The big story out of the American Society of Clinical Oncology meeting over the weekend came from a study about so-called “liquid biopsies.” Major news organizations headlined it as follows:
Washington Post: ‘Liquid biopsy’ study offers hope for a blood test to find cancer
TIME magazine: A blood test for cancer gets closer
I think many if not most news consumers are likely to draw the wrong conclusion from these messages.
As these stories later point out with varying levels of urgency, the study involved blood samples from about 15,000 people already known to have cancer. And the researchers compared results from these samples with biopsy tissue samples that were available for about 400 patients. The point was to compare the profile of the tumor as assessed by the blood sample with the existing surgical biopsy sample.
The researchers were not trying to “find cancer” as the Post misleadingly suggests. Nor did the study involve a “blood test for cancer” as TIME states. The cancer had already been found and diagnosed using a surgical biopsy. (Despite the hyped suggestions of some companies marketing these tests, the idea that blood tests can reliably find cancer in healthy people without a biopsy is nowhere close to being a reality.)
Yes, the study does point to possible use for these blood tests in monitoring tumors after an initial diagnosis — a point that the TIME magazine story makes in its closing paragraph. And the Post acknowledges, 10 paragraphs in, that surgical biopsies will remain the “gold standard.” But whether the tests represent a realistic “alternative to painful biopsies,” as the New York Times suggests, is certainly open to question and could have used some clarification up front.
To its credit, the New York Times was the only outlet that I saw which noted that for about 15 percent of the patients, no tumor DNA was detected in the blood at all. And even a consultant to the company that makes the test admits that that’s a significant concern. “There are simply tumors that do not shed DNA into circulation at detectable levels, so we are bound to miss them,” he commented.
Most stories that I read didn’t scratch the surface of major questions and concerns that still surround this technology and its clinical application.
For example, even if the tests work as advertised for monitoring cancer relapse, how do we know that this information actually leads to better outcomes that people care about like improved survival? That’s an issue raised by HealthDay’s coverage but none of the other stories that I perused. (HealthDay’s headline was also much more informative and accurate than most others: “Liquid Biopsy May Help Doctors Track Changes in Tumors.”)
While it might seem like earlier treatment of cancer relapse can only be a good thing, that’s not necessarily proven, as oncology researcher Mike Thompson, MD, PhD, of Aurora Healthcare in Milwaukee, pointed out to me. He noted that in ovarian cancer, earlier treatment based on raised levels of a cancer biomarker, CA125, didn’t prolong survival compared with treatment initiated based on a clinical exam.
Richard Hoffman, MD, told me he was concerned that these tests would eventually be marketed for general cancer screening — an application that many companies are indeed working on.
I would then wonder about false positive results—finding DNA from tumors that the immune system would eliminate. The New York Times article notes that these liquid biopsies are not currently used to diagnose cancer but rather to monitor disease progression—but that’s how PSA [prostate specific antigen] testing was introduced.
I’m also wondering about selection bias—why were biopsies available for only 400 patients when the company had 15,000 samples to work with?
Only a large clinical trial can truly tell us how accurate and effective the test might be for improving patient care, but that’s another important concern that the stories I read did not call attention to.
As breast surgeon Deanna Attai, MD, told me, “There’s lots of promise, but it’s far from ready for prime time and not likely to eliminate the need for tissue biopsy in the near future.”