Episode 176 - March 13, 2026

Summary

• The desk pushed back on Sam Fazeli’s macro worry that ~$100 oil from the Iran–Israel–United States war could reignite inflation, rates, and a biotech risk-off. Paul Matteis countered that data readouts are being rewarded and followed by big fundraisings, and that a war-induced recession might pressure rates down, not up. Josh Schimmer’s structural case: biotech is “no longer the casino type,” but “less speculative, more mature, and more GARP-y,” becoming cash-flow-positive “in a way that we’ve never seen before” and insulated from both oil and AI.
• Vinay Prasad departs the FDA in April, and Josh’s verdict is that he was “a very competent regulator, but not necessarily the most appropriate leader.” He “sees through all of the games that get played by companies,” and his stringent bar was “defensible”; reports of deteriorating morale may have cost him the role more than any decision. Paul pointed to DNL310’s Hunter syndrome PDUFA on April 5 (open-label biomarker data) and the possibility of a uniQure refiling path as tests of FDA flexibility; Sam placed the new-equilibrium question in the next three to four months. The later leucovorin/autism FDA news reinforced Paul’s view that the agency remains “odd” and unpredictable, not that an approval was established.
• Servier, with roughly €7B in annual revenue, bought Day One and its OJEMDA (tovorafenib) oncology business. OJEMDA is the first FDA-approved drug for relapsed/refractory patients with BRAF fusions or rearrangements, V600 mutations, and related alterations; sales were $155M in 2025 with $225–250M guided, while ex-US rights are licensed to Ipsen. Servier’s prior Agios oncology acquisition, including Voranigo, may offer synergies. The recurring La Lettre rumor of an AstraZeneca exclusivity window on Abivax was denied by the company, but Graig Suvannavejh sees the cytokine-space profile as attractive to acquirers; the potential revenue at risk was presented as roughly $150B–$200B, with uncertainty.
• IDEAYA’s darovasertib plus crizotinib phase 2/3 uveal-melanoma topline data are due at month-end: the target is median PFS above ~5.5 months, versus 7 months in phase 1/2 and 2–3 months with existing standards of care. A KOL on Graig’s morning call put probability of success as high as 80% (Graig sits at 65%); with only HLA-restricted KIMMTRAK ($400M annualized) approved, an accelerated-approval path could put the drug on the market next year. Sam raised the concern that PFS might not translate into an OS signal, but no substantive answer followed in the transcript.
• Roche’s giredestrant missed in first-line metastatic breast cancer in PERSEVERA on what Josh Schimmer believed was a palbociclib backbone, after LIDERA showed an adjuvant benefit versus placebo. Josh reads oral SERDs as becoming “more of an either/or with the CDK4/6 inhibitors.” The failure surprised a Street that had concluded giredestrant would work post-LIDERA, but it leaves Sermonix Pharmaceuticals’ lasofoxifene second-line opportunity less threatened.
• Vertex’s povetacicept IgAN interim showed a highly statistically significant proteinuria reduction without much differentiation from the anti-APRIL approaches, yet the $100B+ stock was up 10% at one point on “quote unquote expected” news — Paul’s signal of underlying sector appetite. Otsuka’s pricing for the first drug in this class was perhaps twice expectations. Xenon’s phase 3 epilepsy efficacy was better than phase 2, “extremely rare in CNS,” and Xenon and Dianthus each raised roughly $700M on data.
• NMIBC is shifting from monotherapy toward doublets, and Josh thinks CG Oncology can commercialize alone against J&J. Relmada posted close to an 80% one-year complete-response rate with its gemcitabine/docetaxel gel in a small phase 2 dataset; its stock rose about 40% in a week and roughly 2,200% over a year from a $0.25 share price in April last year. Multiple lines can coexist because about half of patients on monotherapy still need something else. Josh said J&J’s early claims of data superiority over CG “completely fizzled”: “a Goliath of a pharma company… not only going after them, but being wrong.”

Deep dive

1. Sam plays the bear on oil and rates; the desk says biotech has changed

• Sam’s opener, deliberately against last week’s “nice, bullish, warm, fuzzy” episode: the Iran–Israel–United States war has oil “hovering around $100 again,” feeding through urea fertilizer and liquefied natural gas into fertilizer and food costs and potentially back toward higher rates — the same rate regime blamed for the bad years before the August turn. His question: if the worry sticks, “what do you think is gonna give first — IPOs?”
• Paul’s pushback, self-described “opposite of an expert on the macro piece”: this was a good week — readouts got rewarded and ended up in big fundraisings, and even Vertex was “really rewarded even though the data weren’t even the bull case.” If the war drags into recessionary indicators, “maybe that actually attenuates the risk of a rate increase.” His honest concession: “the biggest real issue for biotech is just it’s super high beta and it’s a risk-off sector.”
• Josh’s structural case: biotech is insulated from oil prices and from “what’s going on in the AI and software world” — “no longer the casino type,” but “less speculative, more mature, and more GARP-y,” becoming cash-flow-positive “in a way that we’ve never seen before.” Graig’s close: companies can still raise capital on good data, and “it’s still gonna be a pretty good year for biotech.”

