The medical establishment is currently obsessed with a narrative that sounds like pure science fiction. They want you to believe that a simple vial of blood can act as an early-warning radar system, catching the return of cancer long before traditional scans can see a thing. The media covers every minor trial with breathless enthusiasm. Startups raise hundreds of millions of dollars promising peace of mind in a purple-top tube.
It is a beautiful illusion. It is also fundamentally dangerous.
The lazy consensus dominating oncology right now suggests that finding circulating tumor DNA (ctDNA) early always translates to saving lives. It does not. Behind the marketing gloss of liquid biopsies lies a messy biological reality defined by technical limitations, crushing financial incentives, and a devastating lack of clinical utility data. We are rushing to deploy a technology before we even understand whether acting on its results does more harm than good.
I have watched clinical teams look at these ultra-sensitive blood reports and panic. I have seen millions of dollars spent treating numbers on a page rather than actual, visible disease. We are creating a new class of patients: the pre-recurrence survivors, people trapped in a medical limbo where they are told they have cancer DNA in their veins, but no doctor can find the tumor to treat it.
The Biology They Conveniently Ignore
The pitch for ctDNA monitoring relies on a deceptively simple premise. Tumors shed pieces of their DNA into the bloodstream as cells die. If we sequence your blood and find mutations matching your original tumor, your cancer is back.
But biology refuses to be that simple.
First, consider the phenomenon of clonal hematopoiesis of indeterminate potential, or CHIP. As humans age, stem cells in our bone marrow naturally acquire mutations. These mutated white blood cells spill into the bloodstream and shed their own DNA. To an amateur lab assay, a CHIP mutation looks identical to a recurring solid tumor mutation. If you have a history of breast or colon cancer, a routine blood draw might pick up a TP53 or KRAS mutation. The immediate assumption is recurrence. The reality is often just a normal byproduct of human aging.
Distinguishing between true tumor recurrence and CHIP requires dual sequencing of both the plasma cell-free DNA and the patient's own white blood cells. Many commercial tests skimp on this step because it doubles the sequencing cost. The result? Patients are subjected to terrifying, unnecessary workups based on genetic noise.
Furthermore, the rate at which tumors shed DNA varies wildly. A highly aggressive, poorly vascularized tumor might shed massive amounts of DNA into the periphery. A slower-growing, well-encapsulated recurrence might shed absolutely nothing. A negative blood test does not mean you are cancer-free, and a positive test does not automatically tell you where the fire is.
The Illusion of Early Intervention
The core argument for tracking ctDNA is that catching a recurrence early allows for earlier treatment, which leads to better outcomes. This sounds like common sense. It is actually a textbook example of lead-time bias.
Imagine a scenario where a patient has a colon cancer recurrence that will become visible on a CT scan in exactly two years. If we do regular imaging, we find it in 2028, treat it, and the patient survives until 2032. Total survival time from detection is four years.
Now, imagine we use an ultra-sensitive blood test on that same patient. The test detects a microscopic spike in ctDNA in 2026, two years before the scan would show anything. We start toxic chemotherapy immediately. The patient still dies in 2032. Total survival time from detection now looks like six years.
The enthusiast proclaims the blood test extended the patient's life by two years. In reality, the patient didn't live a single day longer. They simply spent two additional years living as a cancer patient, undergoing grueling treatments for an invisible target. The test did not change the trajectory of the disease; it merely changed the date on the diagnosis.
We have seen this play out before with PSA screening for prostate cancer and mammography for early breast cancer. We traded localized overdiagnosis for massive systemic overtreatment. With liquid biopsies, we are preparing to make the exact same mistake on an industrial scale.
The Math of a Single Drop of Blood
Let us look at the raw mechanics of a blood draw. A standard tube of blood contains roughly 10 milliliters of plasma. Within that plasma sits a tiny amount of cell-free DNA, usually measured in nanograms.
If a patient has a massive, metastatic tumor burden, tumor DNA might make up 5% or 10% of that total pool. That is easy to detect. But if we are looking for the earliest whisper of a recurrence, we are hunting for what is known as a low mutant allele fraction (MAF). We are looking for one mutated fragment of DNA floating among ten thousand normal fragments. This is a mutant allele fraction of 0.01%.
At 0.01% MAF, the laws of physics and sampling statistics begin to work against you. In a standard 10ml blood tube, there may only be two or three actual physical fragments of the mutated tumor DNA present. If the phlebotomist draws the blood a little too slowly, or if the tube sits on a counter for an extra hour before processing, the normal white blood cells will lyse. They burst open, flooding the sample with healthy genomic DNA, completely diluting those two or three precious tumor fragments.
The test comes back clean. The patient breathes a sigh of relief. Six months later, they are in the emergency room with a palpable mass. The test did not fail because the technology was bad; it failed because you cannot sequence what is not physically inside the tube.
Chasing Ghosts in the Clinic
What happens when the test is positive but the scans are negative? This is the dirty secret of the liquid biopsy boom.
When a ctDNA assay returns a positive result for recurrence, the standard clinical protocol is to order a full-body CT scan or a PET-CT. If the tumor is smaller than a few millimeters, the scans will show nothing.
Now the oncologist faces an impossible choice. Do you start systemic chemotherapy, immunotherapy, or a targeted agent based purely on a genetic reading from a blood machine? If you do, you are treating a patient without knowing where the disease is, how large it is, or whether the therapy is actually shrinking the target. You cannot measure a response if you cannot see the tumor.
Alternatively, do you tell the patient, "The blood test says your cancer is back, but we cannot find it, so we are just going to wait three months and scan you again"?
The anxiety this inflicts is predatory. It ruins whatever quality of life the survivor had left. They are forced to live with a ticking time bomb inside their minds, waiting for the shadow to finally appear on an X-ray so their doctor can justify treating them.
Follow the Money
We must look at who benefits from the rapid adoption of these tests. The development of next-generation sequencing assays is an incredibly lucrative business. Companies charge upwards of $4,000 per test. They want these assays integrated into standard surveillance guidelines, meaning every cancer survivor would get tested three to four times a year for five years.
Do the math. That is tens of thousands of dollars per patient, scaled across millions of survivors globally. The financial incentive to push these tests into the clinic far outpaces the scientific evidence validating them.
The heavy hitters in clinical trial design are trying to bring sanity back to the discussion. Ongoing trials are finally asking the right question: Does changing treatment based on ctDNA detection actually improve overall survival?
Until those trials read out, we are flying blind. We are buying into the marketing hype of tech companies that view healthcare through the lens of software updates rather than human physiology.
A Hard Pivot to Real Utility
Am I saying we should throw liquid biopsies in the trash? No. The technology is phenomenal for specific, defined use cases.
If a patient already has known, visible metastatic disease, a blood test is a fantastic way to identify targetable mutations without forcing the patient to undergo a painful, invasive tissue biopsy. It is also an excellent tool for monitoring whether a known tumor is responding to a specific drug in real-time. If the ctDNA drops precipitously after a week of therapy, you know the drug is working.
But using it as an early warning system for recurrence in healthy survivors is a misuse of the tool.
Stop asking if a blood test can tell patients their cancer is coming back. It can. The real question is whether knowing that information earlier changes the ultimate outcome, or if it just stretches out the period of human suffering. Right now, the evidence points firmly toward the latter.
We must stop treating numbers on a laboratory printout and start treating the actual human being sitting in the clinic. Turn off the hype machine. Wait for the survival data.
