Themes & takeaways — JPM 2026¶
Cross-day synthesis of the recurring threads, surprises, and named entities across Day 1, Day 2, Day 3, and Day 4.
Source caveat: Unlike the Nextflow vault (which has primary transcripts), JPM is invite-only — every claim here is one degree removed from the room, synthesized from trade-press reporting (STAT, Fierce Biotech/Pharma/Healthcare, Endpoints, BioPharma Dive, Evaluate, etc.) and primary company press releases.
The five threads¶
1. AI in pharma — from partnership to ownership¶
The single loudest theme. JPM26 marked a phase change from "we have an AI collab" to "we own the AI company":
- Lilly + NVIDIA $1B AI Co-Innovation Lab (Day 1) — largest pharma–AI infrastructure commitment ever announced at JPM; NVIDIA BioNeMo + Vera Rubin compute.
- AstraZeneca acquires Modella AI (Day 2) — described in trade press as the first big-pharma acquisition of an AI firm; multi-modal foundation models for oncology.
- Pfizer announced plans to "embed AI across the business" (Day 2).
- Waystar projected RCM TAM expanding from $20–25B to $55B with agentic AI (Day 1).
The closing-week framing in Alex Hogan's STATus Report: AI is "table stakes" — the conversation has shifted from will you do AI to show us the ROI and the compliance story (echoed by Hogan Lovells on AI compliance + BIOSECURE).
Counterweight: Regeneron's George Yancopoulos publicly questioned current AI applications in medicine on Day 4 — calling for careful oversight. The conference's most prominent scientific voice pushing back on AI maximalism.
2. Licensing displacing M&A¶
JPM26 had ~$8.3B in announced deals (Pharmaceutical Technology) but no pharma megamerger. The pattern:
- AbbVie + RemeGen $5.6B PD-1×VEGF licensing (Day 1) — biggest single deal headline of the week.
- Novartis + SciNeuro $1.7B Alzheimer's antibody pact (Day 1).
- Novartis + Zonsen PepLib $50M radioligand license (Day 2).
- AZ + Modella AI — undisclosed terms, tuck-in (Day 2).
- Boston Scientific + Penumbra $14.5B (Day 4) — the one megadeal, and it's medtech, not biopharma.
- Pre-JPM Lilly + Ventyx $1B+ acquisition cited as the largest biopharma deal near the conference window.
Fierce Biotech canonical framing: "All the meetings, none of the megadeals." One attendee told Fierce some companies appear to be holding announcements for the TD Cowen Health Care Conference in March.
3. China is the story¶
Two distinct angles converged:
- China-origin assets in big-pharma deals — RC148 (RemeGen → AbbVie, Day 1), brain-shuttle Alzheimer's antibodies (SciNeuro → Novartis, Day 1), peptide radioligand (Zonsen PepLib → Novartis, Day 2).
- Chinese pharma stage time — Hengrui presented on Day 4 with >$3.2B 2025 sales through Q3, 21 NDA/BLA approvals, $1.5B HK IPO, ~$16B cumulative out-licensing (incl. GSK partnership). Kelun-Biotech presented bispecific/multispecific conjugates.
BioPharma Dive's framing: US biotech's "Sputnik moment." Novartis CMO Shreeram Aradhye (Day 3) cautioned against simple speed comparisons — different ecosystem, trial-enrollment economics, payer structure — but did not dispute the trend.
4. Obesity = the gravitational center¶
Every major pharma had to address it:
- Pfizer (Day 1, Day 2) — $10B Metsera acquisition anchors 10 Phase 3 starts in 2026; Bourla compared cash-pay obesity demand to Viagra.
- Amgen — MariTide Phase 2 reveal (Day 1, reprise Day 3): up to 20% mean weight loss in obesity-only, 17% in obesity + T2D; potential first once-monthly T2D therapy; six global Phase 3 underway.
- Novo Nordisk (Day 2, Day 3) — oral Wegovy ~3,100 patients in first week since Jan 5 launch; Doustdar publicly hungry for M&A "with intention" after losing Metsera bid; pulling back from non-obesity/diabetes indications.
- Roche (Day 2) — teased entry into obesity as one of three growth pillars.
