Episode 154 - September 12, 2025

Summary

• The financing window shows signs of reopening, but the hosts disagree on how wide. LB Pharmaceuticals’ clinical-stage schizophrenia IPO was well received, rose and held, while Maze ($150M PIPE), Rapport ($250M follow-on), CAMP4 (recap) and Amylyx ($175M) all raised in one week; H.C. Wainwright’s Richard Gormley told Brad Loncar, “for the last two years our companies have been in survival mode and now they’re actually thriving.” Eric Schmidt’s caveat: most raisers were recaps or upper-mid-caps with good data — “I’d be a little bit cautious in thinking that it’s game on for fundraising.”
• A drafted China-biotech executive order is a major overhang, and the worst-case version could freeze the deal machine. Per the NYT’s leaked copy, pharma is lobbying to preserve cheap, derisked Chinese assets while investors, including Peter Thiel, Jared Kushner, the Koch brothers and Sergey Brin, push to tighten the screws; mandatory CFIUS review of licensing deals “has the potential to kill all of these deals,” per Loncar, and the FDA could be told to ignore China data at all stages.
• Brad Loncar argues the real threat is structural, not geopolitical: pharma may simply bypass U.S. biotech. Big pharma has sourced 50%+ of its pipeline from U.S. biotech for 20-plus years, but GSK–Hengrui-style deals ask, “do we even need to go through U.S. biotech?” while the FDA remains “the most inflexible” on getting drugs into the clinic. “That’s where China has been eating our lunch.” Chris Garabedian’s chips analogy was that Nvidia and AMD can sell to China while paying 15% of revenue to the U.S.; Sam Fazeli added that Merck’s canceled London R&D facility shows the issue also affects U.K. jobs.
• Novartis paid $1.4B for Tourmaline, whose stock rose almost 90%, revisiting the inflammation–atherosclerosis thesis after its 2018 canakinumab CRL. The asset, pacibekitug, is an IL-6 antibody Pfizer shelved — another SpringWorks/Cerevel-style monetization of Pfizer castoffs, headed to Phase 3 in atherosclerotic disease.
• Revolution Medicines posted a 55% ORR with its pan-RAS inhibitor in first-line pancreatic cancer versus high-30s for FOLFIRINOX and low-40s for NALIRIFOX. Eric was categorical that “these drugs of theirs flat-out work,” while PRMT5 inhibitors from Bristol Myers Squibb, Amgen and Tango represent another promising mechanism after 30 years of almost nothing working in PDAC.
• Eric thinks the market got Summit’s 25% HARMONi selloff (~$5B off a ~$20B cap) wrong; Sam Fazeli thinks the release may have been an “own goal.” Eric saw “more consistency than inconsistency” across geographies and calls the filing obsession “completely off the mark” — filing would start the IRA clock and be “probably net-present-value-destroying.” Sam’s rebuttal: subtract U.S. patients and the European hazard ratio shows zero benefit on short follow-up, while the discussant attributed most efficacy to VEGF, not PD-1.
• Takeda announced two Phase 3 trials for its in-house orexin-2 agonist in narcolepsy — “an intractable disease literally up until a week ago.” Roughly 80% of treated patients were essentially normalized; peak-sales estimates run $1.5B–$3B against Alkermes’ once-daily challenger, which is still pre-Phase 3.
• The FDA’s new transparency regime cuts both ways: near-daily FAERS dumps knocked Soleno down 25% on an event the investigator deemed not drug-related, while genuinely useful adcoms get canceled. Eric’s line — “correlation is not causation. That’s not something that MAHA always understands” — pairs with a co-host’s warning on Capricor’s surprise CRL posting: “every communication between the company and the FDA turns into a public press release. Is that really healthy?”

