Reflecting on 2024: A Year of Growth and Validation (And Looking Ahead to 2025)

As we enter 2025, we reflect on an extraordinary year at PhaseV. Our AI/ML technology solutions have achieved remarkable outcomes in clinical development across diverse therapeutic areas and phases. 

With excitement for what lies ahead, here are some key highlights we achieved over the past six months.

Customer Highlights: Seven leading Big Pharma companies leveraged our ML/AI-guided platforms for adaptive trial design, causal-ML insights, indication selection & expansion, and more.

Research Partnerships Highlights: We leveraged extensive databanks, registries, and expert research centers to develop new causal disease models that enable the prediction of treatment responses and the optimization of clinical development decisions. Recent examples of our research collaborations include the UHN Lupus Clinic and the Kidney Research Network.

PhaseV Platform Advancements Highlights:

• Enhanced ML capabilities for trial design and new LLM features.

• Advanced visualization tools for multi-stakeholder decision-making in adaptive trials.

See more about these platform advances and capabilities below

Scientific Contributions Highlights:

• Published our research on causal responder detection, CATE estimation, oncology treatment analysis, and leveraging RWD to predict disease progression in Oncology.
• Presented our findings at key events including the ASA Workshop, DIA, Eular, Bayes, and ML4H.

Data Growth: We expanded our proprietary data lake to include over 300,000 patients and strengthened data partnerships and collaborations, leveraging clinical and real-world datasets to optimize clinical development

Team Expansion: We are thrilled to welcome Brad Carlin , PhD as our new Senior Director of Statistics and Data Science. Dr. Carlin, who has been with us as an advisor since the start and is now joining us full-time, brings a wealth of expertise and experience to support client projects and drive internal development.

New Location: We have relocated to a vibrant new office space in Kendall Square, Cambridge, a hub for life science innovation and collaboration.

Looking Ahead: Our Vision for 2025

PhaseV is poised to make 2025 a year of unparalleled impact. Here are our key priorities for the new year:

• Pushing the Boundaries of Causal ML: 

Achieve unparalleled accuracy in identifying heterogeneous treatment effects, even within small sample sizes, enabling the precise identification of patient populations to maximize treatment impact and increase the probability of success.

Harness causal modeling to revolutionize indication selection and expansion, optimizing clinical development decision-making on a transformative scale.

• Power Any Baysian or Adaptive Trial Design: Streamline even the most complex trial designs with cutting-edge optimization algorithms and enhanced simulation capabilities.
• External Data to Enhance Each Project: Expand our clinical and RWD network to provide our customers with comprehensive datasets seamlessly integrated with our solutions to uncover the most optimal development pathways.
• Advanced ML Analytics for Clinical Operations: Deploy advanced ML solutions to tackle operational challenges, from trial site optimization to supporting open-label trials and beyond. 

These initiatives will empower our partners to achieve unprecedented efficiency, maximize trial success, and bring life-saving treatments to patients faster.

Join Us at JPM 2025 Let’s connect at JPM in San Francisco this January! Book a meeting with us at info@phasevtrials.com.

We Are Always Up For A Challenge!

What is the biggest challenge you’re facing in clinical development? Share it with us at info@phasevtrials.com, and we’ll provide tailored insights on how our AI/ML technology can deliver the solution you need.

PhaseV’s Platform for the Design and Optimization of Adaptive Clinical Trials 

• Enhanced optimization capabilities 

We've made exciting progress in enhancing how we design clinical trials through our advanced optimization capabilities. Traditional trial design often involves carefully selecting various trial parameters - like sample sizes, decision thresholds, and timing of analyses - primarily based on past experience. Our innovative machine learning approach transforms this process by efficiently exploring a vast range of options to identify the most effective trial designs. 

This breakthrough enables sponsors to optimize critical factors like trial duration, cost, and the probability of successfully identifying effective treatments. Our capabilities now extend across a broad spectrum of trial designs, including early-phase studies and those with discrete parameters, providing unparalleled support for diverse clinical development needs.

* Research Department, PhaseV, Y. Racah et al, Model-Guided Parameter Optimization for Complex Innovative Trial Designs, ASA BIOP RISW 2024.

• New LLM capabilities for trial design

Our new LLM capabilities, integrated into our ML-based trial design, leverage data from published literature and clinicaltrials.gov to enable sponsors designing trials to better estimate the expected placebo response in their studies, effectively conducting a “meta-analysis on the fly”. 

These two capabilities were recently presented in a demo at the ML4H event.

• Enhanced Visualization:

One of our core focus areas when providing a platform for multiple stakeholders (statistical, clinical, operational) is enhancing visualizations to effectively communicate complex information. A few notable updates to our platform include a redesigned trial schema that now features the “classic” CONSORTdiagram, side by side with new and interactive visualizations tailored to adaptation-specific decision-making within trials:

Additionally,our interactive widget for early phase trials,allows users to quickly view their input toxicity and efficacy profiles, as well as the optimal biological dose (OBD) and maximum tolerated dose (MTD):