- The competition has shifted from drug-data tournament to commercial/distribution warfare: oral GLP-1s, monthly dosing, cash-pay channels.
5. Policy / regulatory headwinds as constant background noise¶
Threaded through every day:
- Sanofi Day 3 — RSV recommendation narrowing on Beyfortus, recent FDA rejection, MFN pricing arrangement skirting tariff threats; Sanofi singled out the Dec 24, 2025 CRL on tolebrutinib for "skipped advisory committee" reasons.
- Pazdur Day 2 — first public remarks since departing FDA, warning on politicization and voucher-program transparency.
- Mehmet Oz Day 2 — fielded hospital-exec questions on Medicaid and vaccines off-site.
- Trump MFN pricing — Yancopoulos (Day 4) explained why Regeneron was the only one of 17 companies receiving presidential pricing letters that hadn't yet struck a deal.
- AI compliance + BIOSECURE — Hogan Lovells flagged as top device-supply-chain concerns.
Surprises worth flagging¶
| Surprise | Day | Why it stood out |
|---|---|---|
| $1B Lilly–NVIDIA AI lab | 1 | Largest pharma–AI infrastructure commitment ever announced at JPM |
| AZ buys Modella AI outright | 2 | First big-pharma acquisition of an AI firm; partnership → ownership shift |
| Boston Scientific / Penumbra $14.5B | 4 | Only true megadeal of the week, and it's medtech not biopharma |
| AbbVie / RemeGen $5.6B PD-1×VEGF | 1 | Largest single biopharma deal of the week; another China-origin asset |
| Oral Wegovy: 3,100 patients in first week | 2 | Real-world demand validation for the first oral GLP-1 obesity drug in the U.S. |
| Hengrui's $16B cumulative out-licensing | 4 | Concrete number behind the "China is the story" framing |
| Yancopoulos publicly skeptical of AI in medicine | 4 | Highest-profile science voice pushing back on AI hype on closing day |
| Pfizer compares obesity demand to Viagra | 1 | Bourla's framing of the cash-pay obesity opportunity |
| Regeneron — large M&A "in most cases dramatically overpays" | 3 | Public capital-discipline posture as Dupixent LOE looms (2031) |
| "Slowest JPM for news in a long time" | 1 | Brad Loncar's Day 1 framing — set the tone for the week |
| Pazdur's first post-FDA public remarks | 2 | Insider warning on agency politicization |
| Sanofi tolebrutinib CRL with skipped AdCom | 4-week | Concrete example of FDA dysfunction cited at JPM |
Tools, assets & entities — quick index¶
Where each surfaced:
| Item | Day 1 | Day 2 | Day 3 | Day 4 |
|---|---|---|---|---|
| NVIDIA BioNeMo / Vera Rubin | ✓ | |||
| Modella AI (multi-modal FMs) | ✓ | |||
| Lilly Orforglipron (delayed) | ✓ | |||
| Pfizer Metsera / MET-097i | ✓ | ✓ | ||
| Amgen MariTide (peptibody) | ✓ | ✓ | ||
| Novo Wegovy / oral Wegovy | ✓ | ✓ | ✓ | |
| AbbVie RC148 (PD-1×VEGF, RemeGen) | ✓ | |||
| Summit ivonescimab (PD-1×VEGF, ref'd) | ✓ | |||
| BMS pumitamig (PD-L1×VEGF, BioNTech) | ✓ | ✓ | ||
| BMS Cobenfy / milvexian / admilparant | ✓ | |||
| AZ baxdrostat / camizestrant / gefurulimab | ✓ | |||
| AZ ADCs / cell therapies | ✓ | |||
| Roche giredestrant (SERD) | ✓ | |||
| Regeneron Dupixent franchise | ✓ | ✓ | ||
| Sanofi Beyfortus (RSV) | ✓ | |||
| Sanofi tolebrutinib (CRL) | ✓ | |||
| Biogen high-dose Spinraza (EC nod) | ✓ | |||
| BridgeBio Attruby (acoramidis) | ✓ | |||
| Sarepta Elevidys (DMD) | ✓ | ✓ | ||
