Pfizer’s oncology R&D strategy: Jeff Legos on speed, breadth and novel combinations

Oncology projects currently take up about 40% of Pfizer’s overall R&D budget. Behind this construction is Jeff Legos, Ph.D., who became the firm’s chief oncology officer last year following a high-profile move from Novartis.

After steering positive key pivotal readouts for a trio of Novartis’ blockbusters—Kisqali, Pluvicto and Scemblix—Legos is now tasked with fleshing out Pfizer’s $43 billion Seagen acquisition into an antibody-drug conjugate empire while accelerating a PD-1xVEGF bispecific licensed from 3SBio toward 2028 readouts.

For the closely watched PD-1xVEGF competition, Legos’ strategy for Pfizer is anchored by three pillars: speed, breadth and novel combinations.

“We are operating under very expedited timelines in terms of the first wave of studies,” Legos said in an interview with Fierce Biotech.
 

Slamming the gas

Pfizer cleared regulatory requirements and completed the licensing agreement for 3SBio’s PD-1xVEGF bispecific toward the end of July 2025. For an upfront payment of $1.25 billion and potential milestones of up to $4.8 billion, Pfizer gained ex-China rights to SSGJ-707, now also known as PF-08634404.

By the end of 2025, Pfizer would have had regulatory discussions with the FDA and launched seven clinical trials, including three phase 3 studies, for the PD-1xVEGF drug. And the company plans to start another five studies by June this year. Normally, it would take about a year from clinical ideation to a study start, Legos noted.

As to breadth, Pfizer is examining all of the existing indications of Merck & Co.’s Keytruda, trying to figure out where the PD-1xVEGF inhibitor can go, with additional considerations paid to disease prevalence and unmet medical need, Legos said. 

The urgency to quickly expand PF-08634404’s clinical program highlights a thrilling race to develop a successor to displace Keytruda, the current immuno-oncology king. To justify its price tag, Pfizer has talked about how PF-08634404 is differentiated from other candidates from a chemistry and biological mechanism perspective. But the phase 3 programs that Pfizer has outlined so far appear similar—if not identical—to what its rivals Summit Therapeutics/Akeso and BioNTech/Bristol Myers Squibb are undertaking for their PD(L)1xVEGF candidates.

Some of the same settings where Pfizer is running phase 3 include a first-line non-small cell lung cancer study pitting PF-08634404 against Keytruda in their respective combinations with chemotherapy across both squamous and nonsquamous histologies, against Roche’s Tecentriq and chemo in first-line extensive-stage small cell lung cancer, and phase 3 tests in first-line gastric cancer and first-line colorectal cancer.

PF-08634404 is a tetravalent antibody with four binding domains, two each for PD-1 and VEGF. The structure provides more robust cooperative binding and create a so-called “daisy chain” effect that could bring more cancer-killing T cells to the tumor microenvironment, Legos explained.

“Despite that [scientific rationale], it does come down to clinical data, and that’s why we are required to test it in these very well-defined patient populations,” Legos said.

Even though Keytruda has established itself as the standard of care in first-line NSCLC, only about 30% of patients benefit from immunotherapy, and only 30% of those patients benefitting experience long-term, durable responses, Legos observed. And in colorectal cancer, immunotherapy remains less effective in microsatellite-stable tumors. The unmet need prompted Pfizer to study whether PF-08634404 can improve patients’ outcomes, he said. 
 

Can PD-1xVEGF live up to its promise?

Despite billions of dollars being poured into the PD-(L)1xVEGF space, the main criticism against the class is its seemingly weak ability to turn slowing tumor progression into prolonging patients’ lives. Summit and Akeso’s first-in-class ivonescimab recently showed a 26% death risk reduction in Chinese patients with previously treated EGFR-mutated NSCLC at the final analysis of the phase 3 HARMONi-A trial. But when it comes to the global HARMONi trial, the overall survival data failed to meet statistical significance.

Legos argues that those data shouldn’t discount the broader potential of the PD-1xVEGF mechanism. Neither PD-1 nor VEGF inhibitor works very well separately in EGFR-mutated NSCLC, so it’s “not so surprising that the overall survival data is more borderline,” he said.

And the problem may not be PD-1xVEGF anyway.

“When you design a study with smaller patient numbers, in order to meet predefined endpoint around progression-free survival, you may not have enough power to sufficiently demonstrate the overall survival benefit,” Legos said.

For PF-08634404, Pfizer plans to make overall survival at least a primary endpoint in all its phase 3 trials.

“It comes down to the intention of the trial and your statistical analysis plan up front, and then ultimately the patient population that we are studying,” he said.

Answers to the question of whether Pfizer’s PD-1xVEGF drug can be the next Keytruda will start to emerge in 2028, when the company expects readouts from the first wave of its phase 3 programs.

