PRECISE CURATE.AI: A prospective feasibility trial to dynamically modulate personalized chemotherapy dose with artificial intelligence.

Abstract

1574 Background: Most treatment guidelines recommend chemotherapy at maximum tolerated doses, which does not always lead to optimal efficacy, but implicitly results in toxicity. To overcome this challenge, we developed CURATE.AI, a small data, AI-derived platform that harnesses only a patient’s own prospectively/longitudinally acquired data to dynamically identify their own optimal and personalized doses. We subsequently harnessed CURATE.AI to dynamically modulate individualized chemotherapy doses for patients in a prospective clinical trial. Methods: We conducted an open-label, multi-center, single-arm, prospective feasibility trial in patients diagnosed with advanced solid tumors and treated with single-agent capecitabine, XELOX or XELIRI (+/- biologics) (NCT04522284). The standard-of-care (SOC) capecitabine dose was 1000 mg/m2, unless adjusted by clinician to account for patient’s comorbidities and organ dysfunction. Using an AI-discovered second-order correlation between patient-specific variation of capecitabine doses and corresponding tumor marker (CEA, CA19-9 or CA-125) readouts for each cycle, CURATE.AI generated individualized patient digital avatars and recommended bespoke dose for the subsequent cycle. The clinicians were permitted to accept CURATE.AI dose recommendations, or reject the recommendations and dose based on clinical judgement. Results: Since August 2020 we recruited ten patients: single-agent capecitabine (n = 1), XELOX (n = 6), and XELIRI (n = 3). As of 20 Jan 2022, one patient remains on the trial. The prescribed dose was on average reduced by 20 % (± 13.8 %) as compared to the projected SOC dose. The nine reported patients completed 3.9 cycles (± 2.2 cycles), with the longest participation lasting 8 cycles. CURATE.AI recommendations were considered in 27 out of 40 total dosing decisions and accepted for prescription in 26 of those decisions. The reasons for not considering CURATE.AI included insufficient time from patient recruitment to the first dose administration and complex medical circumstances at the time of the dosing decisions. Conclusions: CURATE.AI has been successfully incorporated into the clinical workflow of dynamic dose selection in the treatment of solid tumors under a clinical trial. Prospective validation of CURATE.AI led to a reduction of an average prescribed capecitabine dose, which alongside additional preliminary findings may eventually play an important role in improving patient response rates and durations compared to SOC. Results from the PRECISE CURATE.AI trial support the initiation of a randomized clinical trial and potential expansion towards other oncologic indications. Clinical trial information: NCT04522284.

Hybrid Email and Outpatient Clinics to Optimize Maintenance Therapy in Acute Lymphoblastic Leukemia

“…Patients can upload the CBC report and receive back drug prescriptions with a possible visual aid of daily dose distribution and a reminder message on their mobile phones to take the drugs. With these data stored on a database, machine learning/artificial intelligence tools can further optimize dosing 28 and aid clinicians.…”

N-of-1 medicine

“…Importantly, this approach demonstrated that while small data could successfully achieve global optimisation of efficacy and safety from a massive parameter space of combinatorial permutations, a prospectively acquired and carefully designed set of experiments was critical to pinpoint a list of actionable regimens. [ 12 37 38 44 45 46 47 48 49 50 51 52 ]…”

“…Through emerging methodologies of harnessing data to drive healthcare decision-making, the fields of precision and personalised medicine are being increasingly impacted by the fields of digital health and digital medicine. [ 12 13 14 15 16 17 18 19 20 21 22 23 24 25 ] Wearable devices are increasing our access to longitudinal data to drive digital biomarker development, and the rapid rise in actionable data is empowering new ways of driving clinical decision support. [ 26 27 28 29 30 ] Realising clinically actionable strategies to correlate this data, big or small, into optimal and personalised treatment guidance may profoundly impact the ability to sustainably deliver truly personalised N-of-1 medicine at the required scale for all stakeholders involved.…”

Deep Learning and Drug Discovery for Healthy Aging

“…Recent work has shown that substantial dose reductions can maintain stable disease in human solid cancer treatment. 17 In addition, evolutionary dynamics has been used to reduce total drug doses needed for human prostate cancer therapy. 18 Given the aging-related nature of osteoporosis and other bone diseases, modulated dosing in accordance with physiological and disease changes that inevitably take place during chronological age progression may help prolong efficacy.…”