Episode 165 - December 5, 2025

Summary

• All four co-hosts enter 2026 constructive, with the sector’s character fundamentally changed: winners are “not propped up on a dream” but on real launches. Paul Matteis lists the drivers — drug-pricing threat “relatively minimal,” more M&A coming, XBI nowhere near all-time highs — and Brian Skorney adds large-cap pharma still trades at about 15 times while the five-year XBI chart “still looks relatively cheap,” quoting Rod Wong that “unfortunately, the FDA is now the major remaining source of policy uncertainty.” Graig Suvannavejh calls it “the third inning” of the rally; Brian’s one macro watch item is interest rates, given XBI’s link to the long investment cycle.
• The IPO over/under became the episode’s live wager: Josh initially says 15, Paul echoes 15, Josh then puts the over-under near 50, Graig says 50, Brian takes the over, and Paul later says he may be under 50 while hoping for 40s, 50s, or even 60. Josh’s case: three years of a closed window accumulated “a crop of such mature, de-risked companies that really deserve to be public.” Graig’s guardrail is that the public market can value development-stage assets but has never sustainably analyzed preclinical ones; when preclinical IPOs return it’s “total FOMO… it winds up being the signal of the crash,” though probably not in 2026 or 2027.
• Rick Pazdur’s resignation roughly three weeks after taking CDER removes what Brian called “the last adult in the room,” with Tracy Beth Høeg — credited with the disputed 10-childhood-COVID-deaths analysis — installed as interim director. Brian notes essentially the entire 20–30-year FDA old guard is gone and that despite Vinay Prasad’s claims he isn’t micromanaging, “all the signals are that he is inserting himself into decisions all over the place” — neither Brian nor Paul has spoken to anyone who’s had a CBER meeting with Prasad present. Brian’s practical advice: get Prasad to sign your meeting minutes.
• The uniQure reversal is the tradeable FDA question: after May Breakthrough Therapy designation and an agreed statistical analysis plan, November meeting minutes were consistent with FDA’s view that the Phase 1/2 study was not well-suited to support an AMT-130 BLA — “you’re supposed to believe that this May meeting was a sham,” per Paul. His read-through for single-arm gene-therapy pivotals (Lexeo, Rett, genetic cardiomyopathies): spring-era alignments may have reflected champions who have since left, but Paul remains cautiously optimistic that carefully prespecified agreements with current CBER decision-makers should be believed. Josh, a believer in the therapy, says the situation “threw me for a loop.”
• Capricor’s positive Phase 3 in DMD sent the stock up about 400%, and Brian’s approval logic is blunt: “this is the first therapy to hit cleanly in an RCT in this population — how do you not approve it?” Josh frames a higher data bar for therapies whose mechanisms are not well understood: allogeneic cardiosphere-derived cells may need replication in a well-controlled trial, and Capricor appears to have supplied it. Graig describes consistent signals across trials and flags the Nippon Shinyaku deal’s unusually rich 30%–50% tiered royalty, with FDA review likely next year.
• Paul’s contrarian read on Bristol’s Cobenfy ADEPT-2 delay: the update “reads bad” — site irregularities, added patients, readout pushed to the end of 2026 — yet “the probability of success of this trial arguably goes up versus how we felt a week ago.” An IDMC or similar group reviewed data and recommended enlarging the sample rather than stopping, and both Bristol and MapLight (+20%) rallied; the real risk in Alzheimer’s psychosis remains therapeutic index: three-times-daily dosing, a food effect, and a side-effect-sensitive population.
• Praxis cleared its scariest hurdle — a positive pre-NDA meeting despite having overridden a DMC futility recommendation — and plans to file ulixacaltamide for essential tremor in early 2026, with a second epilepsy NDA potentially following an interim stop for overwhelming efficacy. Against the investor concerns Josh raised about the futility flip and dropout imbalance, Brian’s response is that the point estimate was reportedly unchanged — the interim had only a little more than 40% of ultimate enrollment and was massively underpowered — and “they basically had the full effect size at week four,” before most dropouts. At a little above $4 billion market cap, ET peak “could be something much more substantial.”
• Janux’s selective disclosure of its PSMA×CD3 masked T-cell engager data — justified by management as protection against fast-following China competitors — halved the stock to near cash, and Paul isn’t buying the defense: “usually opacity is discounted… most good spaces are competitive.” Josh warns this secrecy dynamic may recur as China biotech develops, while Graig says it makes evaluating data quality and advising clients harder. Josh also flags Otsuka’s Voyxact for IgA nephropathy, priced near $400,000 a year, as the kind of ultra-premium launch that could eventually reignite the pricing debate that ends the rally; Graig tees up Terns Pharmaceuticals’ potentially best-in-class CML data at ASH, with the stock up about 250% since abstracts.