However, the landscape of cancer care is forever evolving, and PD-(L)1xVEGF is not the only force that has its eyes on the prize. Notably, Merck is investing aggressively in its Kelun Biotech-partnered TROP2 antibody-drug conjugate sacituzumab tirumotecan (sac-TMT) but has been quiet on its own PD-1xVEGF bispecific after licensing the asset from China’s LaNova Medicines—now part of Sino Biopharm—in late 2024.

“I’m not sure it’s an either-or, because they’re doing two very different things,” Legos said of a potential showdown between PD-1xVEGF bispecifics and TROP2 ADCs.

ADCs, as targeted missiles to deliver a cancer-killing payload, can achieve immediate cytotoxic cell death.

“But frankly, we’re not just looking for improvements in response rates,” Legos continued. “Without engaging the immune system, I think the benefit of just a single modality may be limited in terms of its fullest potential.”

Merck is already pairing sac-TMT with Keytruda in phase 3 trials in multiple tumor settings. As to whether a TROP2 ADC can be combined with a PD-1xVEGF bispecific, “those experiments would have to be done,” Legos said.
 

Expanding Pfizer’s oncology pipeline

Combination with other novel agents, including ADCs, is already on Legos’ agenda, forming the third pillar of his strategy in hopes of truly differentiating PF-08634404 from its peers. One of the phase 3 studies being planned pairs PF-08634404 with Pfizer’s Astellas-partnered ADC Padcev in first-line bladder cancer, a setting where the Padcev-Keytruda combo has recorded standing ovation-worthy data. 

Following the Seagen buy, questions persist regarding Pfizer’s ability to build a sustainable ADC pipeline while avoiding the path of Gilead Sciences, which has not produced a new ADC beyond the TROP2-directed Trodelvy in the six years since its $21 billion acquisition of Immunomedics. Pfizer discontinued a B7-H4 ADC a year ago and has repeatedly pared back efforts around its RemeGen-partnered HER2 ADC disitamab vedotin.

Legos remains optimistic that new ADCs are coming out of Pfizer. Among eight key readouts that Pfizer expects this year across the group, sigvotatug vedotin, an ADC targeting integrin-beta6, is expected to report phase 3 data in second-line nonsquamous NSCLC in the first half of 2026.

With its biology experts and partners, Pfizer is profiling novel antigens to target for next-generation ADCs and is examining “a variety of permutations,” such as bispecific antibody constructions and novel payloads, Legos said.

“Instead of just focusing on traditional cytotoxic payloads, which I think have done very well, we’re now leveraging our small-molecule chemistry, and these could be low molecular weight inhibitors, they could be molecular glues, they could be targeted protein degraders,” Legos said. “These can become the basis of payloads where you can deliver very high dose, high concentration of a specific inhibitor directly to the cancer cells, thereby minimizing any safety or tolerability challenges but optimizing a therapeutic index.”

The mix-and-match has already begun. Instead of the vedotin payload, Pfizer is testing another integrin beta6 candidate with a TOPO1 payload in phase 1 dose escalation, and is developing a third intravesical molecule with a unique payload in non-muscle invasive bladder cancer to complement Padcev, according to Legos.

“We are working on modifying all of those different variables to bring forward what we believe to be a potentially practice-changing ADC in our priority disease areas,” Legos said.

The Pfizer oncology pipeline currently includes traditional small molecules, regular antibodies, ADCs and bispecifics, including T-cell engager. But the company does not have any CAR-T projects. Pfizer Ventures was a top investor in Capstan Therapeutics, an in vivo CAR-T biotech that was recently bought by AbbVie. 

Legos said he’s “fairly agnostic to” modalities and is “completely open […] to pursue innovation wherever it originates.”

“We have all of the building blocks that hopefully helps us accelerate the time to patient, but we are open and interested to explore all types of innovation, irrespective of modality,” he said.

While the large deals for Seagen and 3SBio’s PD-1xVEGF drug have put rejuvenation of oncology at the top of Pfizer investors’ mind, obesity has grabbed their attention lately in the forms of Pfizer’s $10 billion takeover of Metsera and a $1.9 billion biobucks deal for new GLP-1 therapies from Fosun Pharma.

Pointing to oncology’s 40% allocation of Pfizer’s entire R&D budget, Legos said he’s not worried about competition for resources internally with metabolic diseases but sees potential for mutual lessons sharing.

“I’m very hopeful about the potential intersections of cancer and the insulin axis as it relates to the future of care,” Legos said. “For a company that actually can be a leader in obesity and in oncology, it’s an opportunity to think about how cancer care is done on the backbone of GLP-1, amylin, GIP as part of the future.”