A San Diego blood test caught 95% of early pancreatic cancers

Pancreatic cancer is among the deadliest malignancies precisely because it is so hard to catch early, when treatment still has a real chance of working. Most cases are found only after the disease has spread, which is why researchers have spent years chasing a simple blood test that could flag a tumor at Stage 1. A screening platform developed at the University of California, San Diego has become a widely cited example of that ambition, after a pilot study reported that it identified roughly 95 percent of the earliest-stage pancreatic cancers it was tested against.

The appeal of a blood-based screen is easy to understand. A single tube of blood is cheap to draw, carries little risk, and could in principle be added to the routine tests people already receive during a checkup. That is why the prospect of catching a cancer as lethal as this one from an ordinary blood draw has drawn sustained attention from researchers and clinicians, and why a reported detection rate near 95 percent for the earliest stage stood out when it emerged.

How the UC San Diego blood test works

The test is built on a technology called high-conductance dielectrophoresis, which uses an electrical field to isolate tiny particles called extracellular vesicles from a small blood sample. These vesicles are shed by cells throughout the body and carry proteins that can hint at what their parent cells are doing, including whether a tumor is present. According to UC San Diego Health, an artificial-intelligence model then analyzes the pattern of tumor-associated proteins captured on those vesicles to estimate the likelihood that cancer is present.

The platform grew out of more than a decade of work at the Moores Cancer Center and was commercialized by Biological Dynamics, a San Diego company that spun out of the university. Several of the study’s authors were affiliated with the company, a disclosure the researchers included in their published findings.

What the 95 percent figure actually measured

The headline number comes from a pilot study that compared blood samples from 139 patients with Stage 1 and Stage 2 pancreatic cancer against 184 people without the disease. In that group, the platform flagged about 95.5 percent of Stage 1 cases, the point at which surgery and treatment are most likely to succeed. The University of California described the result as an early but striking signal that a routine blood draw might one day surface tumors that current imaging and symptom-based diagnosis routinely miss.

Because the study was small and drawn from a defined set of patients and controls, its authors framed the results as preliminary. A screening test’s real-world value depends on how it performs across large, diverse populations, including how often it produces false positives that send healthy people into unnecessary follow-up testing. A high detection rate in a curated study group does not automatically carry over to the general public, where cancer is far rarer and the balance between true and false alarms shifts.

The distinction between sensitivity and specificity is central to that judgment. Sensitivity describes how many genuine cancers a test catches, while specificity describes how well it avoids flagging healthy people. A test tuned to catch nearly every early tumor can still stumble if it also labels large numbers of cancer-free patients as suspicious, which is why researchers report both figures and why regulators weigh them together before endorsing a screening tool.

Why early pancreatic detection matters so much

Pancreatic ductal adenocarcinoma is notorious for producing few symptoms until it is advanced. By the time jaundice, weight loss, or abdominal pain drives a patient to a doctor, the cancer has often already spread to nearby organs or lymph nodes, sharply narrowing treatment options. Survival statistics reflect that pattern: patients diagnosed at the localized stage fare dramatically better than those found later. A blood test that reliably caught the disease at Stage 1 could shift a meaningful share of cases into that more treatable window.

The organ’s location compounds the problem. The pancreas sits deep in the abdomen behind the stomach, which means small tumors rarely cause noticeable trouble and are hard to see on routine exams. There is also no broadly recommended screening program for the general population, unlike the established tests for breast, colon, and cervical cancers, leaving most patients to be diagnosed only after symptoms appear. A reliable blood-based screen could begin to fill that gap, particularly for people at elevated risk because of family history, certain genetic mutations, or conditions such as new-onset diabetes in older adults.

Related research on blood-based detection has pursued the same goal through different chemistry. A separate line of work described in the peer-reviewed literature has focused on classifiers that read molecular signatures in circulating material to distinguish early pancreatic cancer from benign conditions, underscoring how competitive and active this field has become.

What still has to happen before a test reaches patients

A promising pilot is a starting point, not an approval. To move from a research result to a clinical tool, a screening test generally must be validated in larger prospective studies, clear regulatory review, and prove that it changes outcomes rather than simply detecting more disease. Researchers also have to define who should be screened, since testing broad, low-risk populations can generate more false alarms than genuine early catches.

For now, the San Diego platform stands as one of the more encouraging demonstrations that extracellular vesicles can carry an early cancer signal detectable in a single tube of blood. Whether that translates into a widely used screening test will depend on the slower, less headline-friendly work of large-scale validation, a process the university’s researchers have said must come next.

This article was produced with the assistance of AI and reviewed by Morning Overview editors prior to publication.

More from Morning Overview