New model “Serial CTRS” shows promise in predicting survival for advanced NSCLC patients
Revolutionizing Lung Cancer Prognosis with AI
In a significant development in oncology, Onc.AI has introduced a groundbreaking artificial intelligence (AI) model aimed at enhancing prognosis predictions for lung cancer patients. This innovative approach leverages radiomic analysis to assess immunotherapy responses in individuals with late-stage non-small cell lung cancer (NSCLC). The findings were presented at the 2024 Annual Society of Clinical Oncology (ASCO) meeting, highlighting the potential of AI in transforming cancer care. Distribution de Communiqués de Presse+2BioSpace+2Synapse+2Synapse+2Distribution de Communiqués de Presse+2BioSpace+2
The Challenge of Predicting Immunotherapy Responses
Immunotherapy, particularly immune checkpoint inhibitors (ICIs), has become a cornerstone in treating NSCLC. However, predicting which patients will benefit from such treatments remains a challenge. Traditional biomarkers often fall short in accurately forecasting patient responses, leading to unnecessary treatments and associated costs. Onc.AI’s AI-driven model addresses this gap by analyzing radiomic features from pre-treatment CT scans to predict progression-free survival (PFS) and overall survival (OS) outcomes. Synapse+1aacr.org+1BioSpace
Collaboration with Pfizer: A Step Forward
Onc.AI’s collaboration with Pfizer has been instrumental in validating the AI model’s efficacy. The partnership focused on assessing the model’s ability to predict responses to Sasanlimab, an investigational PD-1 agent, in metastatic lung cancer patients. The results demonstrated significant hazard ratios for PFS and OS, indicating the model’s potential in stratifying patients based on their likelihood of benefiting from immunotherapy. Synapse+2BioSpace+2Distribution de Communiqués de Presse+2Distribution de Communiqués de Presse+2Synapse+2BioSpace+2
Clinical Validation Across Diverse Cohorts
The AI model underwent rigorous testing across multiple NSCLC cohorts:Distribution de Communiqués de Presse+1BioSpace+1
- 1L Real-World Cross-Validation Cohort: PFS Hazard Ratio (HR) of 0.49 in all-comers and 0.28 in patients receiving single-agent immunotherapy.Synapse+2Distribution de Communiqués de Presse+2BioSpace+2
- 1L Real-World Holdout Cohort: PFS HR of 0.18 in patients receiving single-agent immunotherapy.Targeted Oncology+7BioSpace+7Distribution de Communiqués de Presse+7
- 2L+ Clinical Trial Cohort (NCT02573259): PFS HR of 0.30 and OS HR of 0.29 in NSCLC patients treated with Sasanlimab.Distribution de Communiqués de Presse+2Synapse+2BioSpace+2
These findings underscore the model’s robustness and its applicability across different patient populations. genialis.com+4Synapse+4BioSpace+4
Implications for Clinical Decision-Making
The integration of AI into clinical workflows offers a non-invasive tool to aid oncologists in treatment planning. By accurately predicting patient responses to ICIs, the model can help avoid ineffective treatments, reduce healthcare costs, and improve patient outcomes. Moreover, it facilitates personalized medicine by identifying patients most likely to benefit from specific therapies. PMC+1Targeted Oncology+1Synapse+1BioSpace+1
Advancements in AI and Radiomics
Radiomics involves extracting quantitative features from medical images, providing insights into tumor characteristics that may not be visible to the naked eye. Onc.AI’s model utilizes deep learning algorithms to analyze these features, offering a more nuanced understanding of tumor behavior and treatment responses. This approach represents a significant advancement in precision oncology.
Future Directions and Regulatory Considerations
While the AI model shows promise, its integration into clinical practice requires careful consideration of regulatory frameworks and validation studies. Ensuring data privacy, addressing potential biases, and obtaining regulatory approvals are essential steps in bringing such innovations to the bedside. Collaborations between technology companies, pharmaceutical firms, and regulatory bodies will be crucial in this endeavor. Targeted Oncology
Conclusion: A Transformative Leap Toward AI-Powered Precision in Lung Cancer Prognosis
Onc.AI’s AI-driven model marks a significant milestone in the application of artificial intelligence in oncology. By enhancing the prediction of immunotherapy responses in NSCLC patients, it paves the way for more personalized and effective cancer treatments. As the medical community continues to embrace technological advancements, such innovations hold the promise of transforming patient care and outcomes.ASCO Publications
FAQ : Everything You Need to Know About Onc.AI’s AI Breakthrough in Lung Cancer Prognosis
What Is Onc.AI and How Does It Contribute to Lung Cancer Treatment?
