ProBio: A Biomarker Driven Study in Patients With Metastatic Prostate Cancer

ProBio is an outcome-adaptive, multi-arm, open-label, multiple assignment randomised biomarker driven platform trial in patients with metastatic hormone-sensitive and castration-resistant prostate cancer.

Patients will be randomised to control or experimental treatment class arms. Patients in the control arm will receive standard of care following national guidelines and will remain within the control arm throughout the course of the trial. Patients in the experimental arm will be randomised to a treatment class (consisting of one or multiple drugs) based on a biomarker signature. The biomarker signatures are defined as tumour properties or mutations in certain genes/pathways identified in the scientific literature as important in prostate cancer treatment response. The biomarker signatures are identified using a gene panel specifically designed for advanced prostate cancer.

Alterations in the following genes/pathways or combinations thereof constitute the biomarker signatures:

• Androgen receptor

• DNA-repair deficiency

• TP53

• TMPRSS2-ERG gene fusion

• PI3K pathway alterations

Patients in the experimental arm can be randomized to the following treatments classes:

for mHSPC

• AR signalling inhibitors (Abiraterone acetate, Enzalutamide, Apalutamide)

• Taxane-based chemotherapy in combination with ARSi (Docetaxel plus Abiraterone acetate, or Darolutamide)

• PolyADP Ribose Polymerase Inhibitors (Niraparib plus Abiraterone Acetate)

for mCRPC

• AR signalling inhibitors (Enzalutamide, Abiraterone acetate)

• Poly ADP Ribose Polymerase Inhibitors (Niraparib plus Abiraterone acetate)

• Selective AKT Inhibitor (Capivasertib plus Docetaxel)

ProBio will use outcome-adaptive randomization, adapting the randomization based on the observed progression free survival (PFS) within biomarker signatures. Treatments will initially be assigned to patients based on the biomarker signatures for which that treatment is most likely to be effective. The trial will be analyzed within a Bayesian framework, which allows for calculations of the probability for each treatment that it is superior to standard of care within a given signature. Each experimental arm will be evaluated for efficacy relative to the control arm with the same biomarker signatures.

Participants and treating physicians will be blinded to ctDNA profile of each patient. The biomarker signatures will thus not influence treatment choice among controls (reflecting today's standard of care).

Further, ProBio will use the sequential multiple assignments trial (SMART) concept, where each patient who progresses within the trial will re-enter the trial and be re-assigned to another treatment based on the patient's current ctDNA profile. Patients will be withdrawn after in total maximal three randomized consecutive treatments after inclusion into the study.

The randomization probabilities within the experimental arm are defined in proportion to the probability that each treatment is superior to standard of care within a given biomarker signature, and therefore change as data accumulates in the trial and knowledge accumulates for what biomarker signatures and specific treatments that are more probable to be effective.

Trial results will be evaluated regularly by an independent data and safety monitoring board (DSMB). The DSMB will evaluate treatment-signature combinations with respect to:

• Graduation for superiority: A treatment-biomarker signature combination will be graduated from the trial if it has a Bayesian predictive probability of success in a future confirmatory phase III trial exceeding a pre-specified threshold (85%).

• Termination for futility: Treatment-biomarker signature combinations will be dropped from the trial for futility when success probabilities drop sufficiently low (less than 10% using a minimum of 20 patients assigned to the specific treatment-biomarker signature combination).

• Alternatively, if the maximum sample size of 300 and 150 patients (for mHSPC and mCRPC, respectively) assigned to a treatment biomarker signature is reached without graduation for superiority, assignments to that combination will end.

ProBio is a platform study. This means that new treatments and biomarker signatures can be added to the experimental arm in the future. This will be done after protocol amendments.