Last Week Tonight in BioPharma: Week of June 29th, 2026
Anthropic wants to make drugs, a new "diva" in KRAS, Ipsen goes shopping, BridgeBio gets a "billy", and more!
Welcome back to Last Week Tonight in BioPharma (LWTB). Hope you had an enjoyable Fourth of July weekend. LWTB is here to catch you up on what you missed while you were away.
Roche’s divarasib became the latest RAS inhibitor to register impressive data (this time in NSCLC). Your favorite chatbot, Claude, became your competitor overnight, with Anthropic announcing that its new “Claude Science” offering will also be leveraged by the company to develop its own therapeutics. Plus, Euro-midcap Ipsen unloads $2.5B in M&A dollars in just three days.
All that and a couple more below!
If you subscribe to Big Pharma Sharma (BPS) for my free weekly Last Week Tonight in Biopharma series and you like what you read, please consider upgrading to paid to read my best work.
If you want to show your support another way, you can always leave a donation to the BPS beer and wine fund below
📡 PRESS RELEASE DECODER
What the press releases actually mean
Roche’s Divarasib Beats Lumakras and Adagrasib Head-to-Head in Phase 3 KRAS G12C NSCLC — Third-to-Market With Best-in-Class Data
📅 July 1, 2026 | 🏢 Roche / Genentech (RHHBY),Amgen(AMGN), Bristol Myers Squibb (BMY),RevolutionMedicines(RVMD) | 💊 divarasib (GDC-6036) vs sotorasib (Lumakras) and adagrasib (Krazati) | 🏷 Phase 3 Readout / KRAS / Second-Line NSCLC
Genentech reported that Krascendo-1 (NCT06497556, n=338), the global head-to-head Phase 3 trial of a KRAS G12C inhibitor against the two approved incumbents, hit both its primary and key secondary endpoints. Divarasib delivered statistically significant and clinically meaningful improvements in progression-free survival versus sotorasib or adagrasib, and hit overall survival at the interim analysis. Discrete efficacy numbers have not been disclosed and will come at an upcoming medical meeting. No new safety signals; the safety profile was consistent with prior divarasib data. The trial excluded patients who had previously received a KRAS G12C or pan-KRAS/RAS inhibitor.
Roche plans to submit divarasib for approval in second-line NSCLC in 2027, per its recently disclosed regulatory submission list. Roche has guided peak divarasib sales of CHF 1–2 billion (~$1.2–2.5B), a range Jefferies analysts have echoed for the 2L opportunity alone. Divarasib holds FDA Breakthrough Therapy Designation (2022) and Orphan Drug Designation (2026) in KRAS G12C NSCLC.
Two additional Phase 3 trials are ongoing: Krascendo-2 (divarasib + pembrolizumab vs chemo + pembro in 1L advanced KRAS G12C NSCLC, NCT06793215) and Krascendo-3 (adjuvant divarasib monotherapy vs immunotherapy or observation in resected stage II–IIIB KRAS G12C NSCLC, NCT07541170).
🧠 BPS Take: Lumakras was FDA-approved in 2021, Krazati in 2022. Full-year 2025 sales were ~$368M and ~$240M respectively — both meaningfully below launch-era analyst expectations. History says me-better targeted oncology drugs with better data tend to supplant incumbents. Osimertinib entered EGFR behind gefitinib and erlotinib, then moved to first-line and now generates >$6B/year while the first-gen agents lag behind. Alectinib entered ALK behind crizotinib and now holds roughly 60% of the market. Lorlatinib is doing the same to alectinib in CROWN. In molecularly defined NSCLC oncologists switch fast when the OS data shows there is a better drug available.
Divarasib fits that template cleanly. It won on OS in a head-to-head study against existing KRAS G12C inhibitors. Lumakras and Krazati are modestly successful commercial products, but have certainly never achieved the scope and scale once expected of them when they first launched and would seem to be easily displaced.
Two things worth watching before the read-throughs get overdone. First, the actual PFS and OS deltas — “statistically significant” at interim doesn’t tell you whether this is a 2-month or 6-month OS improvement. The absolute magnitude will dictate how easily divarasib displaces Lumakras and Krazati. Second, Krascendo-2 in first-line is where the CHF 2B high case lives; second-line is a $500M–$1B franchise on its own.
