Commercialization of a Software Platform to Increase the Efficiency of Human Clinical Trials using Patient-Level Machine Learning Models

ABSTRACT Neurological and neurodegenerative disordersincluding Alzheimer s diseaseADParkinson s diseasePDHuntington s diseaseHDand amyotrophic lateral sclerosisALSare characterized by heterogeneous disease progressionAs a groupthese complex diseases are influenced by an interplay between several genetic and environmental factorsDrug development for these diseases could benefit from a framework that selects more homogeneous participants for inclusion in clinical trialsbut current methods have largely failed in this regardIn ALSboth slowly and rapidly progressing patients have been identified as confounding statistical analyses of clinical trialsStatistically speakingheterogeneous disease progression rates in ALS clinical trials contribute to both trial arm misbalances and high variances of study populationsThese issues increase the sizeduration and cost of clinical trials in neurological diseases and contribute to high failure ratesIn our successful phaseSBIR grantwe used ALS as a model disease to develop our neurological disease product prototype based on machine learning disease progression models that improves trial arm randomization and stratifies participants according to disease progressionWe worked with a clinical trialist to develop a user interface capable of returning a randomization group assignment in real timeOur ForecastOneTrials prototype product has passed IRB review and will be used in asite clinical trial starting in fallThe objectives of this phaseSBIR are to convert the ALS neurological product prototype into a commercializablerobustscalablemarket ready product with a versatile API and to develop a pipeline of models and applications that will expand our product offerings of drug development and clinical trial optimization tools for additional neurodegenerative indicationsAimConstruct a robustscalable platform infrastructure for the neurological product based on a centralApplication Programming InterfaceAPIthat receives user input and returns actionable information in real timeThe prototype ForecastOneTrials tool will be developed into a robustscalableCFR partcompliant productusing ALS as a first indication that allows the rapid deployment of additional applications and disease indicationsAimDevelop a pipeline of neurodegenerative models and applications to populate the neurological productWewill add additional disease indications to the product created in Aimincluding ADPD and HDThe suite of marketable products that we create following this phasegrant will answer drug development needs of a full portfolio of neurodegenerative diseasesUltimatelywe see a series of machine learning models and applications aimed at solving drug development issues for multiple disease areasincluding oncologycardiologyinfectious diseasesmetabolic disordersautoimmune diseases and respiratory diseasesetcThese models and applications will vastly increase the speed and efficiency of drug developmentresulting in fastercheapermore efficient drug trials that yield numerous new medications to ease human pain and suffering NARRATIVE This work builds on ForecastOneTrialsa prototype drug development tool created during our successful phaseSBIR grant using machine learning generated predictions of patient disease progressionThe phaseprogram will result in a suite of marketable products designed to answer drug development needs of a full portfolio of neurodegenerative diseasesThese models and applications will vastly increase the speed and efficiency of drug development in multiple disease areasresulting in fastercheapermore efficient drug trials that yield numerous new medications to ease human pain and suffering