Pfizer's Next-Gen Pneumonia Vaccine Beats Prevnar 20; BMS Deploys Anthropic AI

Pfizer's next-gen pneumonia vaccine puts up antibody responses eight to fifteen times higher than Prevnar 20 in infants — and the company is advancing straight to Phase 3.

BMS hands Anthropic the keys to a thirty-thousand-employee enterprise, the FDA fires another senior review leader, and twenty deaths in Japan put the spotlight on Amgen's Tavneos.

Welcome to The Pharma Closeout for Wednesday, May 20th, 2026. I'm Alex Mercer.

And I'm Maya Patel. Let's get into it.

Pfizer dropped Phase 2 data today on its experimental twenty-five-valent pneumococcal conjugate vaccine — 25vPnC — tested head-to-head against Prevnar 20 in infants. After the third dose, antibody levels came in approximately 8.8 times higher than Prevnar 20. After the fourth dose, that gap widened to roughly fifteen times. Five additional serotypes, twenty-five strains total, and Pfizer's already heading into Phase 3.

The fold-changes in serotype-specific IgG are striking. But immunogenicity margins in Phase 2 don't automatically close the gap to reductions in invasive pneumococcal disease — that's still the Phase 3 deliverable. What's harder to dispute is the strategic necessity. Prevnar 20 is a multi-billion-dollar revenue line. Merck's Vaxneuvance has been chipping at it. Pfizer needed a successor that makes the coverage superiority argument inarguable, and these margins do exactly that.

The five additional serotypes aren't marginal picks, either. They cover strains contributing to residual disease burden in well-vaccinated populations — the kind of argument that lands with advisory committees and reshapes formulary conversations.

Pfizer's real edge right now is tempo. They're defining what the next-generation standard looks like. Every month of Phase 3 enrollment is a month Merck can't gain meaningful pediatric share, because the goalpost is already moving.

That's the best case. But let's be honest — pediatric enrollment timelines are notoriously hard to compress, and the FDA will expect comprehensive safety data across all twenty-five serotypes. That's not a dataset you can rush. The execution risk is real.

It is. But Pfizer has run this exact regulatory pathway twice — Prevnar 13 and Prevnar 20. The bridging study infrastructure is well-established. The risk isn't regulatory novelty. It's whether Pfizer can move fast enough to keep the competitive window from narrowing.

And this all lands at a moment when the vaccine division needs the win. COVID franchise wound down, RSV market getting crowded — moving 25vPnC into Phase 3 is the clearest signal yet that Pfizer's vaccine platform still has a growth story beyond the pandemic era. Investors have been waiting for exactly this kind of pipeline signal from the vaccine unit, and today's readout delivers it.

If Phase 3 confirms what Phase 2 is showing, Pfizer carries the pneumococcal revenue line into the next decade. If it stumbles, that franchise is exposed at precisely the wrong time. The confirmatory data here is not optional — it's existential for the program.

Shifting to the deal side. The headline today is Bristol Myers Squibb announcing a strategic agreement with Anthropic to deploy Claude Enterprise as the shared intelligence platform across its entire global operation. That covers research, clinical development, manufacturing, commercial, and corporate functions — more than thirty thousand employees. Financial terms were not disclosed, but the scope is unlike anything we've seen from a single pharma-AI collaboration.

Every big pharma company has an AI announcement these days. Most are pilots scoped to one function. This is structurally different. BMS is positioning Claude as shared infrastructure across every division, and the agentic framing is the tell — this isn't a chatbot answering questions. It's a system designed to connect workflows, surface institutional knowledge, and take action across platforms. If it delivers at scale, every other pharma AI strategy gets benchmarked against this.

BMS isn't alone in making enterprise AI bets — their peers have been striking similar deals. But the all-function scope sets a new bar. On the acquisition front, Lilly closed today on Engage Biologics for up to 202 million dollars, adding a non-viral DNA delivery platform called Tethosome. Still preclinical, but it extends Lilly's growing genetic medicines footprint.

Non-viral delivery is the piece of gene therapy that still lacks a dominant solution — viral vectors create immunogenicity barriers that limit redosing. At 202 million for a preclinical platform, Lilly is buying optionality, not a finished product. Given the acquisition pace they've been setting all year, they're clearly comfortable stacking these bets and letting the science sort out which ones scale.

