Precision Promise

Precision Promise was built by the Pancreatic Cancer Action Network with the statisticians behind I-SPY 2 and GBM AGILE and opened in January 2020 at a network of United States pancreatic cancer centres. Patients with metastatic pancreatic adenocarcinoma entering first- or second-line treatment were randomised adaptively between standard chemotherapy (gemcitabine with nab-paclitaxel or modified FOLFIRINOX) and experimental arms, with overall survival as the primary endpoint and a design that let a promising arm expand seamlessly into a registration-quality stage. Tumour and blood samples were collected to search for subgroups that responded.

The first experimental arm, SM-88 (racemetyrosine), was discontinued in 2022 for lack of efficacy. The second, pamrevlumab, an antibody against connective tissue growth factor given with chemotherapy, did not improve survival and was closed after its interim analysis. No further arms graduated, and the registry records the trial as completed with about 500 participants enrolled.

Precision Promise is the cautionary member of the platform family: the machinery worked, but a platform is only as good as the drugs offered to it, and pancreatic cancer had few strong candidates during the years it ran. It leaves a well-characterised control cohort and biobank, and lessons about how a disease charity can sponsor a registration-grade platform, that later pancreatic trials draw on.