Illumina Spinoff GRAIL to Trial Liquid Biopsies for Early Detection of Cancer

By Aaron Krol

January 11, 2016 | Illumina, the dominant manufacturer of DNA sequencing technology, has formed a spinoff company called GRAIL to experiment with a simple blood test for early detection of all types of cancer.

Illumina, whose instruments accounted for more than 90% of all DNA data collected worldwide in 2014, already has its fingers in many of the most lucrative medical uses of genomics. The company’s line of verifi tests, acquired with Verinata Health in 2013, tap into the fast-growing market for non-invasive prenatal tests, which detect fetal disorders through a blood draw from the pregnant mother. A collaboration with several large pharma companies has also produced a panel test called TruSight Tumor 15, which searches for mutations that could present new treatment options in cancer cases.

GRAIL, however, will be the first independent business Illumina has created to pursue a still-nascent clinical program. Its goal, said Illumina’s former CMO Richard Klausner, now an executive at GRAIL, on a conference call yesterday, will be to develop “a universal, direct measure of cancer,” far more accurate than proxy tests like mammograms or PSA tests for prostate cancer.

The technology involved is liquid biopsy, in which DNA and RNA molecules shed by tumors are picked up circulating freely in the bloodstream. By extremely deep sequencing, or by isolating tumor nucleic acids from other genetic materials in the blood, researchers in both academia and private industry have shown it is possible to find telltale signatures of cancer in simple blood draws. Companies including QIAGEN, RainDance, and Genomic Health are all working to commercialize liquid biopsies as alternatives to more invasive cancer testing.

None, however, have demonstrated that liquid biopsies can accurately predict the emergence of malignant cancer at an early stage.

GRAIL will have access to enormous resources to experiment with liquid biopsy as a preventative test in stage one and two cancers. The company has raised more than $100 million at launch from ARCH Venture Partners, Bezos Expeditions, Sutter Hill Ventures, and Illumina itself, which remains the majority owner. GRAIL may also find it cheaper to do ultra-high-throughput sequencing than other companies in the liquid biopsy space, virtually all of which buy their instruments and reagents from Illumina―and given the multi-billion-dollar opportunity for an early stage pan-cancer test, and the clout that comes from its Illumina ties, GRAIL will likely have little trouble raising new rounds of funding if necessary.

The bar for releasing a test of this nature, however, is daunting. Early testing of any kind is more likely to pick up benign cancers, never destined to cause health problems, than young malignant tumors that demand prompt treatment. For GRAIL’s liquid biopsy to benefit patients, the company would need to uncover genetic markers that can distinguish harmful cancers from benign ones with exquisite specificity―and find these markers in datasets where tumor DNA may represent as little as one out of every 100,000 sequencing reads.

To demonstrate the accuracy of such a test, GRAIL will need to mount clinical trials on a scale normally reserved only for post-market studies of ultra-common medications like statins or aspirin. In the conference call announcing the launch of GRAIL, José Baselga, the current President of the American Association for Cancer Research, Physician-in-Chief at Memorial Sloan Kettering Cancer Center, and now Chair of GRAIL’s Advisory Board, said that the company plans to launch an “unprecedented, large population-based study” in which hundreds of thousands of healthy patients may be enrolled. Each participant would be regularly tested with GRAIL’s technology over a period of months or years, until enough had developed cancer to gauge the test’s real-world specificity and sensitivity.

This massive trial is likely the only form of study that could satisfy the FDA that an early pan-cancer test is accurate and safe. As Illumina CEO Jay Flatley acknowledged, a liquid biopsy of this nature will need to have an “extraordinarily low false positive rate.”

That’s something Illumina itself has struggled with, as some researchers have raised concerns about false positives in non-invasive prenatal tests like verifi. These tests, like liquid biopsies for cancer, seek out signals from low-abundance free-floating DNA in the blood; their specificity has not been assessed in formal clinical trials in the U.S., because they are sold as non-diagnostic screening tests through a non-FDA-regulated laboratory testing structure.

Illumina has worked successfully with the FDA in the past, winning a historic clearance in 2013 for its MiSeqDx sequencer and associated tests for rare genetic disease. However, the FDA did not demand a prospective clinical trial for that instrument. More recently, Illumina has begun co-sponsoring clinical studies with Memorial Sloan Kettering for liquid biopsies in breast and lung cancers, but these studies are still in very early stages. They will now be merged into GRAIL’s larger effort.

GRAIL will spend the duration of 2016 choosing a panel of genetic markers that the company feels have the best chance of identifying malignant cancers early. “There is no more powerful vision for reducing cancer mortality around the globe than early detection,” said Klausner on the conference call.

The company hopes to begin clinical trials in 2017.