A Phase 1B Study of Canakinumab, Spartalizumab, Nab-paclitaxel, and Gemcitabine in Metastatic Pancreatic Cancer (PC) Patients

This is an open-label multi-center phase Ib study to confirm the recommended phase II/III dose of canakinumab and spartalizumab in combination with nab-paclitaxel and gemcitabine. The study will recruit patients with metastatic pancreatic adenocarcinoma treated in the first line setting. The starting dose level of canakinumab explored will be 250 mg Q4W ("starting dose level"). In case of unacceptable toxicity of the starting dose level of canakinumab, the dose of canakinumab will be de-escalated to the "-1 dose level" administered as 250 mg Q8W, while other components of the combination stay at the same dose as the starting dose level. Patients will be observed for dose limiting toxicities (DLTs) for a minimum duration of 56 days (8 weeks). To achieve study objectives and to ensure the adequate number of DLT evaluable patients, the study will recruit approximately ten patients to have at least 6 evaluable patients per dose level of canakinumab. Additional approximately ten patients (to have at least 6 additional evaluable patients) may be enrolled at lower dose level in case a dose de-escalation is necessary. Dose confirmation will be guided by an adaptive Bayesian logistic regression model (BLRM) based on any DLTs observed for two cycles of treatment (i.e. 56 days, or 8 weeks). The adaptive BLRM will be guided by the Escalation with Overdose Control (EWOC) principle to control the probability of DLT in future patients on the study. BLRM is a well-established and widely used method to estimate the recommended dose for expansion (RDE) or maximal tolerable dose (MTD) in clinical trials in patients with cancer with small sample size. The use of Bayesian response adaptive models for small datasets has been endorsed by academic publications (Babb et al. 1998, Neuenschwander et al. 2008, Neuenschwander et al. 2010, Natanegara et al. 2014), by the European Medicines Agency (Guideline on Clinical Trials in Small Populations, 2007) and it constitutes an important aspect of the FDA's Critical Path Initiative (Clinical Path White Paper, FDA, 2004). The Bayesian analysis incorporates prior toxicity data of single agent and drugs combinations together with the currently available data to predict the probability of DLT and excessive toxicity of a dose level of interest. The Bayesian method is be based on a Meta-Analytical-Combined (MAC) approach (Spiegelhalter 2004, Neuenschwander 2016) to combine all historical and concurrent data. Prior toxicity information included in the BLRM model was obtained from three studies with canakinumab as a single agent and combination of canakinumab and spartalizumab (PDR001X2101, ACZ885I2202, PRD001X2103) and from a phase I/II study of nab-paclitaxel + gemcitabine (Von Hoff D, et.al., 2011). Simulation was used to illustrate the recommendation from BLRM under a set of hypothetical scenarios with assumed number of evaluable patients and DLTs. The decisions on a recommended dose will be made by the Investigators and the Sponsor in a Safety Review meeting when at least 6 DLT evaluable patients per dose level will be observed for DLTs for a minimum duration of 56 days (8 weeks). Safety review will be based upon the review of all relevant data available including treatment tolerability and safety information together with the BLRM summaries of DLT probability, PK, PD, and preliminary activity information (if available) at the time of the meeting. Patients will be treated until disease progression per RECIST 1.1, unacceptable toxicity, or until the patient or treating physician decides to stop treatment. Pharmacokinetic (PK) and immunogenicity (IG) samples will be collected at specific time points throughout treatment. Each treatment cycle is 4 weeks. All patients must be followed for safety up to 150 days after the last dose of spartalizumab or canakinumab, or 30 days after the last dose of the combination chemotherapy, whichever the later. After the end of safety follow-up, patients will be followed for disease progression if discontinuation of treatment is due to reason other than progression, and for survival (via telephone call or onsite visit if a patient happens to be visiting the site) until the end of study The study completion is defined as when the last patient has completed the study treatment, safety follow up, and completed survival follow up period up to 1 year from first treatment, whichever is later or in the event of an early study termination decision, the date of that decision.