A Protein-Based Blood Test for Multi-Cancer Diagnostics.

Background/Objectives: Conventional cancer screening relies heavily on imaging and invasive procedures, leading to high false-positive rates and limited uptake, while leaving several high-mortality cancers without routine screening options. This study evaluated a protein-based multi-cancer early detection (MCED) test designed to detect five high-burden cancers with high sensitivity, specificity, and tissue-of-origin (TOO) accuracy. Methods: Serum from 141 patients with confirmed breast, lung, colorectal, ovarian, or pancreatic cancer and 119 healthy controls was analyzed using a 16-parameter protein biomarker panel. The assay measured extracellular protein kinase A (xPKA) activity, additional kinase activities, and cancer-associated antibodies (IgG, IgM). A supervised, rule-based classification framework was developed for cancer detection and TOO assignment. Results: The MCED test achieved 100% sensitivity across all five cancer types and 97% overall specificity, with 98% TOO accuracy. Importantly, 100% of Stage I cancers were detected. Cancer specificities ranged from 96.6% (breast) to 100% (ovarian, pancreatic, and colorectal). Conclusions: This protein-based MCED approach demonstrates exceptional performance for multi-cancer detection and TOO identification, including robust early-stage detection, and may reduce the downstream diagnostic burden relative to the existing system.