Valerie Neff Newitt
July 2026—Being able to distinguish the subset of patients with stage IV colorectal cancer who have a chance at cure with ablation or resection from those who do not has been the aim of two University of Chicago physicians who developed a biomarker for colorectal liver metastases called PersonaCRC.
There are two elements to the test, says Sean Pitroda, MD, professor of radiation and cellular oncology at UChicago Medicine and developer with Ralph Weichselbaum, MD, Daniel K. Ludwig distinguished service professor, chair of radiation and cellular oncology, and co-director of the UC Ludwig Center for Cancer Research. Dr. Pitroda is chief medical officer of PersonaDx, the company behind PersonaCRC; Dr. Weichselbaum is chief scientific officer.
One element of the test is the molecular subtype of the colorectal liver metastasis: canonical, immune, and stromal. The second element is the integrated risk category that combines the molecular information and five clinical factors known as the clinical risk score, taken from the patient chart, to classify the patient’s metastases as low, intermediate, or high risk. “These have the strongest implications for survival, patterns of failure, and treatment planning,” Dr. Pitroda says.
Caris Life Sciences performs the RNA sequencing of the liver metastasis sample. “We use AI to analyze 150 individual genes,” Dr. Pitroda says, “and the analysis of those genes produces the molecular subtype.”
Among their publications is a 2018 article on the integrated molecular subtyping that defines a curable oligometastatic state in colorectal liver metastasis (Pitroda SP, et al. Nat Commun. 2018;doi:10.1038/s41467-018-04278-6). The oligometastasis hypothesis, they wrote, suggests a spectrum of metastatic virulence in which some metastases are limited in number and organ involvement and potentially curable with surgical resection or other loco-regional therapies. They and their coauthors provided what they described as a “framework for integrated classification and treatment of metastasis” and said their findings “support the biological basis of curable oligometastatic colorectal cancer.”
“Metastasis can be limited both in number and in pace,” says Dr. Weichselbaum of an understanding he notes surgeons had recognized long ago. Since general chemotherapy and targeted therapies are often not curative, he says, the thinking is that in select patients, localized interventions such as surgery, high-dose radiotherapy, or radiofrequency ablation could be used to cure metastasis.
There is a sizable subset of CRC patients who, if they get ablative treatment, he says, “would probably be cured or at least have a good chance to be cured.”
Dr. Pitroda says it was the first demonstration of a molecular basis for curability in stage IV CRC.
“If we can identify those patients,” he says, “we can give them curative treatments, not just palliative therapies. So what we’re trying to do at PersonaDx, and what we’ve done in colorectal cancer, is identify how you can accurately select those patients who are going to have curable or long-term survival versus those patients in whom their disease may manifest initially with limited presentation but will eventually become widespread and incurable with our modern approaches.”
This question differs from that of mutation-to-drug matching, Dr. Pitroda notes. Instead, it’s about matching curative versus palliative treatments to a patient’s biology, and he describes the framework they use as a “new way of thinking about how to classify metastases.”
The PersonaCRC test, in providing a biology-to-treatment match, he says, “offers that level of granularity that no other test currently offers.”
It was the phase three New EPOC randomized clinical trial data (Katipally RR, et al. JAMA Oncol. 2023;9[9]:1245–1254) that validated the molecular subtypes defined in the 2018 study published in Nature Communications.
The patients enrolled in the clinical trial had resectable or suboptimally resectable colorectal liver metastases. “The primary endpoint was to see if a drug called cetuximab improved outcomes beyond standard chemotherapy for these patients who were undergoing chemotherapy plus surgery of the liver metastases. It was a negative study,” Dr. Pitroda says. “But because the data were high quality and the samples were available, we were able to take the analysis we did from the 2018 paper, derive a biomarker only in that 2018 paper data set, then independently validate using the New EPOC samples. So the New EPOC [trial] served as a completely independent validation.” It was a European cohort, not American, he adds, so the chemotherapy regimens differed and the patients were slightly higher risk. “They tended to operate on higher-risk patients,” he says.
