema’s two separate agonists. He did not believe that architecture alone explained the readout, but if it did, the implication for Zealand’s combination approach would be negative.
Sam said both studies appeared not to include lifestyle changes, adding “at least in the amycretin trial”; he put the placebo-relative weight-loss figure at 24%. His preferred interpretation was a positive read-through for Zealand’s amylin program, including its planned GLP-1 combination; Zealand rose about 4%.
Novo’s roughly 7.5% gain could not erase the larger valuation reset. Despite retaining a market value above $300 billion, its shares remained below the August 2023 SELECT cardiovascular-data level, after once moving above $90 and later reaching roughly $140.

3. Public biotech’s bounce lacked conviction, while private capital regained appetite

Tim said the XBI entered JPMorgan down almost 20% over three months, and reasonable M&A announcements failed to trigger the hoped-for conference rally. Investors were declining wall-crosses—“I don’t really care what your PIPE deal is”—and choosing to hold existing positions.
The subsequent roughly 5% rise appeared driven by heavy short covering at one or two pressured funds, not “a giant inflow of fresh money.” Tim’s conclusion was deliberately cautious: sentiment remained poor, even if the squeeze showed why this was becoming “not a good time to be short this market.”
Capital markets were functioning for the right issuers. Ascentage completed its IPO, though below hoped-for pricing, while Disc Medicine’s follow-on was upsized; Chris’s distinction was that financings born of strength can attract demand, whereas cash-desperation deals impose dilution without favorable valuation support.
Private investors sounded more optimistic because their time horizons are longer and funds must deploy dry powder. Luba saw fewer mega-rounds, renewed interest in oncology and novel mechanisms, but continued risk aversion through a preference for clinical-stage assets; Chris expected product and proof-of-concept stories to outrank the platform exuberance of 2018-2021.

4. M&A pipelines suggest ITCI was an opening signal, not the whole year

Tim reported that Stifel’s M&A and licensing workload was at a “historic high,” with pharma active across the board and multiple billion-dollar-plus transactions underway. The ITCI deal’s size may be rare—$10 billion-plus acquisitions cannot recur constantly—but he rejected the idea that one JPMorgan Monday exhausted demand.
Luba’s count supported the direction: transactions with more than $75 million upfront rose from six in 2023 to 17 in 2024, versus 20 at the 2020 high. With the obvious caveat that “nobody has a crystal ball,” she expected 2025 could exceed that prior peak.
Sam calculated that Chinese assets represented about 9% of pharma agreements, with 19 deals in the fourth quarter alone. Upfronts and total values showed little visible discount versus U.S. or European assets, while oncology remained 25-30% of activity; the unresolved question is whether buyers accelerated signings ahead of possible U.S. restrictions.

5. Washington’s health-agency pause divided the panel on intent and consequence

Sam’s Washington contacts considered the NIH stoppage—meetings, travel and other activity—more extensive than a normal transition. He hoped it was temporary but paired it with RFK Jr.’s earlier promise: “We’re going to give infectious disease a break for about eight years.”
The WHO withdrawal reinforced his concern. U.S. support had been roughly $700 million over the preceding year—“a quarter of the overspend on Air Force One”—and funded work including polio eradication, tetanus and flu programs; Sam questioned what leverage justified leaving that activity in limbo.
Luba heard immediate anxiety from biotechs whose FDA conversations had slowed or gone unanswered. She also said NIH grants and review activity had paused. Her counterweight was that FDA teams expedited many reviews before Christmas, making the current interruption look like a short reprioritization; she advised companies to allow “an extra few weeks” before assuming lasting damage.
Tim warned that the administration was throwing “sand in the gears” of institutions funding basic science just as China moved from Cultural Revolution-era hostility toward academics to aggressive scientific and capital investment. Chris resisted a “science versus anti-science” frame, arguing that reform could address inconsistent FDA feedback and inefficiency; the disagreement was whether disruption is a productive tactic or an ill-timed geopolitical risk.