Moving through the rest of the tape. Gilead expanded its API supply partnership with Korea's Yuhan Corporation — a 140-million-dollar contract bringing cumulative orders over the past two years to roughly 632 billion won. That's Asia-Pacific supply chain diversification at real scale. In venture, cAMPfield Therapeutics closed a 180-million-dollar Series A for immunology and inflammation — one of the larger early-stage I&I rounds this year. Amgen announced a CFO transition, with Peter Griffith retiring and Dittrich coming in from Galderma, where he served as CFO after prior stints leading finance at Shire and Sulzer. And Roche signed a generic licensing deal with the Medicines Patent Pool for Xofluza, expanding access to the flu antiviral in lower-income countries.

The thread connecting today's deal tape is infrastructure. BMS in operational AI, Lilly in genetic medicine delivery, Gilead deepening its manufacturing base, Amgen refreshing financial leadership. None of these are pipeline fireworks — they're the foundation companies lay when they're planning for the next five years, not the next earnings call. And the 180 million in fresh I&I venture capital says the early-stage money is following the same long-horizon thesis.

On the regulatory front, more instability at the FDA. Høeg has been fired, and the agency has reshuffled leadership for two top drug review offices. That continues a CDER turnover pattern we've been tracking — and at this point, the churn itself is becoming the regulatory risk.

When review division leadership keeps cycling, institutional knowledge walks out and companies lose the informal channel intelligence that helps them read how the agency is thinking. It's not just one personnel change — it's the cumulative drag on predictability. And we're heading into a heavy stretch of PDUFA decisions and advisory committee meetings, which makes this exactly the wrong time for leadership vacuums.

Also worth noting on the regulatory front. ImmunityBio's supplemental BLA for Anktiva plus BCG in BCG-unresponsive papillary NMIBC has been accepted by the FDA, with a PDUFA decision expected in the coming months. The CIS indication Anktiva currently covers is real but limited in patient volume. Papillary tumors are where the numbers concentrate in BCG-unresponsive disease — this expansion is where Anktiva's commercial thesis either scales or stalls.

And a safety signal that deserves close attention. Twenty people have died in Japan after taking Amgen's Tavneos — avacopan — approved for ANCA-associated vasculitis. Reports point to serious liver dysfunction. Amgen's partner Kissei issued a safety notice urging prescribers to exercise caution. Amgen says it supports the drug, but a post-marketing signal at this magnitude typically sets regulatory proceedings in motion.

Hepatotoxicity signals at this frequency don't stay contained in one market. When Japanese regulators move toward label changes or prescribing restrictions — and the trajectory here suggests they will — the EMA and FDA reassess their own risk-benefit calculations. Prescribers everywhere start adjusting their behavior well before the formal label catches up. That's the commercial risk Amgen needs to get ahead of.

What to watch. Parabilis Medicines has publicly filed for an IPO after raising 800 million dollars privately and closing a 125-million-dollar partnership with Regeneron. If they price successfully, they'll be among the best-capitalized biotechs to hit the public market this year.

Relay Therapeutics is also targeting the market with a 175-million-dollar offering, backed by early data in rare vascular diseases that positions them against Novartis's Vijoice. Two clinical-stage biotechs testing the IPO window in the same stretch tells you biotech sentiment is holding stronger than the macro noise would suggest. If both price well, the pipeline of companies waiting behind them will move faster.

On the deal landscape, first-quarter biopharma M&A already hit 84 billion dollars per Reuters, and the pace isn't letting up. GHO Capital and CBC Group are merging to form what could be the largest healthcare-specialist private equity investor — a signal in itself. When capital pools at that scale consolidate, competition for quality targets intensifies and deal timelines compress. Assets that would have run a longer process a year ago are moving in weeks. If you're in business development, plan accordingly.

That's your Pharma Closeout for Wednesday. Pfizer's 25vPnC putting Merck on notice in the pneumococcal market, BMS making the broadest enterprise AI commitment we've seen in pharma, Lilly stacking another technology acquisition with Engage Bio, and the FDA's leadership churn casting a shadow over every pending filing. If this briefing saves you time, follow us on Spotify and drop a rating — it helps other pharma professionals find the show. We're back tomorrow.

Have a great evening, everyone. See you then.