Tempus AI's IPS Test Outperforms Standard Biomarkers in Predicting Immunotherapy Response
核心洞察
Tempus AI's Immune Profile Score (IPS) test demonstrated superior accuracy in predicting immune checkpoint inhibitor outcomes compared to conventional biomarkers including tumor mutational burden (搜索), microsatellite instability (搜索), and PD-L1 (搜索) across four independent validation cohorts.
The multimodal biomarker identified 13% of microsatellite stable colorectal cancer (搜索) patients who showed strong real-world overall survival with immunotherapy treatment, potentially expanding treatment options for previously overlooked populations.
IPS classified 17% of rare metastatic solid tumor patients as "IPS-High" despite not meeting cancer-specific FDA-approved immunotherapy labels, showing significant survival differences that could guide treatment decisions.
Tempus AI announced results from a new clinical validation study demonstrating that its algorithmic Immune Profile Score (IPS) test more accurately predicts patient outcomes with immune checkpoint inhibitors (搜索) than conventional biomarkers. The multimodal biomarker combines clinical and immune-related markers from DNA and RNA analysis to predict immunotherapy response, potentially transforming patient selection for cancer treatment.
Superior Performance Across Multiple Cancer Types
In four independent validation cohorts of pan-cancer metastatic solid organ cancer patients, IPS demonstrated superior prognostic utility with a hazard ratio of 0.45 compared to tumor mutational burden (搜索) (TMB), microsatellite stability (MSS), and PD-L1 (搜索). The test showed prognostic value independent of these conventional biomarkers, suggesting it captures additional biological information relevant to immunotherapy response.
The study revealed particularly striking results in microsatellite stable colorectal cancer (搜索), where IPS identified 13% of patients who demonstrated strong real-world overall survival with immune checkpoint inhibitor treatment (HR=0.2). This finding is significant as this patient population might have been overlooked using conventional biomarkers alone, potentially expanding immunotherapy access to patients who could benefit from treatment.
Expanding Treatment Options for Rare Cancers
Beyond colorectal cancer (搜索), the study highlighted IPS's potential impact on rare cancer treatment. The test classified 17% of patients with rare metastatic solid tumors (搜索) as "IPS-High," despite these patients not falling within cancer-specific FDA-approved immunotherapy labels. The significant difference in median real-world overall survival between "IPS High" versus "IPS-Low" patients (HR=0.26) indicates that immunotherapy could be a relevant and potentially life-saving option for patients who might otherwise be missed by current selection criteria.
"This study builds on previous findings showing that IPS scoring can identify patients who may have better overall survival with ICI therapy. These results are a significant step forward in the personalization of cancer care, particularly for patients that might be missed by conventional biomarkers," said Halla Nimeiri, MD, Chief Development Officer at Tempus.
Clinical Implementation and Accessibility
The IPS test is available as an add-on for clinicians ordering Tempus' xT (DNA) and xR (RNA) assays, utilizing data already collected as part of standard sequencing workflows. This integration approach allows physicians to access enhanced immunotherapy prediction capabilities without requiring additional tissue samples or separate testing procedures.
Dr. Nimeiri emphasized the clinical utility of the multimodal approach: "By leveraging a multimodal algorithm, our IPS test provides a more accurate and comprehensive view than traditional, independent biomarkers, empowering physicians to make more confident treatment decisions for their patients while simultaneously moving the field forward in meaningful ways."
Technology and Data Integration
Tempus positions itself as a technology company advancing precision medicine through artificial intelligence applications in healthcare. The company maintains one of the world's largest libraries of multimodal data and operates a system designed to make that data accessible and useful for clinical decision-making. The IPS test represents an application of this AI-enabled approach to provide personalized treatment guidance based on comprehensive molecular profiling.
The validation results demonstrate how algorithmic approaches to biomarker development may offer advantages over single-parameter testing, particularly in the complex landscape of immunotherapy response prediction where multiple biological factors influence treatment outcomes.