SageMedic Corp.

For about two decades, the ruling idea in cancer diagnostics has been to read the genome. Sequence the tumor, find the mutations, match a drug to the mutation. It is a genuinely powerful idea, and it works - some of the time. The uncomfortable footnote is how often "some of the time" turns out to be. By SageMedic's accounting, genomic tests hand an oncologist an actionable answer in roughly 26% of cases. The other three-quarters of patients get a shrug dressed up as data.

SageMedic Corp., a small company operating out of Redwood City, California, has spent its existence pushing a different and slightly heretical question. Instead of predicting what a drug will do to a tumor, why not just watch what it does? Take a fresh biopsy, grow hundreds of live three-dimensional microtumors from the patient's own tissue, expose those microtumors to a panel of FDA-approved chemotherapies and targeted drugs, and observe which ones the cancer resists and which ones kill it. This is called functional precision oncology, and it is the entire premise of the company's flagship product, the SAGE Oncotest.

The appeal is almost embarrassingly intuitive. Genomics tells you what might work; a functional assay shows you what does. And because the test measures behavior rather than inferring it from mutations, it returns something usable far more often. SageMedic puts that number at 88%. The whole thing runs in 7 to 10 days.

Now, one should be careful here. A company reporting its own headline number is not the same as a peer-reviewed randomized trial, and "actionable" is doing some quiet work in that sentence - it means the test produced a clear result a doctor can act on, not that the patient was cured. Functional testing is an old idea that has been tried, in various forms, for decades; the technology's history includes plenty of assays that looked promising and did not pan out. What is different now is the throughput. Growing hundreds of microtumors that preserve a tumor's heterogeneity and its extracellular matrix, and screening them fast, is the kind of thing that has become plausible only recently. SageMedic is betting the plausibility has finally caught up to the intuition.

SageMedic is, in the most literal sense, a family business born from a family loss. Dr. Chris Apfel's mother died of ovarian cancer. His father, diagnosed later with lung cancer, chose to forgo treatment. Apfel - already an accomplished physician-scientist - was, in the company's telling, struck by how limited the progress in cancer treatment had been, and set out to find a more effective approach. He founded the company with his wife, Dr. Brigitte Apfel, a pulmonologist who has focused on lung cancer.

Apfel is not a first-time inventor. Earlier in his career he built a six-factor clinical prediction model, published in the New England Journal of Medicine, that became the Apfel Score - a calculator used in operating rooms worldwide to predict the risk of postoperative nausea and vomiting. He left a clinical department at UCSF, picked up an MBA from Wharton, and turned to cancer. There is a neat symmetry in it: his first famous contribution was a tool for predicting who would get sick from anesthesia; his second act is a tool for figuring out which drug will make a tumor sick instead of the patient.

At the American Association for Cancer Research Annual Meeting in April 2025, SageMedic presented a poster with a deceptively simple title: "Heterogeneity of ex-vivo Tumor Responses in Primary Ovarian Cancer Tissues." The finding underneath it is the whole thesis in miniature. Primary ovarian tumors, tested against standard therapies like carboplatin and paclitaxel, responded in substantially different ways from one another. If every ovarian cancer is biologically distinct, the argument goes, then treating them all with the same default regimen is leaving information - and outcomes - on the table.