2. Prasad’s exit: competent regulator, failed leader — and the tests that follow

• The news: Vinay Prasad departs the FDA in April, successor unknown — and notably no sector bounce this time, unlike his first exit. Josh’s framing, the episode’s sharpest: Peter Marks was controversial for leniency, Prasad for stringency, but Prasad’s bar was “defensible” and he was “incredibly astute… he sees through all of the games that get played by companies” — “a very competent regulator, but not necessarily the most appropriate leader,” with reports of deteriorating morale perhaps costing him the role more than any decision.
• Paul pointed to DNL310’s Hunter syndrome PDUFA on April 5 — a blood-brain-barrier enzyme product with “outstanding data, but open-label data on biomarkers,” making it “the next test of FDA flexibility” — plus what happens with uniQure and whether it finds dialogue and a path to refiling. Sam corrected the horizon from a year to the next three to four months while asking whether a new equilibrium appears before a CBER leader is in place.
• Sam’s lament on both regimes: the director’s job is to ensure reviewers “were doing their job correctly — not to intervene and say, ‘All that work you’ve done, I disagree with.’” Graig’s hedge: “sometimes it’s better to have a devil you know versus a devil you don’t know.”
• The late-episode counterexample — leucovorin and autism: Paul sees political pressure behind getting approval on “very limited data” (though perceived risk is low): “it’s strange to me that something like this could happen and some of these rare disease products couldn’t even get their own fair share of a review.”

3. Housekeeping and history: AEMS, Crowley’s 50-year op-ed, and Sam’s AI-pricing thought experiment

• Graig on the FDA’s new AEMS: FAERS (drugs and biologics), VAERS (vaccines), and a database for animal drugs and animal foods merge into one FDA Adverse Event Monitoring System — legacy platforms cost about $37M per year, with about $120M in savings expected over five years. His read: more seamless information access, “actually a good thing” from an agency more often generating headaches.
• John Crowley’s STAT opinion piece pegs biotech’s 50th anniversary to Genentech’s founding, cites $3.2T in U.S. economic activity, then pivots to pricing: the current administration’s interest in bringing U.S. prices closer to those outside the U.S. may not be good for U.S. biotech. Crowley discusses PBM middlemen and rebates, but does not specifically address Chinese competition.
• Sam’s segue, fresh from four days in China: from one, two, or three drug developers in 2005 — including HUTCHMED — to a sector that took off in the last five to ten years, driven partly by management trained by Western multinationals or biotechs and returning to run Chinese companies. His thought experiment: if AI productivity cuts development time and therefore cost, companies could launch new drugs at lower prices while keeping profitability high, though that could disrupt the pricing of older drugs.

4. Servier buys Day One; the Abivax rumor rides again

• The week’s one real deal: Servier has annual revenue of around €7B and describes itself as the 35th-largest pharma company; it bought Day One’s oncology business centered on OJEMDA (tovorafenib), the first FDA-approved drug for relapsed/refractory patients with BRAF fusions or rearrangements, V600 mutations, and related alterations. Sales were $155M in 2025 with $225–250M guided, while ex-US rights are licensed to Ipsen. Josh’s connective tissue: Servier bought Agios’s oncology assets about half a decade ago, including Voranigo for another brain tumor, whose launch “has actually gone quite well” according to some reports — possible synergies with Day One.
• The recurring rumor: La Lettre claimed Abivax gave AstraZeneca a period of exclusive data access, after which it might seek other buyers or continue discussions; Abivax said “that’s not true.” Graig still thinks the cytokine-space profile is attractive to an acquirer and “wouldn’t put it past anyone” this year or next, given that large pharmas may face roughly $150B–$200B of revenue at risk over five years and may need external sources of innovation. Eric Tokat’s Bloomberg line: $20B+ in M&A is still expected.

5. IDEAYA’s uveal melanoma readout: low bar, big event

• The setup, from Graig’s coverage: darovasertib, an oral PKC inhibitor, plus crizotinib, with phase 2/3 topline median PFS due at the end of the month in uveal melanoma — where the only approved drug is Immunocore’s KIMMTRAK, the very first TCR-based therapeutic ever approved, annualizing about $400M in its fifth year but restricted by HLA status. Patients with existing standards of care in a non-HLA-restricted setting are getting only 2–3 months; phase 1/2 showed a seven-month benefit, and “anything better than perhaps five and a half months” would be meaningful. “Patients sometimes need to get their eye removed, and no one wants to lose an eye, literally.”
• Odds and path: a KOL on Graig’s client call that morning put success probability as high as 80%; Graig is at about 65%. The company has suggested an accelerated-approval pathway — potentially putting the drug on the market next year on strong data.
• Sam asked whether a decent PFS result could fail to translate into a large OS signal, given KIMMTRAK’s opposite pattern of weak PFS but proven overall-survival benefit. The excerpt contains no substantive answer to that question.

