I first heard the phrase “minimal residual disease” (MRD) in the audience at the 2026 Targeted Therapies of Lung Cancer (TTLC) meeting. Max Diehn, MD, PhD, stood at the podium, posing a question that I hadn’t known to ask during his presentation titled MRD in the Management of Early and Locally Advanced NSCLC; Are we There Yet?
The phrase landed with a striking duality. Minimal sounded reassuring. Residual did not.
Together they described something I had long suspected: the possibility that cancer persists even after clear margins are declared or imaging shows no evidence of disease.
It was a suspicion that had followed me throughout my own treatment. Three years after a lobectomy for a stage I lung cancer, a second diagnosis was made. Hoping to conserve lung tissue, I chose stereotactic ablative radiotherapy (SABR) over another surgery. Despite my oncologist’s confidence in the stopping power of 60 Grays, I never shook the fear that some daughter cells had been missed.
Already rattled by the return of the disease, the uncertainty of whether we had beaten the cancer tipped me into a clinical depression. Two years later, a third stage I tumor was found and treated with surgery. Newly available biomarker testing revealed a driver oncogene. If molecular testing had been available and used in my case, it may have identified the cancer earlier and informed more tailored surveillance or decisions about adjuvant therapy. The state of the science today could have altered the trajectory of my cancer journey.
MRD refers to cancer cells left in a patient’s body after treatment of the original lung cancer. They are undetectable by conventional imaging and are a harbinger of recurrence. With the advent of circulating tumor DNA (ctDNA) assays, it is now possible to identify lung cancer before it becomes radiologically visible.
The approach is individualized. Tumor tissue is sequenced to identify patient-specific mutations, which are then used to create a custom diagnostic molecular test. The personalized assays can detect those same mutations among thousands of normal cells in subsequent blood draws.
Sensitivity of the assays is improving rapidly. Detection thresholds have advanced from one mutant cell in 10,000 with single-strand sequencing to one in 250,000 with duplex sequencing and now approach one in a million with phased variant techniques. The signal is small, but the implication is large.
Why aren’t we there yet?
MRD testing is covered by Medicare because the FDA has accepted evidence that the assays can detect cancer before imaging can; however, its role in guiding therapy remains under study. While current MRD assays are excellent at confirming when cancer is present, there can be false negatives. Improvements in sensitivity may address this, but for the time being, there is a lack of evidence that MRD-negative values can be used to de-escalate adjuvant therapy. There are no NCCN Guidelines for MRD in early-stage lung cancer. It is considered an evolving area within the “clinical trials” context of surveillance.
MRD introduces the possibility of forewarning. A positive result signals an elevated risk of recurrence. A negative result, while not definitive, offers greater reassurance than imaging alone. In both cases, uncertainty remains, but it is no longer unstructured.
That distinction matters more to patients than clinicians may appreciate.
For patients, the value is clear. We wake at 3 am and inventory sensations. We interpret a cough, a twinge, or ambiguous ground-glass on a scan report. The absence of data does not create peace; it creates narrative. And the narratives we create are often worse than the probabilities physicians hesitate to share.
We want to know anyway.
The difference lies in how value is defined. In medicine, information is often judged by whether it changes treatment. But patients do not experience information solely through its effect on treatment decisions. There is another form of value at work here, one that is less often raised in clinical discourse. Call it epistemic value: the value of knowing something meaningful about one’s condition even when that knowledge does not immediately translate into intervention.
MRD testing offers that kind of knowledge. It does not eliminate uncertainty, but it changes its structure. In the absence of MRD data, the time between scans is undifferentiated. A three-month interval is simply waiting. Recurrence, when it appears, often feels abrupt—not because it is biologically sudden, but because it has been invisible.
Anxiety in survivorship is not driven solely by risk. It is driven by ambiguity. Many patients would choose a quantified, even unfavorable probability over an undefined possibility.
Knowing a patient is MRD-positive can shape prognosis discussions, influence adjuvant therapy, and aid interpretation of equivocal imaging findings.
Knowing my MRD status changes how I inhabit time. It replaces a binary narrative–disease or no disease–with a probabilistic one.
The question, then, is not only whether MRD should guide treatment. It is whether we are prepared to recognize the value of knowing before we can act.
For those of us who live in the interval between treatment and recurrence, that value is already clear.
