Modern precision oncology has built its foundation on DNA-based biomarkers — mutations, tumor mutational burden, and specific oncogenic drivers like EGFR or ALK. Yet the data on treatment response tells a more complicated story. Targeted therapies aimed at known driver mutations achieve a primary non-response rate of roughly 10–20%, but nearly all patients eventually develop acquired resistance. Immunotherapy, meanwhile, is far less predictable: used alone, it fails to help 40–60% of patients from the outset, largely because PD-L1 expression and tumor mutational burden are imperfect predictors of who will actually respond. Chemotherapy, the oldest tool in the arsenal, sees non-response rates of 70–80% and near-universal eventual resistance. Even combination immunotherapy-chemotherapy regimens, while improving on single-agent immunotherapy, still leave 50–55% of PD-L1-low or negative patients without adequate response. Across every therapy class, a substantial population of patients is left without a reliable way to predict, in advance, whether treatment will work — these are the “invisible patients”.

A tumor’s genetic mutation profile does not necessarily reflect how its genes are actively expressed, how the tumor microenvironment is behaving, or how a specific patient’s biology will metabolize and respond to a drug in real time. This is precisely the gap that OncoDxRx’s PGA (Patient-derived Gene expression-informed Anticancer drug efficacy) platform is designed to address. For these invisible patients, a blood-based, gene-expression-informed efficacy test offers a less invasive, potentially more dynamic way to guide treatment selection than repeat tissue biopsies or DNA panels alone, tools like PGA represent an attempt to close that gap — not by replacing DNA-based diagnostics, but by adding the functional, expression-level data needed to make invisible patients visible again.

The PGA assay operates on a fundamentally different biological premise: that the “hyperactive” transcriptome encodes functional identity and drug susceptibility with a fidelity that somatic mutations alone cannot capture. Where DNA catalogues WHAT A TUMOR CAN DO, RNA reports WHAT IT IS DOING. This distinction is clinically decisive, PGA does not require the patient to carry a targetable mutation — it requires only that the tumor be overexpressing genes, which every living tumor inevitably does.