BioNTech has added evidence that PD-(L)1xVEGF-A bispecifics can unseat Keytruda. The German biotech reported an 18-month overall survival rate of 69.7% in a phase 1b/2 breast cancer trial, handily beating the cross-trial benchmark and bolstering its argument that its bispecific works in a broader population.
Biotheus, the company BioNTech is buying for $800 million upfront, ran the single-arm study to assess the PD-L1xVEGF-A bispecific BNT327, also known as PM8002. People with locally advanced or metastatic triple-negative breast cancer (TNBC) took the bispecific and nab-paclitaxel as a first-line therapy. Earlier readouts looked at responses and progression-free survival (PFS) but lacked overall survival (OS) results.
Researchers shared the first look at OS data from the Chinese trial at the San Antonio Breast Cancer Symposium Tuesday. In the 42 patients in the intention-to-treat analysis, BioNTech saw 12-, 15- and 18-month OS rates of 80.8%, 78.1% and 69.7%, respectively.
Merck & Co.'s KEYNOTE-355 trial, which BioNTech referenced, reported an estimated OS at 18 months of 47.8% in TNBC patients who received Keytruda and chemotherapy, with median OS of 17.2 months. For the BNT327 study, median OS data are yet to reach maturity as of the data cutoff for the trial, which had a median follow-up time of 19.5 months.
The Keytruda study was a much larger, global, placebo-controlled phase 3 trial. Bearing in mind the usual caveats about cross-trial comparisons, the available evidence suggests patients may live longer on BNT327. And the case for BioNTech’s asset looks particularly strong in subpopulations where traditional PD-(L)1 checkpoint inhibitors have struggled to move the needle.
Regulators are yet to approve a traditional checkpoint inhibitor in TNBC patients with PD-L1 expression of less than 10 on the combined positive score (CPS). Median OS in the Keytruda trial was 23 months in patients with CPS of 10 or more, compared to 14.7 months in women with lower PD-L1 levels.
In contrast, the overall 18-month OS rate in the BNT327 trial was underpinned by the results in people with low CPS. The 18-month OS in patients with CPS of 10 and higher was 55.6%. The figure climbed to 65.6% in people with CPS between 1 and 10, and rose again to 76.9% in the sub-1 CPS population.
There are reasons to question whether BioNTech will see that pattern of higher OS rates at lower CPS in a larger trial. The phase 1b/2 CPS subpopulations are small, with nine to 16 women in each category, and the trend only emerged in the 18-month data. No clear pattern emerged in the 12- and 15-month data. After 15 months, the OS rate was highest in the CPS-10 subgroup.
While the details could shift in subsequent trials, the results are broadly supportive of BioNTech’s plans to expand beyond the current use case for checkpoint inhibitors by validating BNT327 in patients with CPS of less than 10. In the population, where Keytruda isn’t approved, chemotherapy achieved a median OS of 15 months and a four-year OS of 15% to 20% in Merck’s TNBC trial.
The population is one of a number of settings where existing immuno-oncology therapies are yet to win approval. BioNTech believes it could roughly double the total market by validating its bispecific in cancers that account for more than 1.4 million new cases a year in the U.S. and the EU. The next step for BNT327 in first-line TNBC is the planned start of a registrational global phase 3 trial in 2025.
BioNTech is in a highly competitive race to unlock the TNBC market and other opportunities. Akeso and Summit Therapeutics, which set the PD-(L)1xVEGF-A hype train in motion with head-to-head data against Keytruda, are testing their drug candidate ivonescimab in a range of settings. And drug developers including Merck have earlier-stage prospects barreling down the pipeline.