Deep dive

1. Everyone’s bullish on 2026 — the sector finally graduated from dreams to launches

• Paul’s framing, borrowed approvingly from the Canaccord biotech team: today’s excitement is “not propped up on a dream” — Phase 2/3 winners are becoming commercial realities and young companies like Intra-Cellular Therapies and Madrigal “grow up.” Add minimal drug-pricing threat, more M&A, and an XBI nowhere near its highs: “I’m not worried about us entering into another bear market imminently.”
• Graig’s baseball analogy: “we’re like in the third inning of what this rally is looking like,” though he’s “a little concerned that it’s gone up so far and so quickly” and would prefer gradual — while predicting an “incredibly positive and bullish” JPMorgan next month.
• Brian’s counter to the too-fast worry: it fell just as fast — four years of “horrific divergence between the XBI and the S&P.” On a six-month chart XBI looks like “bubble territory”; on five years it looks cheap, and large-cap pharma still trades at about 15 times. He quotes Rod Wong: “unfortunately, the FDA is now the major remaining source of policy uncertainty.”
• Brian’s one macro caveat: interest rates. XBI is “very, very much tied” to Fed rates given the long investment cycle, and the inflation fight isn’t settled.
• Josh’s structural point: “there’s almost no bigger red flag of an industry when stocks can’t perform when they go from that hope-and-dream phase into the execution launch phase” — the sector figured out the launch algorithm, with premium pricing one component. Graig adds that disciplined spending and capital allocation have helped companies vault to profitability instead of avoiding it.

2. The IPO wager: 15 vs. 50, with a preclinical tripwire

• Paul, reflecting on three 2025 IPOs he was involved with — Sionna, MapLight, and LB Pharma — sees mid-to-late-stage companies asking “are we ready to be a public company?” Nothing like 2020, when S-1s listed IND acceptance as a catalyst. Brian adds that deep-pocketed public investors doing private deals have let companies “better incubate themselves and be more than just one-drug stories.”
• The numbers were fluid: Josh initially says 15, Paul echoes 15; Josh then puts the over-under closer to 50, Graig says 50, Brian takes the over, and Paul later says he may be under 50 while hoping for the 40s and 50s, perhaps even 60.
• Graig’s discipline thesis: the public market has never sustainably analyzed preclinical companies — every cycle peak it tries, and “it’s total FOMO… it winds up being the signal of the crash.” Current IPOs are development-stage assets the market can evaluate, so “we’re in a safe place”; preclinical deals will probably return eventually, but not necessarily in 2026 or 2027.
• Paul’s memory of the 2020–22 excess: four of his covered names launched IPOs on the same day and he had to remind himself which company wanted to discuss what — “that was the heyday… it got silly.” He expects a strong 2026 could set up a healthier 2027, with the whole system more disciplined after the XBI’s collapse from roughly 174.

3. Pazdur’s exit and the vaccine flashpoint: “the last adult in the room” is gone

• Brian’s standing call since January: “expect chaos.” The institutional memory is gone — nearly every 20–30-year career leader has left — capped by Rick Pazdur leaving roughly three weeks after taking CDER, following George Tidmarsh’s brief, controversial stint. Media reports point to Vinay Prasad’s controversial internal emails and Pazdur being denied hiring leeway. Interim CDER chief Tracy Beth Høeg, previously Marty Makary’s chief adviser, is credited with the analysis behind the 10-childhood-COVID-vaccine-deaths claim — “more along the lines of this new MAGA-RFK mentality than the old guard.”
• Brian’s warning on Prasad: “he says, ‘Oh, I’m not micromanaging things,’ but all the signals are that he is inserting himself into decisions all over the place.” His practical advice for anyone facing a CBER approval: “I would want him to write or sign any minutes” — though Paul notes nobody he’s spoken to has even had a meeting with Prasad in the room.
• Paul’s lament, “as someone who has venerated the FDA”: “I’ve never seen an FDA leader mention a president’s name in a memo… Pazdur was under a number of different Republicans and Democrats and that has never mattered — to see that matter is kind of a bummer.” He also can’t square the COVID-vaccine obsession: Operation Warp Speed “was probably one of the biggest achievements of the Trump administration.”
• Brian’s theory of why vaccines became the flashpoint: they touch every healthy person in the country, carry real risks such as myocarditis, and imperfect science meets social-media pulpits. His coincidental hedge: passive-immunity players like Cidara and Invivyd could deliver monoclonal protective immunity “without some of the baggage.”

4. uniQure and the single-arm question: was the May alignment “a sham”?

• The facts per Paul: this week’s meeting minutes were consistent with the November 3 view that the Phase 1/2 study was not well-suited to support an AMT-130 BLA in Huntington’s — despite May Breakthrough Therapy designation and an agreed statistical analysis plan for the year-three data read out in September. “You’re supposed to believe that this May meeting was a sham, right?” Company-specific outs include following patients longer, adding those short of three years, or reaching a compromise — “who knows.”
• Paul’s hypothesis is that spring alignments were struck when Nicole Verdun and other since-departed CBER seniors may have been champions. His net position remains “cautiously optimistic that if you’re prespecifying everything and you have an agreement, we should believe it,” while acknowledging that this does not establish the efficacy bar. Josh, a longtime believer in the therapy, says it “threw me for a loop.” Graig flags the delay despite Breakthrough designation: “what’s Breakthrough for? I thought you were supposed to have a bat phone.”
• Brian’s bar for the next crop — Lexeo, Rett, and genetic cardiomyopathies — echoing Prasad: “in the case of Parachute, you don’t need a 500-patient placebo-controlled study.” Endpoints that don’t occur on placebo — a 60% RECIST response in relapsed/refractory oncology “is not chance.” The danger zone is slowly progressive diseases with variable functional endpoints versus database-selected natural-history controls: “I don’t know how to realistically interpret” flat-versus-10%-decline comparisons.
• Graig’s read is that the facts are unchanged but their interpretation and regulatory application have shifted. Peter Marks was willing to approve drugs that might not work and withdraw them later, which Graig says may have been too lenient; if the uniQure data were definitive, “this wouldn’t even be a discussion point.” Josh separately notes that the spring was a transitory period with major leadership changes.