6. Carvykti’s commercial lead now meets safety and convenience competition

J&J reported $334 million of quarterly Carvykti sales, comfortably above expectations and following another third-quarter beat. Brian called it the strongest CAR-T launch to date; the adjacent myeloma economics are enormous, with quarterly Darzalex sales above $3 billion and the drug becoming J&J’s largest-ever brand.
Sam said he thought Carvykti’s overall-survival evidence and deep responses would make Bristol Myers Squibb’s Abecma “history,” at least in terms of meaningful sales. Both Carvykti and the Arcellx/Gilead program produce response rates above 90% with deep minimal-residual-disease effects, so “this battle” turns on delayed neurological toxicity and practical patient management.
Entrenchment favors Carvykti, but Arcellx brings a very strong CAR-T partner in Gilead and a potentially cleaner safety profile. Both must also contend with off-the-shelf bispecific antibodies that can move earlier and reach community centers without the same logistical burden.
Chris used multiple myeloma’s evolution since Celgene launched Revlimid to show how an orphan hematologic cancer can create tens of billions in value. Sam supplied the modeling caveat: endless combinations and treatment lines also make it “one of the toughest markets to model.”

7. AI capital is welcome, but cancer-vaccine timelines remain biology-bound

Brian viewed the announced $500 billion Oracle–OpenAI–SoftBank infrastructure effort as another moonshot-style vision. Investment in biomedical innovation and mRNA cancer vaccines is beneficial, but the promised speed understates the distance from bench to bedside: “I have a love-hate with these sort of grandiose statements.”
Sam granted technology leaders credibility in massive healthcare datasets, imaging, diagnostics, radiology and cardiology, but not automatic authority over chemistry and drug discovery. Moderna’s shares recovered roughly 20-25% from their post-guidance low and gained another 8-10% around the AI narrative; BioNTech, despite its own AI-enabled mRNA vaccine work, rose only about 8%.
Larry Ellison had not named Moderna, and neither story changed the decisive experiment: whether an mRNA vaccine improves response, progression-free survival or survival alongside a checkpoint inhibitor. Chris echoed Regeneron’s George Yancopoulos: machine learning can recognize patterns, but experienced drug developers must contextualize its output.

8. Agentic AI shifts the debate from answering questions to executing work

Luba defined AI agents as systems capable of autonomous decisions, unlike assistants that merely generate content for a user to sift through. Healthcare, biotech and life sciences create a cited 30% of global data volume, while a McKinsey projection put workforce automation at 40% over the next 20 years.
NVIDIA’s announced work with Illumina and IQVIA illustrated the operating model: specialized agents for research assistance, R&D, clinical-trial workflows, manufacturing and supply-chain optimization, commercialization and marketing. Building proprietary agents internally could become differentiated IP.
Chris’s unresolved issue was accountability when an autonomous system diagnoses or prescribes. Sam preferred plainer terms such as “automated lab assistant”; Luba countered that biotech’s own vocabulary confuses machines more, while Chris noted that new terminology also serves marketing by forcing users to reconsider familiar automation.

9. Biogen externalized discovery as broader GLP-1 evidence kept expanding

Biogen’s research restructuring implied unspecified staffing cuts and a footprint aligned with a “prioritized pre-clinical portfolio.” Brian contrasted that with a decade earlier, when Wall Street viewed Biogen as possessing large-cap biotech’s most attractive pipeline and research organization.
Brian linked the failures of strong preclinical programs to CEO Chris Viehbacher’s greater emphasis on external innovation and M&A. Chris posed the trade-off: whether cutting internal R&D could improve earnings, stock price, stock currency and cash, enabling the acquisition of de-risked products capable of replacing the “melting ice cube” MS franchise.
Sam closed with a January 20 Nature Medicine analysis reporting 40 favorable events versus 18 adverse ones for GLP-1 agonists. The positives included Alzheimer’s disease, schizophrenia and reduced suicidal ideation; arthritic changes, including rheumatoid arthritis, stood out negatively. Whether the findings reflect weight loss or GLP-1 agonism remained unresolved: “Time will tell.”

Episode 129 - January 31, 2025 Episode 127 - January 17, 2025