The results of this study arm were published simultaneously with the full BEECH study results in Annals of Oncology.2 The results demonstrated that among 16 patients who were intensively sampled during the phase 1 portion, those with a long PFS showed ctDNA suppression after just 8 days of treatment. The researchers found that day 1 of the second treatment cycle was the optimal moment at which to observe a change in ctDNA that could predict a longer PFS. Patients with a ctDNA ratio of less than 0.25 had an improved PFS compared with patients with a higher ratio, which suggested a lack of ctDNA suppression (P =.0003).

In phase 2, Dr Turner and colleagues sampled ctDNA in 42 patients at the start of the study and, following the findings of the preliminary sampling, at day 1 of the second treatment cycle. The median PFS was 11.1 months among women whose ctDNA diminished, compared with 6.4 months among women whose ctDNA levels remained high (hazard ratio (HR), .20; CI, 0.083-0.50; P <.0001). 

A separate look at the 2 treatment groups — paclitaxel plus capivasertib versus paclitaxel plus placebo — revealed no difference in PFS prediction.

The researchers also checked whether monitoring ctDNA during the treatment could pinpoint when a tumor became resistant to the therapy. They followed mutations in ctDNA every 4 weeks from treatment start to disease progression among 50 patients in the phase 2 study. 

Although ctDNA dropped below detectable levels in several patients before a resistance-linked rise, many patients had changes that were difficult to interpret, with levels that tended to fluctuate at and near the lowest levels — a phenomenon the researchers referred to as “bumping along the bottom.” Often, the rise from this nadir was too slow for the researchers to identify its exact timing.