AstraZeneca’s Strensiq Successor Efzimfotase Alfa Hits Pediatric MULBERRY but Misses Adult HICKORY — Franchise Succession Story Now Depends on Salvaging the Bigger Half of the Market
📅 June 28-29, 2026 (ICCBH Montreal) | 🏢 AstraZeneca ($AZN) / Alexion | 💊 efzimfotase alfa (ALXN1850) vs. Strensiq (asfotase alfa) | 🏷 Phase 3 Full Data / Hypophosphatasia (HPP)
AZN presented full data at ICCBH from three Phase 3 trials of efzimfotase alfa, a Q2-week SC enzyme replacement therapy versus Strensiq’s 3-6x weekly dosing.
MULBERRY (pediatric naïve, n=29): Hit primary. RGI-C bone score median difference 1.67 vs placebo (p=0.0003); RSS median difference -1.00 (p=0.0008). CHESTNUT (pediatric switchers, n=43): Bone health maintained; comparable TEAE rate. HICKORY (adolescent/adult naïve, n=124): Missed primary — 6MWT no better than placebo. AZN attributed the miss to strong placebo performance and reported nominal significance in the pediatric-onset subgroup.
Safety: ISR rate 5x lower than Strensiq registrational trials; 98.8% ISR-free days. Commercial: Strensiq 2025 sales $1.7B. AZN peak guidance $3-5B, contingent on adult expansion (~80% of HPP TAM). Manufacturing cost “way under” Strensiq per CEO Marc Dunoyer. Strensiq US patent expiry ~2028, EU 2032. Regulatory filings planned; no BLA date confirmed.
🧠 BPS Take: This isn’t the readout AZN needed. Efzimfotase alfa was designed to grow a $1.7B franchise to $3-5B on the back of adult expansion, and HICKORY missed. Parsing the data in these sorts of rare diseases goes over much easier with regulators than in larger indications, so AZN will still push to get adults on the approved label, and the pediatric-onset subgroup nominal significance gives them the argument to make. How successful this effort is will determine whether this franchise stays in the $2B range or starts to move up to the $3-5B projections.
Strensiq’s 2028 US patent cliff means AZN has to defend or replace regardless, Q2-week dosing is a patient-experience win versus 3-6x weekly injections, lower manufacturing cost helps margin and formulary negotiations, and CHESTNUT means AZN can hold the pediatric base while payer contracts reset. This is a real defense against the most near-term threat, which is post-2028 biosimilar Strensiq. Novel competitor drugs are still in early stage and ways away from approval.
🌐 CONNECTING THE DOTS
When the outside world meets biopharma
Anthropic Launches Claude Science and Announces Its Own Drug Programs for Rare and Neglected Diseases
📅 June 19, 2026 (Jumper hire) & July 1, 2026 (Claude Science launch) | 🏢 Anthropic (private, pre-IPO); | 💊 Undisclosed rare/neglected disease programs; PKU shown as demo | 🏷 AI / Platform / Strategic Positioning
At an event for pharma executives, biotech founders, and researchers on Tuesday, Anthropic launched Claude Science, positioned alongside Claude Code and Claude Cowork as a third flagship product. It integrates 60+ curated skills across genomics, single-cell, proteomics, structural biology, and cheminformatics; runs on the researcher’s own HPC or laptop; connects natively to NVIDIA BioNeMo (Evo 2, Boltz-2, OpenFold3); generates reproducible figures and manuscripts with auditable code trails; and includes a reviewer agent that checks citations and calculations. Beta live for all paid Claude subscribers.
At the same event, Anthropic announced explicitly that it will use Claude Science to pursue its own drug candidates for rare and neglected diseases. Eric Kauderer-Abrams, Anthropic’s head of life sciences, framed drug discovery as central to the company’s mission, stating: “Our mission is to develop AI that serves humanity’s long-term well-being, and we believe that by far the greatest opportunity to do that is in the life sciences.” Alexander Tarashansky, who led Claude Science’s development, demonstrated the system autonomously identifying drug candidates for phenylketonuria (PKU) live on stage. Kauderer-Abrams previously told CNBC in October that Anthropic wants “a meaningful percentage of all of the life science work in the world to run on Claude, in the same way that that happens today with coding.”
Anthropic’s paid pharma customer base already includes BMS (Claude deployed to 30,000+ employees), Sanofi (workforce-wide), Novo Nordisk (NovoScribe CSR automation), and named partnerships across most of the top 15 pharmas. The company is approaching its first profitable quarter with an IPO expected later this year.
