244MO - A phase II trial targeting disseminated dormant tumor cells with hydroxychloroquine, everolimus or the combination to prevent recurrent breast cancer (“CLEVER”)

Background

Breast cancer (BC) recurrence may follow a dormant phase in which quiescent cells reside in niches such as bone marrow (BM). Dormant BM disseminated tumor cells (DTCs) are independently associated with BC recurrence/death. We investigated whether targeting dormancy through autophagy inhibition (hydroxychloroquine (HCQ), and/or mTOR signaling (everolimus (EVE) in DTC+ BC survivors was feasible, reduced DTCs and/or prevented recurrence.

Methods

The CLEVER trial (NCT03032406) is a randomized, phase II trial in patients (pts) diagnosed within 5 years, with positive nodes, triple-negative disease, high-risk Oncotype/Mammaprint, and/or residual disease post-neoadjuvant therapy who completed all treatment except endocrine therapy. DTCs were detected in BM aspirate (BMA) by IHC with pan-CK antibody AE1/AE3. DTC+ pts were randomized to six 28-day cycles (C) of HCQ (600 mg BID), EVE (10 mg daily) or both (+/- 3-month (m0) observation period). If DTC+ persisted after C6, pts received another 6C HCQ+EVE. DTC assessment was done after C3, C6, C12 (if applicable) and 6-mo after end of treatment. Adverse events (AE) were assessed by CTCAEv4. Primary endpoint was feasibility, defined as >75% completion of C6C without G3/G4 AE. Secondary endpoints were safety, DTC response rate (RR) and 3-year RFS. DTC RR was analyzed with Bayesian Poisson regression models.

Results

184/197 eligible pts had baseline BMA, 55 (30%) were DTC+, 53 were randomized: HCQ (n=15), EVE (n=15), and HCQ+EVE (n=23). 13 patients had repeat after 3-mo observation. Feasibility endpoint was met. There were no G4/5 toxicities. At a median follow up of 42 mo (range 7-60), 1 pt recurred in lung (after 2 cycles EVE), 1 pt developed new contralateral breast cancer. The posterior probabilities that HCQ, EVE, and HCQ+EVE reduced DTCs by at least 80% after C3 vs observation alone are 99.1%, 98.2%, and 99.9%, respectively.

Conclusions

The CLEVER trial provides proof-of-principle that therapeutic targeting of dormant BC is feasible and active in eliminating DTCs by targeting dormancy-specific mechanisms. Follow up for recurrence and survival is ongoing.

Clinical trial identification

NCT03032406.

Editorial acknowledgement

Legal entity responsible for the study

University of Pennsylvania.

Funding

National Cancer Institute R01CA208273, Novartis (Drug Only).

Disclosure

A. DeMichele: Financial Interests, Institutional, Research Funding: NOVARTIS, Neogenomics, Pfizer, Genentech, Calithera. I. Makhlin: Financial Interests, Personal, Advisory Board: AstraZeneca, Bayer, Gilead. L. Chodosh: Financial Interests, Personal, Expert Testimony: Teva Pharmaceuticals, Eisai, Sanofi, Lilly, Whittaker, Clark and Daniels, Imerys, Colgate, Sterigenics. All other authors have declared no conflicts of interest.