5. Capricor’s clean Phase 3 hit: mystery mechanism, higher bar — apparently cleared

• Capricor’s deramiocel, allogeneic cardiosphere-derived cells for DMD, — “probably one of the most controversial names of the year” — posted positive Phase 3 data; the stock ripped about 400% this week. Graig describes consistent signals across trials on cardiac function and peripheral muscle despite intense Twitter scrutiny of the statistics, and flags the Nippon Shinyaku partnership’s 30%–50% tiered royalty ahead of a likely 2026 FDA review.
• Josh’s regulatory framework: when biology and mechanism align a priori, smaller datasets may get the benefit of the doubt through the “plausible mechanism pathway”; when the mechanism is not well understood, a higher data bar is legitimate. The FDA demanded replication in a well-controlled trial, and the new data appear to provide it. Paul adds that many approved neuropsychiatric drugs have mechanisms that remain incompletely elucidated.
• Brian’s bottom line: “this is the first therapy to hit cleanly in an RCT in this population — how do you not approve it, understanding the other things that are approved?”

6. Cobenfy’s ADEPT-2 delay: bad on the surface, bullish underneath

• The news: irregularities at trial sites in the Alzheimer’s-psychosis (ADP) study; Bristol will add patients, exclude troubled sites, and push the readout to the end of 2026. Paul’s tell: “stocks are all about expectations” — the announcement “reads bad and yet Bristol went up, and MapLight, a recent IPO with a muscarinic, went up over 20%,” because the pre-existing fear was silence since the study completed in late July.
• Paul’s tea-leaf reading: an IDMC or similar group reviewed data and recommended adding patients to restore power, and Bristol engaged the FDA. One read is that the group saw some sort of signal. “My interpretation was that the probability of success arguably goes up versus how we felt a week ago.”
• The enduring question for muscarinics in ADP, from the ex-Cerevel/Karuna coverage: therapeutic index — three-times-daily dosing, a food effect to navigate, and a population more sensitive to psychiatric side effects.

7. Praxis passes the smell test — two NDAs potentially headed for review

• Brian’s setup: ulixacaltamide’s Phase 3 program in essential tremor — “pretty much a graveyard of failures” — was recommended stopped for futility, Praxis overrode the DMC, and the final data hit overwhelmingly across primary and secondary endpoints in two studies, consistent with Phase 2. The pre-NDA meeting was “very positive, encouraging” with no major asks; the company plans to file in early 2026. At a little above $4 billion market cap, the ET peak “could be something much more substantial.”
• Josh’s question captured the investor concerns: futility-to-overwhelming-success “doesn’t pass the smell test,” and there is a substantial drug-versus-placebo dropout imbalance. Brian’s response: the interim had only a little more than 40% of ultimate enrollment — “massively underpowered” — and management has indicated the point estimate did not change, only the variance; tipping-point, IPTW with MMRM imputation, and COVID-imputation analyses are mostly still positive with p-values below 0.05. “The thing that really sells it to me is they basically had the full effect size at week four,” before most dropouts.
• The second shot: one epilepsy drug in developmental and epileptic encephalopathies stopped at interim for overwhelming efficacy, with data at AES. Josh expects this could leave Praxis with two totally different neuro NDAs under FDA review next year, which Brian calls “pretty rare.”

8. Janux’s secrecy gambit, $400K pricing, and the ASH teaser

• Janux’s PSMA×CD3 masked T-cell engager update in prostate cancer “didn’t go so well” — the stock was cut in half and now trades close to cash. Josh relays management’s defense: after doing the hard platform learning, full disclosure would let competitors — increasingly China-based fast-followers — “draft off” its work. He expects more selective disclosure and warns investors may have to live with unsatisfied curiosity.
• Paul isn’t persuaded: only some management teams have “earned the right” to withhold detail, and “usually opacity is discounted as something that is negative… it’s a tough situation to use competitive dynamics as an excuse — most good spaces are competitive.” Graig adds that it makes evaluating data quality and advising clients harder.
• Closing flags: Otsuka’s Voyxact for IgA nephropathy is priced near $400,000 a year. Josh suspects ultra-premium launch pricing could eventually be “what’s going to end the rally,” though hopefully not for years. Graig’s ASH pick: Terns Pharmaceuticals, up about 250% since abstracts, presents Monday in Orlando with a potentially best-in-class CML asset — “obviously it’ll be dependent on data.”