🧠 BPS Take: When I mapped 27 frontier-AI/Pharma partnerships in May, I put Anthropic in the “science partner” bucket because the context window handles 200-page protocols and the safety/compliance posture sells cleanly into a regulated industry. The expanded Claude Science toolkit can certainly be a major value add for drug developers and enables Claude to integrate more deeply with its BioPharma customers.
What I did not expect this soon was Anthropic explicitly announcing that it will be developing its own drug programs. Anthropic’s interest in doing its own biology work was clear two months ago when they bought Coefficient Bio, a sub-ten-person stealth AI biotech founded in 2025, for $400M in stock. Its founding team came from Genentech’s Prescient Design (Nathan Frey, Samuel Stanton), Evozyne (CEO Aris Theologis), and Roivant (Joyce Hong)— basically an acquihire play. Then on June 19th they hired John Jumper, nobel laureate who co-created AlphaFold with DeepMind’s Sir Demis Hassabis. This Claude Science unveil is step three.
Now it remains unclear what this means in terms of commitment level. Are they simply discovering new programs and will license them out to more experienced players to take them the rest of the way? Are they going to develop them completely themselves? Will they just focus on neglected disease or will they expand into more competitive areas? Does Anthropic see itself as just the service/enablement layer in drug development or do they really want to go so far as to commercialize their own drugs?
The neglected-disease framing is smart positioning to start frankly: humanitarian cover, low regulatory scrutiny, no/low direct conflict with any named customer’s asset. Anthropic can even spin the drug unit out as its own rare disease focused biotech, perhaps even as a public benefit corporation or other less capitalist sounding structure. Focus on the thousands of ultra rare diseases with high unmet need but challenging economics, and do this all as a “gift to society” while serving as a demo of Claude Science in the process. If it fails to do anything, then they can let it die on the vine and they still have the underlying service layer with all their pharma partners keeping them in the fold on AI for drug development. If it succeeds, it is true end-to-end proof of Claude Science’s benefit.
For biopharma customers this creates “coopetition” with information asymmetry. Your AI vendor could now be reading across your enterprise data and workflows, arming itself and its pipeline to compete with you down the road. Certainly Anthropic will say that firewalls will be in place and that they would never use a client’s secure data, but we’ve seen tech companies go back on that promise many times before. If you’re integrated with the Anthropic suite now, you have to think critically about how you protect your data, workflows, and internal knowledge. Perhaps GSKs decision to build their own AI capabilities from the ground up, vs partnering with one of the major labs, looks like a smarter decision in retrospect.
💰 FOLLOW THE MONEY
Deals, dollars, and what they signal
Ipsen Spends ~$2.5B in Biobucks Across Two Deals in Three Days — Kartos ($450M Upfront) for Myelofibrosis, Memo ($227M Upfront) for BK Virus in Kidney Transplant
📅 June 27 & June 29, 2026 | 🏢 Ipsen ($IPSEY), Kartos Therapeutics (private), Memo Therapeutics (private) | 💊 navtemadlin (oral MDM2 inhibitor); potravitug (anti-BK polyomavirus mAb) | 🏷 M&A / Late-Stage Oncology + Rare Disease
Ipsen announced two acquisitions in three business days. Kartos (June 27):$450M upfront + up to $1.3B in regulatory and sales milestones. Adds navtemadlin, an oral MDM2 inhibitor being run in the global Phase 3 POIESIS trial (>600 patients, >250 sites) as add-on to ruxolitinib in intermediate/high-risk TP53wt myelofibrosis patients with suboptimal ruxolitinib response. Top-line data expected 2027; potential launch 2028. Prior Phase 1b/2 (n=19 suboptimal responders) showed 42% ≥25% SVR, 32% ≥35% SVR, 32% TSS50, and disease-modifying signals (71% ≥20% VAF reduction, 57% BMF improvement by Grade ≥1).
Memo (June 29): €200M (~$227M) upfront + >€500M in milestones. Adds potravitug, a first-in-class anti-BK polyomavirus antibody entering pivotal Phase 2/3 later this year in kidney transplant patients with BK virus-associated nephropathy. FDA fast-track and EU orphan designations. Close expected Q3 2026 for both. Ipsen’s 2025 revenue mix: €2.5B oncology, €747M neuroscience, €384M rare disease.
