According to results presented at the 2021 ASH Annual Meeting and Exposition, new sequencing technologies may overcome previous limitations of circulating tumor DNA (ctDNA) profiling in patients with central nervous system (CNS) lymphoma. Using ultrasensitive sequencing, researchers were able to diagnose the disease based on ctDNA mutational profiles. They also found that ctDNA status before and during treatment could predict progression-free survival (PFS) and overall survival (OS).
“ctDNA has been established as an important noninvasive biomarker in various malignancies. However, the role of ctDNA in CNS lymphoma has been very limited so far,” said Florian Scherer, MD, Principal Investigator, Department of Hematology, Oncology, and Stem Cell Transplantation, University Medical Center Freiberg, Germany.
Dr Scherer explained that the data are paradigm shifting for the diagnosis and prognostication of primary CNS lymphoma via cell-free DNA from blood and cerebrospinal fluid (CSF), and that liquid biopsy diagnosis has 100% positive predictive value.
Limitations of ctDNA include its very low concentrations in plasma, leading to low detection rates by sequencing technologies and the limited applicability of single-gene assays. To overcome these limitations, Dr Scherer and colleagues applied Cancer Personalized Profiling by Deep Sequencing and Phased Variant Enrichment and Detection Sequencing to tumor biopsies (N = 85), plasma samples (N = 131), and CSF specimens (N = 62) from 92 patients with CNS lymphoma and 44 patients with other brain cancers or inflammatory cerebral diseases. A total of 794 distinct genetic regions were targeted.
Concentrations of ctDNA were then correlated with radiologic tumor burden measures and were tested for associations with clinical outcomes at distinct clinical time points.
Cancer Personalized Profiling by Deep Sequencing detected genetic aberrations in 100% of CNS lymphoma tumor samples, with a median of 288 mutations per patient. The most frequently mutated genes were MYD88, PIM1, and CD79B. Phased Variant Enrichment and Detection Sequencing detected ctDNA in 78% of pretreatment plasma samples and 100% of pretreatment CSF samples, with specificities of 96% and 97%, respectively. ctDNA levels were 80-fold higher in CSF than in plasma: concentrations of ctDNA ranged from 0.0004% to 5.94% allele frequency in plasma and 0.0049% to 50.47% allele frequency in CSF.
Median plasma ctDNA levels in CNS lymphoma were more than 200-fold lower compared with ctDNA concentrations in patients with systemic diffuse large B-cell lymphoma from historical data.
A significant correlation was found between ctDNA concentrations and total radiographic tumor volumes (TRTV) measured by magnetic resonance imaging but not with clinical risk scores or concurrent steroid treatment. Assessment of ctDNA at pretreatment time points predicted PFS and OS both as continuous and binary variables: patients with positive ctDNA before treatment had significantly worse PFS (P <.0001) and OS (P = .0001).
Patients could be stratified into favorable or poor risk groups by combining ctDNA and TRTV as pretreatment biomarkers. Patients with detectable pretreatment ctDNA and a high TRTV had a 2-year OS rate of 14%, whereas those with no detectable pretreatment ctDNA and a low TRTV had a 2-year OS rate of 100%. Furthermore, ctDNA positivity during curative-intent induction therapy was significantly associated with both PFS (P = .0002) and OS (P = .004).
Using a supervised machine learning approach from tumor whole-genome sequencing data and genotyping analyses, the investigators developed a novel classifier to distinguish CNS lymphoma from other CNS tumors noninvasively. The classifier was also based on mutational landscapes in both plasma and CSF. An independent validation on 207 specimens from patients with and without CNS lymphoma was subsequently performed.
They found 100% specificity and 100% positive predictive value for noninvasive diagnosis of CNS lymphoma, with a sensitivity of 57% for CSF and 21% for plasma, “suggesting that a significant subset of CNS lymphoma patients might be able to forego invasive surgical biopsies,” said Dr Scherer.
“We believe that, in the future, ctDNA could help stratify patients into risk groups that might benefit from treatment reduction or treatment intensification. Furthermore, we envision a potential role of ctDNA for surgery-free CNS lymphoma detection or even the identification of occult CNS involvement,” he concluded.