Of 240 patients who underwent hepatic resection for limited colorectal liver metastases, the biological phenotype of each molecular subtype in the validation cohort (147 patients) was concordant with those in the discovery cohort (93), the authors wrote. The low-risk group demonstrated an improved five-year overall survival of approximately 80 percent, superior to the high-risk group at approximately 40 percent.
The New EPOC study validated that “we can correctly identify these biological properties in their data set, but the survival differences we saw were comparable to what we had observed in our own data set,” Dr. Pitroda says. This means that “even though the molecular data was not trained to predict survival, the fact that it translated to another study and strongly predicted outcome independent of clinical factors meant there was truly an intrinsic property of the tumor that was extensible across cohorts and studies, supporting that this is a true underpinning for the molecular property that could then be used moving forward to support our PersonaCRC test.”
The sequencing data (available in about 10 days) is transmitted from Caris to Protean BioDiagnostics of Orlando, Fla. “Once Protean has the data, they make it accessible to PersonaDx to run the PersonaCRC classifier,” Dr. Pitroda says. PersonaDx (whose CEO is Kevin Donnelly, also a PersonaDx co-founder) performs the analysis and integrates the AI, molecular data, and clinical input and generates the final report within 48 hours, which is then made available to the oncologist. The risk groups, molecular/clinical profile, and preferred treatment strategies are explained as follows:
- High risk: Unfavorable subtype and/or high clinical risk. Primarily systemic therapy; local therapy selective or palliative.
- Intermediate risk: Mixed subtype and/or moderate clinical risk. Combination of systemic and local therapies.
- Low risk: Favorable subtype and low clinical risk. Curative local therapy (resection/ablation), less systemic therapy.
Dr. Pitroda foresees the test having applications that go beyond which patients should get a local therapy for liver metastasis. But for now, the data support ordering the test in patients with newly diagnosed liver metastases who are being evaluated for local therapy or patients who have already undergone a liver resection and later present with recurrent disease for which clinicians must decide whether additional local therapy or a systemic approach is the best path. Another appropriate use, Dr. Pitroda says, would be for a patient with a metastasis that is “borderline resectable, with a risk-benefit calculation that’s a little uncertain, or where standard studies give mixed or ambiguous signals about the potential prognosis.”
The test is billed using CPT code 81479 (unlisted molecular pathology procedure) and offered as a CLIA-validated laboratory-developed procedure and is eligible for Medicare reimbursement pathways applicable to LDTs. “We’re working on getting insurance approval,” Dr. Weichselbaum says. Part of the process is showing strong clinical uptake and utility, “so we’re generating the data as the test is rolled out,” Dr. Pitroda says. The cost is $2,265. “If a patient couldn’t afford it,” Dr. Weichselbaum says, “we would find a way to make it available.”
“The theoretical economic case is pretty compelling,” Dr. Pitroda says, citing the cost of chemotherapy per year as $300,000 and of a liver transplant (“the new fad in treating liver mets,” he says) as upward of $700,000 or $800,000.
Dr. Weichselbaum says feedback from medical oncologists and surgeons at National Cancer Institute-designated cancer centers (about half of such centers in the U.S., he says) has been “very strong, even though the patterns of practice differ a little bit.”
Rollout of the test at the University of Chicago was slated for early June. “PersonaCRC offers a new level of insight into disease biology that can help us better understand prognosis and tailor treatment plans,” says Ardaman Shergill, MD, assistant professor of medicine, oncology, at the University of Chicago Medical Center.
Dr. Weichselbaum calls it “the single most determinative test for cure, noncure that’s available medically.
“It defines risk prognostic groups with great differences in survival.”
“What we’re doing,” Dr. Pitroda says, “is adding a molecular dimension to the characterization that the pathologist is providing at the histologic level. It’s building on it but offering a level of detail that unfortunately is invisible to the naked eye or under the microscope.”
“The pathologist is the key nexus for this test to run successfully,” he adds, noting the need for adequate specimens to enable the required RNA expression detail.
“Actually,” Dr. Weichselbaum says, “with all of these tests being developed, the pathologist has become the single most important member of the health care team.”
Valerie Neff Newitt is a writer in Audubon, Pa.