Deep dive

1. The window is open — Eric hopes the market keeps its filter

• The tape: LB Pharmaceuticals’ clinical-stage schizophrenia IPO was well received, rose and held — “good price discovery,” in Chris Garabedian’s read, not a high-flying bubble print. Alongside it: Maze’s $150M PIPE, Rapport’s $250M follow-on after Phase 2 epilepsy data, CAMP4’s $50M-to-$100M recap PIPE with a board change-out, and Amylyx’s $175M to commercialize avexitide for post-bariatric hypoglycemia. Avidity’s post-DMD-data offering traded down as the market read M&A off the table — though Chris notes raises like that are “always tricky to read”; sometimes they provide negotiating leverage.
• Brad’s field report from H.C. Wainwright, via banker Richard Gormley: “for the last two years our companies have been in survival mode and now they’re actually thriving” — companies talking pipeline adds, not survival.
• Eric’s brake-tap: the raisers were either recaps (“certainly not a great thing to do”) or upper-mid-caps with good data. “Hopefully the market has a strong filter… We don’t want to see what we saw three or four years ago. And I don’t think we’re going to, by the way.” LB is “a data point of one” on private-company IPOs.
• Tess on macro: the not-good employment report cheered people, then CPI “wasn’t looking good” — the Fed is in a tough spot, but a rates-driven macro “at least is a world that biotech is kind of used to” after the policy-shock hiatus. Chris adds the XBI is in the 90s, below the $100 breakout, and generalists still need to see performance — the sector’s ex-COVID track record has not earned the risk allocation.

2. Novartis–Tourmaline: back to the inflammation well, $1.4B

• Eric’s anatomy of the deal: Tourmaline went public via reverse merger in 2023 with perhaps narrower ownership, took an IL-6 antibody, pacibekitug, that Pfizer shelved, and drove it to Phase 3 readiness in atherosclerotic disease — “much the way of SpringWorks or Cerevel,” Pfizer assets monetized by others. The stock rose almost 90%.
• The buyer’s logic is the tell: Novartis has pursued the inflammation-to-atherosclerosis link since canakinumab and CANTOS, which showed a MACE reduction yet drew a CRL in 2018. “They’re coming back to the well” with a different mechanism — “and hopefully going to take this into Phase 3 with maybe a better ending.”

3. The China executive order: two lobbies, multiple trade-offs

• The NYT, through Rebecca Robbins and colleagues, obtained a drafted executive order; the White House called it back-burner, but Brad’s framing stands: “two polar-opposite views of the world are lobbying the Trump administration” — pharma wanting to preserve cheap, derisked Chinese assets versus investors arguing the flow guts U.S. biotech. The harshest lever, mandatory CFIUS review of licensing deals, “has the potential to kill all of these deals” by making them time- and resource-burdensome; a softer one would have the FDA ignore China data at every stage, not just for filings.
• Sam’s patients-first case, with his best specimen: serplulimab, a Henlius PD-1 inhibitor, was approved in Germany for first-line extensive-stage SCLC, ranked above Roche’s drug and AstraZeneca’s durvalumab in ESMO guidelines, and is now running a U.S. head-to-head against Tecentriq. He also cited Carvykti and Brukinsa as examples of drugs developed in China. “There’s no rhyme or reason why the best drug for a disease has to be found in the U.S. or in Belgium or in France… whatever happens here, there has to be some control over not curbing the access of patients to the best drugs.”
• Chris’s chips analogy: Nvidia and AMD can sell chips to China while paying 15% of that revenue to the U.S.; perhaps a future order could allow Chinese drugs while taking 10% of revenue. The point was hypothetical, but Chris argued that restricting innovation’s location would not change China’s system and could push it to develop its own chips.
• Brad’s counterargument that “nobody brings up”: China just held a military parade celebrated by Putin and the North Korean dictator — “a communist nation which many of us would argue is unethical and evil” and widely seen as an adversary. Politicians must weigh funding its biotech sector “to the detriment of science and jobs and biotech in our own country.” Brad also pushes back on the jobs concern: a failed U.S. drug can gain “new life” when a company accesses global innovation, and U.S. patients lose access whether Beijing or Washington blocks a good medicine.
• Brad reframes the issue as structural and “bigger than the whole geopolitical” question: he is not worried the U.S. cannot compete on first-in-class innovation — the real crowding is around validated targets and fast followers. The danger is pharma, which has fed 50%+ of its pipeline from U.S. biotech for 20-plus years, deciding via GSK–Hengrui-style deals, “do we even need to go through U.S. biotech?” And the fixable half: the FDA is “the most inflexible” on getting drugs into the clinic quickly — “that’s where China has been eating our lunch.”
• Sam’s tangible coda: Merck canceled its large London R&D facility, while GSK signed its Hengrui partnership a month earlier — thousands of U.K. R&D jobs that “tangibly do not exist today.”

