The FDA’s Oncologic Drugs Advisory Committee (ODAC) on Thursday voted 6 to 3 that AstraZeneca has not demonstrated its oral SERD camizestrant offers a clinically meaningful benefit for treating HR-positive, HER2-negative metastatic breast cancer when a tumor ESR1 mutation is detected before radiographic progression on first-line therapy.
Given the FDA’s own negative review, approval now looks unlikely. It marks an early setback for AZ’s plan of introducing a new drug with potential for $5 billion in peak annual sales.
In a novel design, AstraZeneca is proposing using camizestrant once an ESR1 mutation is detected in a patient while on first-line treatment ahead of disease progression.
The Serena-6 study showed a 56% improvement in progression-free survival (PFS) for switching to camizestrant, used on top of a CDK4/6 inhibitor, compared with continuing treatment with an aromatase inhibitor (AI) and a CDK4/6 drug. Overall survival was immature, but no sign of detriment was observed.
FDA vs. AstraZeneca
The FDA took issue with AstraZeneca’s trial design, arguing it does not address whether this first-line switch approach provides long-term benefit to patients compared with the standard practice of waiting until disease progression to use an oral SERD drug like camizestrant. Although camizestrant extended median PFS by 6.8 months, the FDA is unsure of its clinical meaningfulness because it was not measured from a standard time point.
Besides, Serena-6 did not allow crossover, meaning no patient in the control arm was allowed to receive camizestrant after progression. Fourteen percent of control-group patients received other oral SERDs in subsequent treatment. Crossover “could have allowed some characterization” of AZ’s proposed strategy, Mirat Shah, M.D., an FDA oncology clinical reviewer, said during her presentation Thursday.
But AZ argues that waiting until disease progression to give an oral SERD is less effective because tumors at that time have increased genomic alterations and therefore become harder to treat.
“While disease progression is not immediate in frontline after the emergence of ESR1 mutations, the AI no longer controls these mutant clones. Replacing the AI with camizestrant intercepts this emerging resistance,” Kevin Kalinsky, M.D., division director of medical oncology at Emory University, a paid consultant to AZ, said during his presentation Thursday.
“Since I was in training, patients have wanted an actionable blood marker to inform whether a treatment is working,” Kalinsky said. “With ESR1 mutations, we finally have such a marker, and with camizestrant, a way to effectively target it.”
The FDA also flagged that if implemented, patients would need to be monitored constantly for ESR1 mutations with only few becoming eligible for camizestrant. The burden of testing deepens the agency’s concern of “uncertain clinical benefit.”
Circulating tumor DNA (ctDNA) tests used to detect ESR1 mutations are routinely used today in clinical practice, Massimo Cristofanilli, M.D., from Weill Cornell Medicine, a paid consultant to AZ, said during the meeting.
“Profiling blood-based ctDNA is minimally invasive and ideal for monitoring the genetic evolution of tumor to identify actionable mutation early,” he said Thursday.
The FDA expands its caution beyond AZ’s application, fearing approving camizestrant would open a questionable precedent that, as Shah put it, has “far-reaching implications for many future patients and many future trials.”
“The biggest concern with the Serena-6 trial is that approving the drug camizestrant would include an endorsement of a treatment paradigm that does not have established clinical benefit,” Shah said.
The FDA is concerned that the early switch approach may be adopted in future breast cancer trials, which could “potentially introduce more toxic therapies earlier into” treatment, and might be picked up by trials in other tumor types, “without evidence of long-term benefit,” Shah said.
Toni Choueiri, M.D., an ODAC member from the Dana-Farber Cancer Institute, said he does not believe an approval here will open a Pandora’s box.
That’s because “every trial that’s biomarker-based addresses a unique disease, a unique biomarker and a unique setting of the patient journey,” he said. “Actually, the next trial will build on a new paradigm and come up with a new set of trials that builds on the previous trial and address from the start any potential shortcoming of previous trials, like Serena-6 and underpowered OS.”
Expert panel’s reservation
Advancing personalized medicine requires rethinking practice, Ingrid Mayer, M.D., AZ’s global clinical strategy head of breast and gynecological cancers, said at the meeting.
“We recognize that Serena-6’s innovative design challenges current norms,” she said. “But what gives us confidence in this approach is the consistent benefit across primary, secondary and exploratory endpoints.”
Camizestrant “extends the clinical benefit of frontline endocrine therapy in CDK4/6 inhibition, delaying progression and deterioration in quality of life,” Mayer continued. “It maximizes time on well-tolerated therapy and delays the need for chemotherapy or antibody conjugates, and it is straightforward to use in clinic, empowering patients to adjust treatment before worsening symptoms and disease progression.”
Despite Mayer’s plea, most experts on the advisory committee sided with the FDA. While AZ’s concept of using a biomarker to identify eligible patients for treatment before progression is fascinating, the Serena-6 data didn’t prove camizestrant can be that drug, Michael Kelley, M.D., from the Duke University Medical Center said.
“What is being asked here is early versus late. It’s not early versus never, and that is what essentially was done,” Kelley said during the deliberation before voting against AZ’s application.
Neil Vasan, M.D., Ph.D., from NYU Langone Health disagreed that Serena-6’s design for subsequent treatment makes its results uninterpretable.
“The FDA is correct that this trial does not define an optimal sequence, but the field itself has not defined one in ER-positive metastatic breast cancer,” Vasan said after voting for approval. “Second-line therapy is not a single standard, it’s a plurality of options shaped by biomarkers, toxicities, access timing. This was true in 2021, and it’s still true today.”
Kelley said he is not arguing for a perfectly designed study. “But I do think we need one that answers the question, and one part of that can be offering the experimental drug as an option to be used at a future point, in addition to other standard treatments,” he said.
Several experts on FDA’s panel opined that AZ’s trial failed to show that ESR1, a known prognostic biomarker, is really a predictive biomarker, the treatment of which may benefit patients.
“I really wonder if we are again exploiting the hope of women with metastatic breast cancer that somehow, if they’d known earlier, the trajectory of the disease would be different,” Natalie Compagni-Portis, a patient representative, said after voting no to AZ’s proposal. “But without evidence of longer overall survival, and with so much discussion about quality of life, to have very sparse data on that, it really seems only speculation and hope.”
