The landscape of clinical trials is undergoing a profound transformation, with advanced imaging technologies increasingly integrated into study protocols. Leading this charge is Brainomix, whose artificial intelligence (AI)-powered medical imaging tool, e-Lung, has demonstrated remarkable potential in predicting disease progression and assessing treatment response in patients with progressive pulmonary fibrosis (PPF). New data from a Phase III study, INBUILD, reveals how e-Lung, alongside established quantitative CT measurements, can provide richer, more sensitive insights, potentially accelerating the development of life-saving therapies for interstitial lung diseases.
The Dawn of AI in Pulmonary Fibrosis Research
For decades, the diagnosis and monitoring of progressive pulmonary fibrosis (PPF) have relied heavily on traditional metrics like forced vital capacity (FVC), a measure of lung function. While valuable, these methods can sometimes lag behind the actual pathological changes occurring within the lungs, making it challenging to precisely gauge disease trajectory and the subtle impacts of novel treatments. This is where the integration of sophisticated imaging analysis, particularly AI-driven solutions, is poised to make a significant impact.
Brainomix’s e-Lung technology represents a cutting-edge advancement in this domain. By leveraging AI, e-Lung can analyze high-resolution computed tomography (CT) scans to quantify various aspects of lung disease, offering a more detailed and objective assessment than previously possible. The recent unveiling of data from the INBUILD study, a pivotal Phase III trial, underscores the power of this technology. The study investigated the efficacy of antifibrotic treatment, nintedanib, in 474 patients diagnosed with PPF.
The core of the INBUILD study’s imaging analysis involved a comparative approach, utilizing both Brainomix’s e-Lung software and a research algorithm developed at the University of California, Los Angeles (UCLA). The findings were compelling: both systems consistently and sensitively detected the effects of antifibrotic therapy. This validation from an independent research algorithm further solidifies the reliability and potential of e-Lung as a robust tool for clinical research.
Chronology of Discovery: Unpacking the INBUILD Study Findings
The INBUILD study, identified by its clinical trial identifier NCT02999178, was designed to rigorously evaluate the impact of nintedanib on patients with PPF. The trial’s design allowed for the detailed assessment of disease progression over time, with imaging playing a central role in capturing subtle changes.
Key Milestones and Observations from the INBUILD Study:
- Baseline Assessment: At the commencement of the trial, comprehensive CT scans were acquired from all 474 participants. These scans served as the foundational data for subsequent analysis.
- Intervention and Monitoring: Patients received either nintedanib or a placebo. Throughout the study, regular CT scans were performed at specific intervals, including at week 24 and week 52, to monitor changes in lung structure and disease extent.
- AI-Powered Analysis at Week 24: A critical analysis point occurred at the 24-week mark. Researchers utilized both Brainomix’s e-Lung and the UCLA algorithm to assess specific imaging markers. The data revealed that antifibrotic treatment with nintedanib significantly impacted e-Lung-specific measures, including the reticulovascular score (RVS) and the weighted reticulovascular score (WRVS). These scores are designed to quantify the extent and severity of fibrotic changes in the lung’s intricate network of alveoli and blood vessels.
- Longitudinal Impact at Week 52: By week 52, the analysis continued to demonstrate the sensitivity of these imaging tools to treatment effects. Notably, the total disease extent (TDE), a measure encompassing the overall burden of fibrotic disease in the lungs, was significantly affected by nintedanib treatment, as captured by e-Lung.
- Correlation with Clinical Outcomes: Beyond simply detecting treatment effects, the study also established a crucial link between baseline imaging findings and clinical progression. The analysis revealed that patients with higher baseline quantitative CT measures of TDE exhibited a greater rate of decline in forced vital capacity (FVC) over the 52-week period. This finding is particularly significant as FVC decline is a primary indicator of disease worsening in PPF.
This chronological progression highlights how e-Lung and similar quantitative CT approaches can provide a dynamic and nuanced understanding of disease evolution, complementing traditional functional assessments.
Supporting Data: Quantifying the Impact of e-Lung
The quantitative data generated by the INBUILD study provides concrete evidence of e-Lung’s capabilities. The sensitivity of the AI tool in detecting treatment effects is a key takeaway.
Specific Metrics and Their Significance:
- Total Disease Extent (TDE): This metric, as measured by e-Lung, represents the overall percentage of lung tissue affected by fibrosis. The INBUILD study showed that nintedanib treatment led to a significant positive impact on TDE at both 24 and 52 weeks, indicating a potential slowing or reduction in disease spread.
- Reticulovascular Score (RVS) and Weighted Reticulovascular Score (WRVS): These scores delve deeper into the microarchitecture of the lung, quantifying the severity of changes in the reticular patterns and the associated vascular network. The observed significant impact of nintedanib on RVS and WRVS at 24 weeks suggests that the drug is influencing the underlying fibrotic process at a granular level, which may not be immediately apparent through FVC measurements alone.