4. Revolution’s 55% ORR in frontline pancreatic — plus a new MG mechanism

• Tess’s numbers: Revolution Medicines’ pan-RAS inhibitor posted a 55% overall response rate in first-line PDAC against comps of high-30s for FOLFIRINOX and low-40s for NALIRIFOX — both regimens with significant tolerability problems — with Phase 3 planned against Gem/Nab, the lower-response but better-tolerated standard. PFS and duration of response are still pending, “but certainly really impressive.”
• Eric, unhedged: after “30 years of almost nothing working in pancreatic cancer,” “these drugs of theirs flat-out work” — every confidence the frontline study reads out positively, with PRMT5 inhibitors from Bristol Myers Squibb, Amgen and Tango representing another promising mechanism. Chris’s frame on the strategy: frontline is risky, “but if you win, you win big.”
• Rapid-fire from Eric: Dianthus’ Phase 2 C1s-inhibitor data in myasthenia gravis introduces a new mechanism against the existing FcRn and C5 inhibitors, in what may become a $10B market; the stock reacted well and the company raised capital.

5. Takeda targets narcolepsy’s cause after 25 years — in-house

• Brad, fresh from interviewing CEO Christophe Weber: orexin deficiency was identified as narcolepsy’s cause around 2000; a Japanese co-discoverer started working with Takeda 25 years ago, one agonist failed, and the orexin-2-selective successor was presented alongside two Phase 3 trials at World Sleep Congress — roughly 80% of patients on the drug were returned to a normal lifestyle. “Intractable disease literally up until a week ago.”
• The setup: about 150,000 U.S. patients (“a more common rare disease, for lack of a better term”), peak-sales estimates of $1.5B–$3B, and Weber has long called it the pipeline’s most important program. The debate is Takeda’s twice-daily first-mover versus Alkermes’ once-daily candidate, which showed Phase 2 data at the same conference but still must run and win a Phase 3. Brad’s kicker: everything was developed at Takeda from the start — a counterexample to the outsourced-science critique of big pharma.

6. World Lung: SCLC ADCs hunt for a post-IMDELLTRA niche; Summit’s possible “own goal”

• Sam’s map from Barcelona: B7-H3, DLL3 and SEZ6 ADCs are all crowding second-line SCLC — Daiichi Sankyo’s ifinatamab deruxtecan updated a 137-patient set roughly in line with topotecan and carboplatin, perhaps a little better, with more side effects. Chris also cited an Hengrui/IDEAYA program with a 6.7-month PFS and said AbbVie’s looked similarly strong. Amgen’s tarlatamab looks strong too: DeLLphi-303’s first-line add-on showed 25.3-month OS in 48 patients, while none of the ADC trials show performance in post-IMDELLTRA patients, an important unanswered setting. BioNTech’s BNT327, a PD-L1×VEGF bispecific, showed decent first-line SCLC responses at only 6.5 months’ median follow-up; Sam said, “I’m more of a fan of a PD-L1 approach for these VEGF bispecifics.”
• Eric’s contrarian take on Summit’s Monday selloff — down about 25%, or roughly $5B off a ~$20B cap — after HARMONi in second-line EGFR-positive lung: across PFS, OS, hazard ratios and geographies, “I thought there was more consistency than inconsistency… I’m not so sure the market has it right.” The media’s U.S.-filing fixation is “completely off the mark” — filing early starts the IRA clock, shortens exclusivity, and “would probably be net-present-value-destroying”; if Summit had to choose, “they’d choose not to file.”
• Sam’s rebuttal, with the load-bearing detail: subtract the U.S. patients from the Western forest plot and Europe shows zero benefit — because follow-up is too short — so the press-released 0.7 U.S. hazard ratio is highly sensitive to a small number of patients. The discussant said most of the efficacy comes from VEGF, not PD-1 (“a lot of people thought that was too harsh”), and physicians he questioned said, “I’m not so sure now.” Given Summit had messaged registrational potential, “it might have been a bit of an own goal.” It may have been better to wait six or 12 months.

