Is it all gone?
For countless cancer patients who have undergone treatment, the question of recurrence may never fully leave. This doubt can jockey away the reins of a patient’s life, dragging them along an endless cycle of waiting, scanning, and hoping.
The emotional toll of cancer surveillance cannot be overstated. “Unless you are a physician or a cancer patient, many people do not realize that patients often live under a cloud of uncertainty throughout their entire journey,” Richard Chen, MD, executive vice president of R&D, and chief medical officer, told Inside Precision Medicine. “This uncertainty includes questions like whether the cancer is still present after treatment, if there are small traces remaining, and whether the treatment is effective.”
Christopher Hall, chief executive officer and president of Personalis, has experienced this psychological purgatory for years with his dad, who is 16 years into cancer and recurrence testing with imaging. “Every six months, we inhale deeply as they perform the scans,” Hall told Inside Precision Medicine. “We live with fear because we know they’re not very sensitive. When it does come back, it’s probably going to be too late.”
The growing field of minimal residual disease (MRD) testing changes that dynamic—not by eliminating fear, but by replacing uncertainty with information. By detecting tiny fragments of circulating tumor DNA (ctDNA) in the bloodstream, these tests offer molecular evidence of whether residual disease remains—or whether new disease is emerging. A negative ctDNA test can offer far more reassurance than a clean scan; a positive test can provide an early warning long before that cancer becomes visible.Â
Now, new data from Personalis suggest that the company’s ultra-sensitive NeXT Personal test can detect recurrence in non-small cell lung cancer (NSCLC) far before radiographic imaging. The results, published in Cell, are based on one of the biggest and most comprehensive MRD datasets ever assembled in the field of lung cancer, and they show that molecular relapse occurred in some cases years before radiographic detection.
Chen said, “We can pick up recurrence well before imaging. The median lead time, how much earlier you pick up traces of cancer before it becomes positive on imaging, was about five to nine months depending on the situation. Remarkably, there were some patients that were positive on NeXT Personal years before they showed up positive on imaging. Cancer is a race against time, so any advantage you have in potentially intervening and changing direction earlier could essentially mean the difference between life and death for the patient.”
Extending lead time length
Although ctDNA analysis has become a powerful tool for several cancer types, lung cancer has remained one of the most difficult contexts for MRD detection. The most prevalent subtype, adenocarcinoma, belongs to a category of low-shedding tumors that release only small amounts of tumor-derived DNA into the bloodstream. These limited signals pose a major barrier to assays that cannot reliably distinguish rare tumor fragments from background noise.
Personalis designed NeXT Personal to address precisely this problem, aiming to overcome the limitations that have affected earlier MRD technologies. The company’s approach reaches one part per million sensitivity through a combination of whole-genome tumor sequencing, broad mutation targeting, and computational signal suppression. Each test begins with whole-genome sequencing of the patient’s tumor, allowing the identification of up to 1,800 mutations that are unique to that particular cancer.Â
The collaboration between researchers from the University College London Cancer Institute, the Francis Crick Institute, and Personalis draws on the TRACERx cohort—431 NSCLC patients and nearly 3,000 plasma samples collected across diagnosis, surgery, adjuvant therapy, and multi-year follow-up. Its longitudinal density provided a rare view of how ctDNA fluctuates across the entire patient journey.
The study reported that NeXT Personal detected recurrence well before imaging. The median lead time was about five to nine months depending on the situation, as lead time is heavily influenced by the frequency of ctDNA sampling, stage distribution of patients, and the cadence of clinical follow-up within a study.
For example, the lead time between ctDNA detection and relapse ranged from zero to 1,732 days (median 158 days). Patients with landmarks—ctDNA detection data at time points between days 10 and 120 postoperatively, prior to the start of adjuvant therapy or disease recurrence—exhibited the longest lead time for molecular relapse (median 259 days; range 11–1,430 days).
The study also reported that the detection and quantification of ctDNA even before surgery carries significant prognostic value. Patients who were ctDNA-positive at diagnosis were far more likely to relapse years later. Chen, emphasizing this multi-stage utility, said, “Our findings demonstrate the test’s usefulness at every stage for the patient—before surgery, immediately following surgery, and during follow-up—and that over time, the patient can be tested periodically. The aim is to catch the recurrence as early as possible.”