- Correlation with FVC Decline: The study’s finding that higher baseline TDE was associated with a greater rate of FVC decline is a powerful predictive indicator. It suggests that e-Lung’s ability to quantify disease burden at the outset of a trial can help stratify patients based on their risk of progression, allowing for more targeted therapeutic strategies and potentially earlier intervention.
The consistent correlation between AI-driven imaging metrics and established clinical outcomes, such as FVC, lends significant weight to the argument for incorporating these technologies into standard trial protocols. Brainomix’s assertion that these findings could "secure richer insights from clinical trials that can complement traditional measurements" is supported by this robust quantitative evidence.
Official Responses: Voices from the Forefront of Research
The unveiling of these findings has generated considerable enthusiasm and support from key opinion leaders and researchers involved in the study and the broader field of interstitial lung disease.
Anand Devaraj, Medical Director at Brainomix, emphasized the broader implications of the study’s results: "These findings support the continued development and incorporation of quantitative CT as an objective biomarker for clinical trials. Furthermore, such technology could also hold a place in clinical practice, as it could help to monitor a patient’s disease progression over time." This highlights the dual potential of e-Lung, not only to refine clinical trial design and analysis but also to eventually aid in the day-to-day management of patients with PPF.
Susanne Stowasser, a co-author of the INBUILD study and Head of Pulmonology/Rheumatology, Clinical Development at Boehringer Ingelheim, echoed this sentiment, underscoring the value of imaging in advancing therapeutic understanding: "Detecting meaningful changes in lung structure could build stronger foundations for evidence around antifibrotic therapies, while guiding the design of future clinical studies." Her statement points to the iterative nature of research, where advancements in imaging can inform and improve subsequent studies, creating a virtuous cycle of discovery.
The broader trend of incorporating imaging as endpoints in clinical trials, particularly in fields like cardiovascular and neurological research, is also acknowledged. This wider adoption signals a paradigm shift in how efficacy and safety are assessed, moving beyond purely functional or symptomatic measures to include objective, structural changes.
Implications: Reshaping the Future of Clinical Trials and Patient Care
The implications of Brainomix’s e-Lung technology and the insights gleaned from the INBUILD study are far-reaching, promising to reshape the future of clinical trials and, ultimately, patient care for those suffering from progressive pulmonary fibrosis and other interstitial lung diseases.
1. Accelerated Drug Development: By providing more sensitive and objective measures of disease progression and treatment response, AI-powered imaging tools like e-Lung can significantly expedite the drug development process. This could mean that promising new therapies reach patients who desperately need them sooner. The ability to detect subtle changes early can lead to quicker go/no-go decisions in clinical trials, saving valuable time and resources.
2. Enhanced Clinical Trial Design: The predictive power demonstrated by e-Lung, particularly its correlation with FVC decline, can inform the design of future trials. Researchers can better stratify patient populations, identify those most likely to benefit from a particular treatment, and optimize trial durations. This precision in trial design can lead to more robust and interpretable results.
3. Objective Biomarker for Clinical Practice: As Dr. Devaraj suggests, the potential for e-Lung to be integrated into clinical practice is immense. Currently, monitoring PPF progression can be challenging. The ability to objectively quantify disease burden and its changes over time via regular CT scans analyzed by e-Lung could revolutionize patient management, allowing for more personalized treatment adjustments and proactive interventions.
4. Democratization of Advanced Diagnostics: While sophisticated imaging analysis might seem exclusive, the increasing integration of AI aims to make these tools more accessible. As AI models become more refined and their implementation streamlined, the advanced diagnostic capabilities they offer can be brought to a wider range of healthcare settings.
5. Addressing the AI Governance Challenge: The article acknowledges the inherent challenges associated with implementing AI in healthcare, particularly concerning governance. As AI technologies become more prevalent in clinical workflows, establishing clear ethical guidelines, ensuring data privacy, and maintaining regulatory compliance are paramount. However, the projected growth of the healthcare AI segment, predicted to reach $57.4 billion by 2029 according to GlobalData, underscores the industry’s commitment to overcoming these hurdles.
In conclusion, the debut of data from Brainomix’s INBUILD study marks a significant milestone in the application of AI in medical imaging for clinical trials. The e-Lung technology is not merely an incremental improvement; it represents a potential paradigm shift, offering a more precise, objective, and predictive approach to understanding and treating complex lung diseases like progressive pulmonary fibrosis. As imaging becomes an increasingly indispensable component of research, AI-driven solutions are poised to unlock new frontiers in therapeutic discovery and patient care.
About the Author
