Tumor DNA in Blood Signals Immunotherapy Response
News Sep 02, 2016
A pilot study by NCI researchers suggests that tumor DNA circulating in the blood of patients with cancer might be a biological marker for determining, soon after the treatment has started, whether a form of immunotherapy is likely to work for a given patient.
The results further suggest that doctors could use information about circulating tumor DNA to quickly identify patients who are responding (or are not responding) to the treatment, the researchers reported in the August 1 Clinical Cancer Research. This information, in turn, could allow doctors to modify a treatment plan accordingly.
The findings are from a retrospective analysis of blood samples from patients who received a treatment known as T-cell transfer immunotherapy, also called adoptive cellular therapy. Using the patients’ samples and information about treatment outcomes, the researchers identified patterns of changes in the blood levels of tumor DNA that were associated with complete responses, partial responses, or no responses.
“By looking at the circulating tumor DNA in blood, we were able to determine whether a treatment was working within the first 2 weeks,” said study coauthor Steven Rosenberg, M.D., Ph.D., chief of the Surgery Branch in NCI’s Center for Cancer Research (CCR).
When cancer cells are eliminated from the body, DNA released from these cells may enter the bloodstream and circulate, Dr. Rosenberg explained. Rising blood levels of DNA from cancer cells that are targeted by immunotherapy may indicate that the treatment is hitting its targets.
“This approach gives you a very rapid indication of whether a particular therapy is causing the death of tumor cells that bear the gene you are targeting,” noted Dr. Rosenberg. The results are preliminary, he added, and need to be confirmed by larger studies.
A Need for Biomarkers
In the study, the researchers analyzed blood samples from 39 patients with advanced melanoma who had received T-cell transfer immunotherapy during clinical trials at NCI in the 2000s.
This particular treatment involved several steps. First, immune cells that had infiltrated a patient’s tumors were collected. These tumor-infiltrating lymphocytes were then activated in the lab and grown in large batches. Finally, these expanded killer T cells were infused into the patient.
About 20 percent of patients with metastatic melanoma who receive this treatment have a complete response, but there has not been a way to identify patients who will respond to this therapy.
“It would be helpful if we could use circulating tumor DNA to determine earlier whether a patient is responding to treatment,” said coauthor Richard M. Sherry, M.D., also of CCR’s Surgery Branch. “If a treatment is not working, you want to change directions to help the patient.”
Patterns in Circulating Tumor DNA
For the study, the researchers used a mutated form of the BRAF gene as a circulating marker. Mutations in this gene drive half of all malignant melanomas and were present in all 39 patients.
Dr. Rosenberg’s team had collected and stored blood samples from these patients over the course of treatment, including before and soon after therapy began. Of the 39 patients, 10 had complete responses, 14 had partial responses, and 15 did not respond.
“We looked at early time points after the initiation of treatment and saw patterns in the levels of circulating tumor DNA in patients who responded and patients who did not respond,” said coauthor Mark Raffeld, M.D., of CCR’s Laboratory of Pathology.
For example, in 9 of the 10 patients who had complete responses, blood levels of the mutated BRAF DNA increased sharply during the first 2 weeks after treatment and then fell to zero.
“Most of the tumor cell killing—as judged by the peak of circulating tumor DNA—occurred within 5 to 9 days of starting treatment, and then the patient’s blood levels of circulating tumor DNA would drop to zero in patients who developed a complete response,” said Dr. Raffeld, explaining the pattern.
By contrast, with one exception, patients whose blood showed no early spike in circulating tumor DNA failed to achieve an objective response, the researchers noted. “The results of this pilot study suggest that patients without these ‘peaks’ are unlikely to respond and might need a different therapy,” said Dr. Raffeld.
The Serial Testing of Blood Samples
Clinical indications of a response to T-cell transfer immunotherapy can often be seen within 2 months of a patient starting treatment. “With this approach, we are trying to analyze tumor responses to treatment at the level of DNA,” said Dr. Sherry.
Circulating markers are already being used to monitor patients with various types of cancer for recurrences. The new findings suggest that the approach could be explored for this type of immunotherapy as well, the researchers noted. For instance, they reported that, for a patient who had a recurrence, a blood sample collected before the recurrence had been identified—with standard clinical tests—had elevated levels of mutated BRAF DNA.
“This is a very exciting study,” commented Anirban Maitra, M.B.B.S., a professor of pathology at the University of Texas MD Anderson Cancer Center, who investigates the use of what are often called “liquid biopsies” for patients with cancer. Liquid biopsies are blood tests that provide genetic and molecular information about cancer cells that are more typically obtained by analyzing a piece of tumor that has been removed surgically.
The longitudinal, or serial, testing of blood samples from patients receiving targeted or immune-based therapies is critical for developing new ways to assess the effectiveness of treatments or the emergence of resistance, said Dr. Maitra, who was not involved in the NCI study. “This research provides important supportive evidence for that approach.”
The spike in the mutated BRAF DNA that correlated with treatment response would not have been apparent with one-time sampling, he noted. Furthermore, the short interval between each liquid biopsy “would be inconceivable for tissue-based serial sampling.”
Measuring tumor responses to immunotherapy can be challenging, Dr. Maitra noted, citing the example of the “pseudo-progression” of tumors. This occurs when tumors that are being infiltrated by immune cells in patients who are receiving immunotherapy appear to be progressing when evaluated using standard imaging tools.
“Liquid biopsy-based approaches can potentially circumvent this artifact and provide a more objective measure of treatment response,” he said.
Circulating tumor DNA represents an additional source of information about tumors, noted Dr. Raffeld. “This approach gives you a window into how much tumor is present in a patient, and that information could supplement information from radiology techniques,” he said.
At NCI, Dr. Rosenberg has begun to test the use of circulating tumor DNA in patients receiving immunotherapy for colon cancer. “This is a very important investigational tool, and we have just begun to use it to test the effectiveness of immunotherapies in patients,” he said.