#USA #Boston #PhaseV #Veradigm #GLP-1

Introduction

PhaseV, a leading company in clinical development through machine learning, has introduced a groundbreaking predictive model aimed at improving understanding of patient responses to GLP-1 (glucagon-like peptide-1) medications. This innovative model utilizes real-world patient data sourced from MarketScan claims and Veradigm electronic health record (EHR) systems. By accurately forecasting adverse events related to these medications, as well as tracking weight trajectories and potential treatment dropouts, PhaseV seeks to enhance patient outcomes and optimize clinical trial designs.

The Importance of Adverse Event Prediction

In clinical trials, patients often discontinue medications due to unpleasant side effects, particularly gastrointestinal issues that are common with GLP-1 treatments. Such discontinuation can severely compromise the trial’s ability to determine a drug's efficacy. Therefore, understanding tolerability and the factors influencing patient dropout rates becomes essential in the design of clinical trials for these medications. Unfortunately, there are limited studies addressing how different medication titration speeds affect these outcomes.

Utilizing real-world data allows researchers to appreciate the significant variability in treatment adjustments across individual patients, which may not be possible in controlled clinical trial environments. For instance, a patient experiencing adverse effects may require a slower dose escalation, while inadequate weight loss might necessitate a quicker change in dosage. By combining longitudinal data for each patient, PhaseV’s model iteratively examines the patterns of adverse events, weight control, and dosage changes, leading to a richer understanding of treatment dynamics.

Mechanism of the Predictive Model

At the heart of PhaseV's model is the integration of various patient data points, such as dosage levels, weight changes, adverse events, and dropout occurrences. Unlike traditional approaches that treat these metrics as isolated incidents, this model interlinks them in a time-sensitive manner, effectively capturing the continuous nature of patient experiences. For example, rapid adverse events can influence subsequent dosing decisions, while a patient’s weight response may also dictate their treatment path.

Furthermore, the model equips researchers with the ability to perform “what if” analyses, enabling exploration of how adjustments in titration schedules might impact specific patients’ experiences. This capability is particularly valuable since it allows researchers to anticipate outcomes even when randomized trials confirming these strategies are unavailable.

Validation of the Model

To validate its effectiveness, PhaseV tested the model's predictions against previously unseen patient data. The results were compelling; the model accurately predicted actual patient experiences concerning weight changes and the timing of adverse events and dropouts. Jennifer Meller, MD, MBA, Medical Director at PhaseV, emphasized the significance of this model, stating that for drug sponsors, it offers a fact-based strategy for making dosing decisions, rather than relying solely on inherited assumptions.

Implications for Future Clinical Research

The validation of this model is a pivotal step forward in clinical research, enabled by PhaseV’s expansion into real-world data acquisition, which now includes the longitudinal data from MarketScan and Veradigm. This powerful new tool can influence clinical trials by allowing for evidence-based decisions regarding titration strategies before any patient is administered the drug in trial settings.

About PhaseV

Founded in Boston, PhaseV is at the forefront of leveraging technology to empower biopharma sponsors, contract research organizations (CROs), and investors with refined tools for more accurate decision-making throughout the clinical trial process. With over 100 clinical programs supported for more than 50 global sponsors, including ten of the top 20 pharmaceutical companies, PhaseV's data and analytics capabilities have been demonstrated to significantly reduce development costs and trial duration, while also increasing the probability of successful outcomes. To learn more about their innovative approaches, visit www.PhaseVTrials.com and follow PhaseV on LinkedIn.