Tempus’ AI-Driven IPS Test Yields HR 0.45 Versus TMB, MSI, PD-L1
Tempus’ Immune Profile Score (IPS) test achieved a hazard ratio of 0.45 across four independent pan-cancer metastatic cohorts, outperforming TMB, MSI, and PD-L1 in predicting immunotherapy outcomes. IPS identified 13% of microsatellite stable colorectal cancer cases (HR=0.2) and classified 17% of rare cancer cases as IPS-High, highlighting expanded immunotherapy eligibility.
1. Study Validates Superior Predictive Power of IPS
Tempus AI’s Immune Profile Score (IPS) was evaluated in four independent real-world cohorts of patients with metastatic solid tumors receiving immune checkpoint inhibitors (ICIs). The study reported a hazard ratio (HR) of 0.45 for overall survival in IPS-High versus IPS-Low patients, demonstrating statistically significant prognostic utility that outperforms conventional biomarkers including tumor mutational burden, microsatellite instability status and PD-L1 expression. IPS combines DNA- and RNA-derived immune signatures with clinical features in a single multimodal algorithm, yielding more reliable identification of patients likely to benefit from ICI therapy across diverse cancer types.
2. Identification of MSI-Stable Colorectal Cancer Responders
Among patients with microsatellite-stable colorectal cancer—typically considered poor candidates for immunotherapy—IPS classified 13% as IPS-High. This subgroup exhibited a remarkable HR of 0.20 for real-world overall survival compared to IPS-Low counterparts, suggesting that a meaningful fraction of colorectal cancer patients overlooked by standard biomarkers could derive substantial benefit from ICIs. These findings indicate IPS’s potential to expand the immunotherapy-eligible population in colorectal cancer, guiding physicians toward treatment options that would otherwise be discounted.
3. Expanding Immunotherapy Access in Rare Tumors
The study further assessed IPS performance in a cohort of patients with rare metastatic solid tumors not covered by existing FDA-approved ICI indications. IPS designated 17% of these patients as IPS-High, and this group demonstrated a hazard ratio of 0.26 for overall survival relative to IPS-Low patients. Such results underscore IPS’s ability to uncover hidden responder populations in under-studied malignancies, offering a data-driven rationale to consider ICIs beyond current label restrictions and potentially accelerate clinical trial enrollment for these rare cancer subtypes.
4. Seamless Integration into Clinical Workflows
Tempus AI has made IPS available as an optional add-on for clinicians ordering its xT (DNA) and xR (RNA) sequencing assays. By leveraging data already generated during routine tumor profiling, IPS can be reported alongside existing genomic results without additional biopsies or patient burden. This integration supports rapid turnaround times and empowers oncologists with a single, comprehensive biomarker report—driving more confident treatment decisions and advancing the adoption of precision immunotherapy in everyday oncology practice.