This expanded lead time could change how recurrence is managed. Identifying relapse at a much earlier stage raises the possibility of intervening when disease is still minimal and potentially more responsive to treatment. If molecular methods can reveal relapse long before imaging, clinicians may gain an opportunity to act earlier and patients may gain more time to plan or adjust therapy. “The first time you get imaging is often when you become symptomatic,” Chen said. “But you can imagine—when you reach that point, it’s a pretty serious occurrence. If you can catch it earlier…before imaging has been done, it could be a real life-changer.”
Trusting negatives
One of the most clinically meaningful applications of MRD testing lies in guiding adjuvant therapy choices after surgery. Many early-stage cancer patients face a difficult decision: whether to undergo additional therapy that carries substantial toxicity and side effects, even if it is unclear whether residual disease remains. “These studies are essential,” Chen said. “Not just showing that you can detect the cancer but that if you use it to make a clinical decision, the patients do better long term.”
Because ctDNA testing relies on liquid biopsies (i.e., blood draws) rather than imaging, it can be performed more frequently and with fewer logistical barriers. This makes ctDNA testing well suited for ongoing monitoring across multiple years, especially in patients who do not undergo regular post-treatment imaging.
Once the initial tumor has been sequenced and the personalized assay is built, ongoing testing requires only blood draws. This stands in contrast to CT or PET imaging, which involves radiation exposure, scheduling constraints, and significant costs. Chen said, “Every time you go get a CT scan, there’s radiation involved, but the logistics of getting a blood test are just straightforward. At the very beginning, to develop a test, we need a sample of the original tumor. But that’s just the beginning. After that, we’re talking about years of monitoring and testing.”
Some cancer types, such as breast cancer, often do not involve routine surveillance imaging at all, leaving patients unaware of recurrence until they develop symptoms. “Being able to have a lot more confidence in your health means a lot to patients,” Hall said. “If it does come back, being able to know that you grabbed it at one of the earliest possible times…is critical to extend life.”
The repeatability of testing also helps reduce uncertainty: highly sensitive assays that consistently return negative results provide meaningful reassurance that even tiny amounts of residual cancer are unlikely to be present. “Having a high-sensitivity test, like NeXT Personal, serves two purposes: first, it helps catch the tumor as early as possible when it is in its smallest state, and second, it allows for greater trust in a negative result,” said Chen. “If the test is not sensitive, a negative result cannot be trusted. Having a test that kind of says, look, right now you’re good, makes a big difference for people when it’s negative. If it’s positive, then at least you’re doing what you can to get out of it as soon as possible.”
Raising standard of cancer care
For now, Personalis is focused squarely on oncology—breast, lung, and colorectal cancer, along with therapy monitoring—and not other areas where liquid biopsies are blossoming, such as neurodegenerative disease. “The beauty of our test is that it is actually robust across all our tumor types,” Chen said. “We consistently get down to this one- to three-part-per-million detection limit…across lung cancer, breast cancer, cervical cancer, colorectal cancer, you name it. It’s a generalizable approach.”
Hall believes these cancers are where high-accuracy MRD testing can have the fastest and most profound impact. “There are always interesting problems to solve,” Hall said, “but we’re still early in building the evidence and the traction to impact clinical care. It takes a lot of time and a lot of evidence to change the way doctors practice. We’re going to stay focused on oncology right now, especially breast cancer, lung cancer, therapy monitoring, and colorectal cancer. That’s really where the patients are and the real opportunity of MRD testing really is.”
Hall added, “There are many challenges if you get a patient’s disease status wrong. We’re trying to use blood tests to figure out whether the patient is low- or high-risk or where they are in the journey. We believe that these ultra-sensitive blood tests represent a significant advancement, and we remain committed to leading this market and integrating them into the standard of care.
NeXT Personal is currently a laboratory-developed test (LDT), and Personalis expects ongoing interaction with the FDA, especially as utility studies expand. Regarding insurance, Medicare has recently begun reimbursing the test for breast cancer, and Chen hopes that “lung and other cancers are not too far behind.”
The promise, both for patients and doctors, is straightforward but profound: a better understanding of the patient’s condition months, if not years, before conventional imaging would indicate anything. For patients living under the weight of the question, “Is it all gone?” clarity is what hands back the reins of their lives.
