Abstract Background: Multiple Myeloma (MM) is an incurable plasma cell neoplasm with an estimated incidence of 32,000 patients per year in the United States. Novel agents and advances in precision medicine for selecting drugs tailored to an individual have increased the likelihood of remission and prolonged survival. To maximize efficacy of the identified drug combinations, dosing strategy should also be tailored to an individual, and dynamically dosed as the individual evolves during treatment. Current dosing strategies are unable to personalize and optimize dosing in a combinatory treatment due to a virtually infinite parameter space as well as stochastic biological behavior. As a result, dose adjustments are instead dictated by treatment tolerability, while the efficacy outcomes remain sub-optimal. In this pilot clinical trial we aim to test if the gap in the dosing strategy for combinatory treatment for MM can be addressed with CURATE.AI - clinically validated deterministic AI-derived platform that generates an individualized profile based only on that individual's data, and recommends the next dose towards optimized efficacy within the pre-identified tolerability range. CURATE.AI has already been used to prospectively modulate dosing strategies for indications ranging from oncology (solid tumor) to infectious diseases (HIV/TB) and immunosuppression (liver transplant). Methods: In this pilot study, CURATE.AI will be used to provide dose recommendations for bortezomib and cyclophosphamide, or bortezomib and thalidomide in the presence of dexamethasone at a fixed dose (VCD, or VTD combinations). The tested hypothesis is: CURATE.AI-driven dosing will result in noninferior efficacy with less toxicity and fewer adverse events. The prospective clinical trial design is a randomized, multiphase (Phase 2/Phase 3), parallel two-arm, single-blinded, N-of-1, exploratory pilot trial with 1:1 allocation, noninferiority, listed under Clinicaltrials.gov identifier NCT03759093. The recruitment is expected to commence in the first quarter of 2020 at National University Hospital in Singapore. Twenty participants fulfilling inclusion criteria (males and females over 21 years [Singapore's legal age for adult consent] with newly diagnosed and evaluable MM, ECOG 0-2, and meeting toxicity criteria within 21 days from the start of the treatment) will be randomly allocated (1:1) into the control or the experimental arm. The subjects will be blinded to the arm allocation. Subjects in the control arm will receive dosing of VCD or VTD according to standard of care - physician following the Singapore Myeloma Study Group Consensus Guidelines. Subjects in the experimental arm will receive fixed dexamethasone dose as in standard of care, while bortezomib and cyclophosphamide (for VCD treatment) or bortezomib and thalidomide (for VTD treatment) doses will be dynamically modulated with CURATE.AI for each subject independently (series of N-of-1) based on that subject's longitudinally collected efficacy output measures. The primary outcome measure is the response rate at the end of cycle 4. The results of this pilot study will additionally inform the logistical and practical suitability of N-of-1 randomized trial designs for testing personalization and dynamic dose modulation for oncology. Citation Format: Agata Blasiak, Theodore Wonpeum Kee, Masturah Bte Mohd Rashid, Edward Kai-Hua Chow, Sanjay De Mel, Wee Joo Chng, Dean Ho. CURATE.AI-optimized modulation for multiple myeloma: An N-of-1 randomized trial [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr CT268.