🧠 BPS Take: Two deals in three days from a mid-cap European specialty pharma is a statement of intent, and the deals are internally consistent with Ipsen’s stated three-pillar strategy (oncology, rare, neuro). But the Kartos deal is the more interesting one and it deserves scrutiny.
I wrote about the myelofibrosis landscape in May and flagged navtemadlin as the more credible of the two MDM2 programs left standing. Kartos smartly re-designed POIESIS to enroll suboptimal ruxolitinib responders after the JAK-refractory BOREAS study delivered mixed results. Phase 1b/2 subset showed a disease-modifying signal in a class that has struggled to show anything beyond spleen shrinkage, so Ipsen is seemingly buying a real drug here. I just question whether $450M upfront and $1.75B all-in for an asset with a 2028 potential launch, in a market where Incyte’s Jakafi generic entry hits December 2028 and CALR-directed antibodies from Incyte and J&J are moving toward Phase 3 is worth it.
BridgeBio Raises $1 Billion in Preferred Equity From Sixth Street and KKR: A Sophisticated Capital Markets Move That Is Not a Distress Financing
📅 July 1, 2026 | 🏢 BridgeBio Pharma ( BBIO 0.00%↑ ) / Sixth Street Partners / KKR HealthCare Royalty | 💊 ATTRUBY (acoramidis) + three pipeline launch assets | 🏷 Preferred Equity Financing
BridgeBio Pharma ($BBIO) raised $1 billion in preferred equity on July 1, 2026. Sixth Street Partners led the transaction with $800 million, and KKR HealthCare Royalty provided $133.9 million at close. The initial conversion price was set at approximately $138 per share, representing a premium of more than 100 percent to BridgeBio’s 30-day volume-weighted average price. The structure includes royalty features on revenue from ATTRUBY (acoramidis) and BridgeBio’s three additional pipeline drug launches expected over the coming years.
ATTRUBY targets transthyretin amyloid cardiomyopathy (ATTR-CM), competing primarily with Pfizer’s ($PFE) tafamidis (VYNDAMAX / VYNDAQEL), which generated approximately $4.3 billion in 2025 revenue. Alnylam’s ($ALNY) patisiran and vutrisiran represent RNA-based alternatives in the ATTR space with a distinct mechanism. BridgeBio has reported direct comparison analyses showing ATTRUBY’s superiority to tafamidis.
🧠 BPS Take: There is a lot of finance mumbo jumbo in the press release, so let me try and simplify this a bit. BridgeBio is getting $1B in funding that they’re going to use to ramp up the commercial scale of ATTRUBY and support the launches of three more of its products. The equity comes in the form of “convertible preferred equity”. Instead of selling a bunch of shares to raise the $1B and diluting its existing shareholders right away, Bridge is offering the new investors a 7% annual dividend plus the option to convert their entire preferred investment into common stock at a set price. If the investors want to, they can trade their preferred shares for common stock at $138 per share, double what Bridge’s stock has been trading at (and later at $153). So for them, they are collecting a cool $70M per year while still holding a $1B in equity in Bridge no matter which way the company stock price swings.
So in a sense, while this is structured as permanent equity, it acts similarly to a royalty partnership. BridgeBio gets the $1B scale-up capital fueled by the massive market potential of ATTRUBY and its upcoming pipeline. If those do well in the market, they are betting their stock probably doubles (if not more) within five years and those new shareholders decide to say goodbye to the dividend in favor of shares. By that time the pie has grown so big that the existing investors probably don’t care that they’re getting diluted, because their position is worth 2x or more what it was before.
The structure aligns the new investors to BridgeBio’s long-term success, while also protecting existing investors. The press release reads as a major vote of confidence in ATTRUBY’s commercial prospects in the ATTR-CM market vs. Pfizer’s , which is one of the fastest-growing cardiovascular rare disease markets in pharma, and Pfizer’s ($PFE) $4.3 billion VYNDAMAX franchise. We saw Apogee do a similar sort of deal with Blackrock before they got bought by AbbVie and we could see more of these sorts of structures that allow late-stage biotechs to raise non-dilutive funds to fund commercial launches, while still keeping acquisition options open.
Back next week with more BioPharma strategy takes! Share this with a friend or colleague if you found it helpful.