| Agios Acvezmi (thalassemia, FDA Dec 23) | ✓ | |||
| BioNTech HER2 ADC / Gotistobart / BNT113 | ✓ | |||
| Pfizer/BioNTech LP.8.1 COVID vaccine | ✓ | |||
| SciNeuro brain-shuttle Alzheimer's | ✓ | |||
| Zonsen PepLib radioligand (China) | ✓ | |||
| Novartis radioligand franchise | ✓ | |||
| Kite anito-cel (Arcellx, BCMA CAR-T) | ✓ | |||
| Legend / J&J Carvykti (BCMA CAR-T) | ✓ | |||
| Inventiva lanifibranor (MASH) | ✓ | |||
| Precigen PAPZIMEOS | ✓ | |||
| Hengrui (Hong Kong IPO, GSK pact) | ✓ | |||
| Kelun-Biotech (bispecifics) | ✓ | |||
| BSx + Penumbra (medtech megadeal) | ✓ | |||
| Aktis / Eikon / Veradermics IPOs | ✓ | |||
| Lonza Vacaville (CDMO) | ✓ | ✓ | ||
| Waystar / Iodine (RCM agentic AI) | ✓ | |||
| Clover Health (MA, 1st GAAP profit) | ✓ | |||
| Alignment Healthcare (MA counter-narrative) | ✓ |
Companies / institutions named: Eli Lilly, NVIDIA, AbbVie, RemeGen, Novartis, SciNeuro, Zonsen PepLib, BMS, Pfizer, Metsera, J&J, Regeneron, Biogen, HI-Bio, Moderna, Alnylam, argenx, BridgeBio, Sarepta, Summit, CRISPR Therapeutics, AstraZeneca, Modella AI, Novo Nordisk, GSK, Bayer, Roche, Gilead, Kite, Arcellx, Sandoz, Acadia, Agios, Travere, Mirador, Lonza, BioNTech, Sanofi, Legend Biotech, Boston Scientific, Penumbra, Clover Health, Alignment Healthcare, Inventiva, Precigen, Biocon, Oruka, Vicore, Atea, Corbus, Molecular Partners, Rakuten Medical, Hengrui, Kelun-Biotech, Aktis Oncology, Eikon Therapeutics, Veradermics, Sionna, Maze, Diamond Age (n/a), Waystar, Iodine, Advocate Health, MGB, HSBC Innovation Banking, OrbiMed, Atlas Venture, Takeda.
Standards / frameworks named: MFN (most-favored nation pricing), BIOSECURE, FDA voucher program, FDA plausible-mechanism pathway, HEDIS, GAAP, CCAR.
Predictions to watch¶
From the closing recaps and threaded through the keynotes:
- AI bake-off goes from collab to acquisition. Expect more big-pharma + AI-startup acquisitions (vs. partnerships) in 2026 following the AZ–Modella template.
- The TD Cowen Health Care Conference (March) may absorb the megadeals JPM26 didn't see — at least one Fierce attendee suggested companies were holding announcements.
- Up to ~20 biotech IPOs in 2026 if the Sionna/Maze/Eikon/Aktis/Veradermics windows hold (HSBC's Jonathan Norris). 2025 had <12 IPOs, the lowest since at least 2018.
- Obesity competition shifts to distribution. Novo's DTP/oral-Wegovy push, Amgen's once-monthly differentiation, Pfizer–Metsera's combos — the next 18 months are commercial, not data.
- Policy risk is structural, not transient. Vaccine politics, MFN pricing letters, BIOSECURE, FDA-process scrutiny — all four will continue weighing on biopharma narratives through 2026.
- China-origin licensing is now mainstream. Three of the four top deals at JPM26 had Chinese-origin assets; expect this share to grow.
Caveat¶
JPM is invite-only. There are no public transcripts. Every claim above is reconstructed from third-party trade-press reporting and primary company press releases. Where dating to a specific day was uncertain (some outlets bundle the week), the source notes that. Some headlines (notably the rumored Merck/Revolution Medicines $32B deal) stayed rumors through the week and are not represented above. Treat dollar figures and patient counts as best-effort transcription from public reporting, not official tallies.