A new blood test for colorectal cancer is supposed to catch the disease early. Instead, it's better at finding full-blown tumors than the growths that turn into tumors. This is backwards.
The conventional wisdom about screening is almost sacred: catch disease in its precancerous stage and you've won. Find a polyp, remove it, prevent cancer from ever forming. This is why colonoscopies exist. This is why we poke around inside people's colons in the first place. The whole idea is that screening works by intercepting the disease before it becomes serious. A test that does the opposite—that's almost useless as prevention, even if it's technically good at detecting cancer.
Yet this is exactly what the Shield blood test does, according to its FDA approval data. The test, developed by Exact Sciences, detects colorectal cancer in 83% of cases. That's genuinely impressive. But it catches precancerous polyps in only 13.2% of cases. This inverted performance profile suggests something fundamental is working differently than expected. According to reporting on the test's 2024 approval status, these numbers represent a genuine scientific puzzle: why would a blood biomarker be so much better at identifying advanced disease than at catching it early?
The mechanism here is almost certainly biological rather than technical. Precancerous polyps are small, slow-growing lesions that haven't necessarily begun shedding cancer cells into the bloodstream in large quantities. A blood test, by definition, only detects markers that are actively circulating. Early polyps might be invisible to blood sampling simply because they haven't begun broadcasting their presence systemically. By contrast, established cancers actively shed more tumor DNA and proteins into circulation—they're louder signals. This means Shield isn't actually failing at detection; it's bumping up against the physical limits of what blood-based screening can see. You can't find what isn't in the blood yet.
This reveals a deeper tension in the cancer screening ecosystem. We have tools optimized for different purposes pretending to be universal. Colonoscopies are brilliant at finding precancers because they're visual and invasive. Blood tests are convenient but inherently limited to detecting circulating markers. Shield is honest about this tradeoff in a way that marketing sometimes isn't: it's a test for cancer risk stratification in people with symptoms or high risk, not a replacement for colonoscopy-based prevention. It might help catch the cancers that slip through, but it won't prevent as many as it detects.
The uncomfortable implication is that no single test will revolutionize colorectal cancer prevention unless it can somehow see into tissue directly—which blood tests fundamentally cannot do. Shield might actually be most useful not as primary screening but as a triage tool: a cheap, easy way to decide who needs a colonoscopy urgently. That's valuable. It's just not the narrative of early detection heroics that gets venture capital excited.