Date: Jan 10, 2016 Author: ANDREW POLLACK Source: New York Times (
click here to go to the source)
Some people see it as the holy grail in oncology — a simple blood test that can detect any kind of cancer at an early stage, when it is easiest to treat and possibly cure.
Illumina, the world's largest maker of DNA sequencing machines, said on Sunday that it was forming a company to attempt to develop such a test.
The company, called Grail, has raised over $100 million, mostly from Illumina and the venture capital firm Arch Venture Partners, but also from Microsoft's co-founder, Bill Gates, and Jeffrey P. Bezos, the founder and chief executive of Amazon. Grail, which is majority-owned by Illumina, has also assembled a prominent roster of advisers.
"If this pans out, this could be a real game changer," said Dr. José Baselga, the physician in chief at Memorial Sloan Kettering Cancer Center, who will head the company's science advisory board.
But some experts said that developing such a test would be a daunting task. Even Dr. Baselga seemed skeptical that a pan-cancer screening test could be ready by 2019, the goal of Illumina's chief executive, Jay T. Flatley. Screening for particular types of cancer may be easier, Dr. Baselga said.
Photo
Illumina's chief executive, Jay T. Flatley.
The approach to be used takes advantage of the fact that DNA from cancerous cells can be found in the blood, albeit in minute amounts. The blood test, using intensive DNA sequencing, would look for patterns of telltale mutations.
Many companies are developing similar tests, known as liquid biopsies, that use a blood sample to replace or supplement a conventional biopsy, in which a sample of the tumor is extracted from the body by needle or surgery. In just the last week, one company, Guardant Health, said it had raised nearly $100 million in its latest funding round while another, Exosome Diagnostics, said it had raised $60 million.
But so far, liquid biopsies are mainly used for patients already known to have cancer. The tests help determine the particular mutations in the tumor to help select the best drugs to use. Or they are used to monitor whether treatment is effective or to check for a recurrence.
Detecting early-stage cancer, although an eventual goal of many liquid biopsy companies, is considered more difficult. Many cells, not just cancerous ones, shed DNA into the bloodstream, and many noncancerous cells have mutations that are also found in tumors.
"As you age, you have mutations," said Dr. Luis A. Diaz Jr., associate professor of oncology at Johns Hopkins and co-founder of three companies developing liquid biopsies. Polyps, moles and benign growths also have mutations that resemble those in tumors, he said.
Moreover, Dr. Diaz said, current liquid biopsy technology is often not sensitive enough to pick up multiple mutations at once.
Yet another issue is that some early-stage cancers will never actually cause harm. But once the cancers are found, patients are often treated and exposed to side effects from the treatments. This is an issue that has driven a debate about whether some forms of screening, like PSA testing for prostate cancer or mammography, are even advisable.
"Patients ought to be hesitant until there is really good data that this actually helps people, and they should remember that it could harm people," said Dr. H. Gilbert Welch, professor of medicine at Dartmouth and author of the book "Less Medicine, More Health."
One company, Pathway Genomics, last year began offering a liquid biopsy test aimed at detecting early-stage cancer. The Food and Drug Administration almost immediately challenged the company, saying the test needed regulatory approval. The agency, in a letter to the company in September, also said that it had not found "any published evidence that this test or any similar test has been clinically validated as a screening tool for the early detection of cancer."
Mr. Flatley said that Grail would overcome those problems by doing intensive DNA sequencing on 30,000 to 50,000 people over time, some of whom would have, or would develop, cancer, to figure out the mutation signatures indicating cancer.
The screening test for each patient would also use the "deep," or intensive, sequencing, to check for numerous mutations. Mr. Flatley said a screening test would have to cost no more than $1,000 to be economically feasible.
Sequencing a whole human genome now costs about $1,000 for big users of Illumina's sequencing machines. But costs are dropping and Grail, which does not have a chief executive yet, will get a special discount from Illumina.
"None of our customers can afford to do this today — only Illumina can," Mr. Flatley said in an interview. "We have a sense of urgency to make it work."
