NEW YORK – Circulating tumor DNA analysis at the end of first-line treatment in patients with aggressive B-cell lymphoma had greater predictive value than a single PET-CT scan, and its use may avert unnecessary tests and therapies, researchers reported at the American Society of Hematology's annual meeting Saturday.
In lymphoma, the standard method for evaluating treatment response is a PET-CT scan conducted upon completion of therapy. However, the positive predictive value of that method in determining whether the cancer has resolved following treatment is only about 50 percent, and doctors typically compensate for its shortcomings by subjecting patients to additional tests or scans to decide whether to give salvage therapy in those who haven't adequately responded.
In a presentation at the meeting, Mark Roschewski, clinical director of the lymphoid malignancies branch of the US National Cancer Institute, noted that a PET-CT scan can't measure disease at the molecular level and lacks specificity for lymphoma. As an example, he said that in more than 1,000 patients with diffuse large B-cell lymphoma in the GOYA clinical trial, 88 percent of patients were deemed to be in complete remission by PET scan, but 20 percent of those patients experienced subsequent disease progression.
This uncertainty is acknowledged in practice guidelines. In recommendations published in 2014 for initially evaluating, staging, and assessing the treatment response of lymphoma patients and in the National Comprehensive Cancer Network's guidelines, experts said that patients should be biopsied again or given an additional scan to determine next therapeutic steps if they had a positive PET-CT scan at the end of their first-line treatment.
Although physicians are very familiar with the need for additional tests to adjudicate remission by PET-CT scan, Roschewski said that "what's unclear and is an open question is how often patients are actually referred for second-line therapy on the basis of a single positive PET scan, and the results are not adjudicated."
Roschewski and his colleagues at the NCI speculated that a determination of minimal residual disease (MRD) via phased variant enrichment and detection sequencing (PhasED-seq) could provide a more reliable and complementary method for identifying the lymphoma patients that don't need further salvage therapy. PhasED-seq can sequence multiple mutations on a single cell-free DNA strand and has been shown to be able to detect one in 10 million cfDNA molecules with a low background error rate.
In a study published in Hematological Oncology in June, Roschewski and his collaborators conducted a pooled analysis of six prospective large B-cell lymphoma trials and showed that PhasED-seq was prognostic after just two cycles of treatment. At the meeting, Roschewski shared that in 75 patients in this study who were believed to be in complete remission at the end of treatment with various chemotherapies, targeted drugs, and antibodies according to PET/CT scans and additional testing per standard practice guidelines, a ctDNA analysis showed that 13 had detectable ctDNA and 62 did not.
Roschewski's group hypothesized that most patients that have a positive PET-CT scan as defined by a Deauville score of 4 or 5 actually do not have lymphoma and will not progress, and that a single PhasED-seq assay would be more predictive than a single PET-CT scan to predict clinical outcomes. On the Deauville scale, a 1 or a 2 is a negative result and a 4 or 5 is positive for the presence of disease.
His team is exploring these questions within an ongoing clinical trial sponsored by the NCI in which researchers are testing AstraZeneca's BTK inhibitor Calquence (acalabrutinib) with dose-adjusted commonly prescribed chemotherapy regimens in patients with untreated diffuse large B-cell lymphoma. They analyzed data from patients in the trial who had completed first-line therapy, given a plasma sample, and undergone an end-of-treatment PET-CT scan. Two nuclear medicine physicians at the NCI who were blinded to patients' outcomes in the trial analyzed the scans and the researchers compared ctDNA analysis by PhasED-seq to those results.
Fifty-five patients completed therapy and had tests available. Roschewski said the patient population was well balanced in terms of high-risk features, and 18 percent of patients were over 70 years old. The two-year progression-free survival rate in this population was 84.7 percent after a median follow-up of two months. For patients that had undetectable ctDNA by PhasED-seq after two cycles, the two-year progression-free survival rate was 96.9 percent, as compared to 64.3 percent for those with detectable ctDNA.
At the end of therapy, 40 patients, or 78 percent, achieved a complete remission according to a PET-CT scan, and those patients had a two-year progression-free survival rate of 90 percent, compared to 67 percent for those whose scans indicated presence of disease. However, 78 percent of patients had undetectable ctDNA at the end of therapy, and the two-year progression-free survival rate for those patients was 98 percent.
Among the 14 patients who had a positive PET-CT scan on blinded review, nine underwent additional procedures including five follow-up PET-CT scans and four tissue biopsies. "Only one of those additional procedures detected lymphoma," Roschewski said. "Seven of 14 patients tested negative for MRD that were [PET-CT] positive."
In contrast, seven patients that were PET-CT positive also tested positive for MRD, and two of those patients' disease progressed. One died of a second malignancy and four are still being followed in the trial. Thirty-five patients, or 88 percent, were negative by PET-CT scan and ctDNA testing, and none died or progressed. There were four patients who had negative PET-CT scans but were MRD positive, and all four progressed.
Roschewski noted that they have only about 10 months of data on the five patients who tested PET-CT negative and were MRD positive and have not yet progressed. "It's possible they will progress, but there are cases where the MRD has cleared after therapy, so it's not true that every patient that's [MRD] positive will indeed progress," Roschewski said, adding that he believes the results for MRD negativity are more immediately actionable.
"Only two patients with a positive end-of-therapy PET-CT scan determined blindly by nuclear medicine [physicians] actually had progression," Roschewski said. "Most of these additional procedures that we've gotten used to doing and ordering to adjudicate these [PET-CT] scans do not detect lymphoma."
Based on the data he presented at the meeting, Roschewski concluded that ctDNA detection by PhasED-seq is prognostic after two cycles of first-line therapy and at the end of therapy, that MRD negativity at the end of therapy predicts a very low likelihood of disease progression, and that ctDNA appears to be of greater predictive value than PET-CT scans.
"One of the most important clinical pearls is that salvage therapy should not be delivered on the basis of a singular end-of-therapy PET-CT scan due to the limited sensitivity and specificity [of the test]," Roschewski said.