7. Transparency cuts both ways: FAERS dumps up, adcoms down

• The safety ledger first: Intercept’s drug for PBC was pulled from the market after accelerated approval but rejection of full approval; Chris said the company had been acquired by Alfasigma. Capsida halted its pediatric genetic-epilepsy gene-therapy trial after the first patient dosed died within days. Chris: every time gene-therapy headwinds seem to lift, “we still aren’t out of the woods” on safety.
• Soleno fell 25% on a FAERS-flagged event that the clinical investigator deemed not drug-attributable, amplified by an active hedge-fund short attack. Eric’s verdict on near-daily FAERS releases: “correlation is not causation. That’s not something that MAHA always understands,” and the damage is “not just bad for investors… it causes commotion, it causes distress, and maybe it even prevents patients from getting access to good medicine.” A co-host adds that tiny-N, context-free FAERS events invite misinterpretation that a richer release would help prevent.
• A co-host’s contrast — the through-line of the segment: raw transparency is rising, with adverse-event feeds and more background on CRLs, while the transparency that actually helped — adcoms — is vanishing. Biohaven’s panel was canceled, as was Travere’s for Filspari in FSGS. “Transparency is good when it helps you better understand what to do with a piece of information.”
• Capricor’s CRL was posted without notice, so the company publicized its rebuttal — a co-host warned, “every communication between the company and the FDA turns into a public press release. Like, is that really healthy?” Chris, drawing partly from the Sarepta experience, notes adcoms can themselves be politicized through question-crafting, but backs them “if used appropriately and objectively.” He also flags the SEC action against a former FibroGen chief medical officer for manipulated data as the kind of enforcement “we need to keep all of the biotech companies honest.”

8. Endnotes: Pfizer v. BridgeBio, Arena’s black box, Novo’s PD-1 rhyme

• On Bloomberg’s claim that Pfizer isn’t playing fair against BridgeBio in ATTR-CM, Sam declines to adjudicate but supplies the motive: he says Pfizer’s drug is going off patent in 2020, while consensus has Vyndaqel peaking around 2027 and falling from 2028 onward. “Are we surprised that Pfizer’s fighting tooth and nail?” Pfizer says the video identifying its drug as the only approved drug in the class was “played in error”; BridgeBio says the tactics are constant. “It’s a cutthroat business out there.”
• Brad on Harvey Berger taking the Arena Bioworks CEO seat alongside Stuart Schreiber: Berger “could be sitting on a beach right now” and his excitement is “palpable and genuine,” which lends credibility — but “they’ve disclosed essentially nothing about what they’re working on,” deliberately, to avoid fast followers.
• Novo’s cuts shocked Sam: 11% of the workforce, 9,000 people, including 5,000 in Denmark, a return-to-office order, and a promised “more performance-led culture — which of course makes you wonder what it was before.” He wonders whether it signals direct-to-consumer selling “is not as straightforward as it seems”; shares reacted mostly positively.
• Brad’s history-rhymes tweet: GLP-1s are replaying the PD-1 story — Bristol led, Merck’s Keytruda strategy overtook, and look at the Keytruda/Opdivo sales gap — though here it is less strategic error and more that “Zepbound seems to be a better drug.” Sam’s coda: nobody expected Lilly’s GLP-1/GIP combination to work the way it did, “but let’s not forget GLP-1 came out of Novo” — Novo is positioned to get back.