6. Giredestrant’s first-line miss redraws the oral SERD map

• PERSEVERA — phase 3 giredestrant in first-line ER+/HER2- metastatic breast cancer — was negative, which Graig says “takes a little bit of the luster off.” For his coverage of Sermonix Pharmaceuticals, previously a CNS company that brought in lasofoxifene, the old osteoporosis SERM now in the phase 3 ELAINE III second-line study: a positive giredestrant result might have shrunk its opportunity in patients with the relevant mutation by reducing aromatase-inhibitor use — “we don’t have to worry about that right now, unfortunate for patients.”
• Josh’s synthesis of why: adjuvant LIDERA worked because it beat placebo without a CDK4/6 backbone; PERSEVERA sat, he believed, on top of palbociclib and added nothing — so oral SERDs become “more of an either/or with the CDK4/6 inhibitors,” left to carve out earlier lines where CDK4/6 inhibitors are not used, perhaps because of safety advantages. “Kind of an unusual setup.” Sam’s note on the stock hit: after LIDERA everyone had concluded it would work.

7. Small-cap tour: Alfasigma acquires GSK’s PBC asset, Benitec’s swallow data, Relmada’s 2,200% year, CG vs Goliath

• GSK’s late-stage IBAT inhibitor for PBC, linerixibat, was acquired by Alfasigma for about $300M upfront plus biobucks and royalties — “not a huge sum given the size of the PBC market.” Alfasigma is the acquirer that bought Intercept before Ocaliva was ultimately pulled; for Josh’s Mirum, trailing by a couple of years with its own IBAT inhibitor, having a commercial rival of Alfasigma rather than GSK may provide “a little bit more breathing room.” Mirum’s drug may have higher relative exposure and therefore potentially greater efficacy, with some signals pointing that way.
• Benitec: an AAV gene therapy injected locally into throat muscles for oculopharyngeal muscular dystrophy, now with two years of low-dose follow-up showing continued swallow improvement and one high-dose patient doing well. Josh’s real story is CEO Jerel Banks, a former investor whose “yeoman’s work” — including solving how to measure swallowing for regulators — essentially turned the company around. It remains some distance from a clearly approvable profile.
• Relmada licensed from an Israeli company a unique gel formulation delivering gemcitabine and docetaxel simultaneously for high-risk NMIBC; its small phase 2 dataset showed a close-to-80% one-year complete-response rate, “meaningfully above” some later-stage competitors, including ImmunityBio, J&J, and CG Oncology. The stock rose about 40% on the week and roughly 2,200% over the last year; Sam noted it was $0.25 in April last year. Josh’s theme: the market may already be shifting from monotherapy to doublets, with CG itself combining its oncolytic virus — now in a rolling BLA for high-risk NMIBC — with gemcitabine.
• Sam’s challenge — can CG go up against J&J in urology alone? Josh: yes. Arthur Kuan’s roughly decade-long turnaround built deep relationships in a market dominated by a handful of large urology groups; the J&J and CG products also differ substantially in administration and side-effect profile. With about 50% 12-month complete responses for monotherapy, half of patients may still need something else, so “it’s not an either/or.” J&J’s TAR-200 has set a healthy price, and J&J’s early claims of data superiority “completely fizzled” — “a very unusual situation to see this Goliath of a pharma company almost going after a very small emerging biotech… and not only going after them, but being wrong.”

8. Vertex, the pricing tailwind, twin $700M raises — and the obesity data Sam can’t parse

• Paul on Vertex’s povetacicept IgAN interim: a highly statistically significant proteinuria reduction but not much differentiation from the anti-APRIL antibodies from Otsuka and Vera — the dual APRIL/BAFF efficacy punch did not clearly materialize, though safety looked clean. Yet a $100B+ company was up 10% at one point “on news that was quote unquote expected” — his tell that there’s “a general appetite to own more high-quality biotech companies.”
• The pricing offshoot: Otsuka, first in this class, priced its drug at perhaps twice expectations. Paul thinks that “maybe doubled the TAM here versus what people were thinking a year ago,” and notes “one thing we haven’t complained about a lot in this call in the past year is drug pricing.” Graig’s echo: companies may feel flexibility to price higher, partly because “their desire to launch ex-US is getting smaller and smaller.”
• Xenon’s phase 3 data for azetukalner were better than phase 2, “extremely rare in CNS,” perhaps aided by a well-managed placebo arm. Paul’s old joke: “if Xenon had the Vimpat effect size, the market cap would be cut in half, and yet if Xenon has the Vimpat commercial outcome, that would be a great outcome” — efficacy is a weak correlate of commercial success in epilepsy. Xenon raised over $700M; Dianthus also raised $700M on interim CIDP data in a crowded but large IVIG market.
• Sam’s obesity lightning round: Roche-Zealand data hit Zealand hard, but topline weight-loss percentages without gender balance or baseline BMI “continue to wind me up” — women tend to respond better, and the details matter. AbbVie-Gubra and China datasets show the space “getting super competitive” while Lilly and Novo keep most of the share; a surprise BioNTech management departure closed the episode.