Photo by Larry Young
Surveillance of circulating tumor DNA (ctDNA) can help predict relapse in most patients with diffuse large B-cell lymphoma before there is clinical evidence of the disease, according to a study published in The Lancet Oncology.
Investigators analyzed ctDNA using the clonoSEQ minimal residual disease (MRD) test and found they could predict relapse with a positive predictive value (PPV) of 88% and a negative predictive value (NPV) of 98%.
The test detected relapse a median of 3.5 months quicker than computed tomography (CT) scans.
“Patients with DLBCL with low amounts of disease at relapse have better survival than those with more disease, which is the rationale for surveillance CT scans,” said study author Wyndham Wilson, MD, PhD, of the National Cancer Institute in Bethesda, Maryland.
“Because the ctDNA test detects disease at a molecular level, it detects microscopic disease, which cannot be detected by CT scans, and may improve patient survival. Furthermore, ctDNA is non-invasive and can be employed as frequently needed, unlike surveillance CT scans, which expose patients to radiation and intravenous contrast.”
For this study, Dr Wilson and his colleagues evaluated 126 DLBCL patients who had participated in clinical trials from May 1993 to June 2013 and were followed for a median of 11 years post-treatment.
Surveillance monitoring
To investigate whether ctDNA monitoring could overcome the limitations of standard imaging techniques, the researchers compared serial ctDNA samples to CT scans taken at the same time post-treatment in patients who had achieved complete remission. This was known as “surveillance monitoring.”
The investigators performed surveillance monitoring of ctDNA in 107 patients who achieved complete remission.
The hazard ratio for clinical disease progression was 228 for patients who had detectable ctDNA during surveillance, when compared to patients with undetectable ctDNA (P<0.0001).
Surveillance ctDNA had a PPV of 88.2% and an NPV of 97.8%. And it revealed the risk of recurrence at a median of 3.5 months (range, 0-200 months) before there was evidence of clinical disease.
Interim monitoring
The researchers also analyzed whether the presence of ctDNA at the beginning of the third cycle of treatment predicted relapse, regardless of whether patients achieved complete remission by the end of treatment. This was known as “interim monitoring.”
Of the 108 patients included in the interim monitoring analysis, ctDNA was detected in 24 patients, 15 of whom eventually relapsed. Only 17 of the 84 patients with undetectable interim ctDNA relapsed.
Five years after the interim serum samples were taken, 80.2% of the patients who were negative for ctDNA were relapse-free, as were 41.7% of patients who were positive for ctDNA (P<0.0001).
Detectable interim ctDNA had a PPV of 62.5% and an NPV of 79.8%.
Fourteen of the 15 patients with detectable ctDNA who relapsed did so within 6 months of the end of treatment, as did 7 of the 17 patients without interim ctDNA.
Based on these results, the investigators concluded that surveillance monitoring of ctDNA identifies DLBCL patients at risk of disease recurrence before clinical evidence of disease in most patients, and interim monitoring of ctDNA is a promising biomarker to identify patients at high risk of treatment failure.
This research was funded by Adaptive Biotechnologies, the company developing the clonoSEQ MRD test, as well as the National Cancer Institute.