Onc.AI is a U.S.-based healthcare technology company specializing in artificial intelligence for oncology. At the AACR 2025 conference, it revealed a new AI model that predicts how well patients with advanced non-small cell lung cancer (NSCLC) will respond to immunotherapy. By integrating this tool into standard CT scan analysis, oncologists gain a non-invasive, data-driven solution to personalize treatment. More on this can be found on their official site: www.onc.ai
How Does Onc.AI’s Radiomic Signature Model Work?
The model uses radiomics—extracting hundreds of data points from CT scans—to create a predictive signature of the tumor. These radiomic features are processed through deep learning algorithms that identify invisible patterns linked to treatment success or failure. This signature helps forecast overall survival (OS) and progression-free survival (PFS), improving treatment matching. Learn more from their ASCO 2024 presentation here.
What Was Announced by Onc.AI at the AACR 2025 Conference?
Onc.AI presented peer-reviewed findings confirming its AI model’s ability to predict immunotherapy success in lung cancer patients. The results, validated in multiple patient groups, demonstrated strong hazard ratios and performance across both clinical trial and real-world cohorts. These insights were presented at the American Association for Cancer Research (AACR) 2025 conference and covered by outlets such as BioSpace.
Why Is Immunotherapy Response Prediction So Important in Lung Cancer?
Not all lung cancer patients respond to immunotherapy, and administering it without accurate prognosis wastes time and resources. Onc.AI’s tool helps identify which patients will benefit, sparing others from unnecessary treatments and potential side effects. This enhances quality of life while optimizing the use of cutting-edge therapies like PD-1 inhibitors. A broader explanation is available at Targeted Oncology.
What Role Did Pfizer Play in Onc.AI’s Research?
Pfizer collaborated with Onc.AI to test the model on a large cohort of NSCLC patients enrolled in a clinical trial for its PD-1 inhibitor, Sasanlimab. The AI system achieved a progression-free survival (PFS) hazard ratio of 0.30 and overall survival (OS) hazard ratio of 0.29, which indicates a strong correlation between predicted and actual outcomes. This collaboration lends pharmaceutical-level validation to the AI’s clinical utility. Details were shared via BusinessWire.
What Are Radiomics and How Are They Used in AI for Cancer?
Radiomics involves extracting vast amounts of quantitative data from medical images using advanced algorithms. These data, invisible to the naked eye, provide deep insight into tumor heterogeneity, shape, texture, and evolution. Onc.AI uses this information to train AI models that outperform traditional biomarkers in predicting therapy success.
What Do the Results Say About Onc.AI’s Predictive Accuracy?
In a real-world holdout cohort, Onc.AI’s model produced a PFS hazard ratio of 0.18 in patients on single-agent immunotherapy. That means patients the model identified as likely responders were 82% less likely to progress compared to others. The model also succeeded in three independent cohorts, validating its predictive strength across treatment settings.
Can This AI Technology Be Used in Routine Clinical Settings?
Yes, one of the model’s major strengths is its compatibility with standard-of-care imaging (CT scans), meaning no new hardware or invasive procedures are needed. Clinicians can use Onc.AI’s platform as a decision-support tool to enhance treatment planning. Integration into electronic health records (EHRs) is currently being explored to streamline workflows.
Is Onc.AI’s Model Approved by the FDA?
As of AACR 2025, the model is in advanced stages of clinical validation and may be submitted for FDA Breakthrough Device designation. This status accelerates review for technologies with significant potential to improve patient outcomes. You can monitor updates directly from FDA’s Breakthrough Devices Program.
How Will This Innovation Impact Future Cancer Care?
Onc.AI’s tool signals a shift toward truly personalized oncology, where decisions are driven by predictive analytics rather than trial and error. In the future, it could also be adapted to other cancer types and treatment pathways. With mounting evidence and pharmaceutical backing, tools like this may become standard practice across oncology clinics.
For more information on Onc.AI’s research and developments, visit their official website: Onc.AI
Learn more about the AACR Special Conference on Artificial Intelligence and Machine Learning: AACR AI Conference