Investigators Validate Role of Liquid Biopsy in Predicting Chromosomal Abnormalities in Myeloid Neoplasms

Liquid biopsy using targeted next-generation sequencing (NGS) for early diagnosis and monitoring of patients with myeloid neoplasms is effective and detects chromosomal structural abnormalities, according to a retrospective study conducted at Hackensack University Medical Center.¹

Patients with myeloid neoplasms often have invasive testing done such as bone marrow biopsies, to obtain samples of viable cells for cytogenetic testing with a turnaround time of 7 to 14 days. Alternative options have been explored; however, methods such as fluorescence in situ hybridization which can only detect one abnormality at a time, array technology which requires significant quantity of samples to confirm accuracy, and whole-genome sequencing which is expensive have failed to overcome their associated pitfalls in this space.

To overcome these limitations, investigators leveraged NGS to evaluate chromosomal gain or loss in liquid biopsies of patients with myeloid neoplasms.

“Our primary goal was risk stratifying patients and to understand if we could use these peripheral blood tests in place of more invasive testing; we found that…patients with high sensitivity and specificity were able to be detected if they had myeloid neoplasms, some even had lymphoid neoplasms,” lead author Andrew Ip, MD, explained.

In an interview with OncologyLive^®, Ip discussed the importance of liquid biopsies in detecting myeloid neoplasms and risk stratifying patients. Ip is chief of the Outcomes and Value Care Division and a member of the Division of Lymphoma at John Theurer Cancer Center, Hackensack University Medical Center, in New Jersey.

Ip described targeted panel used in the study as having, “close to 300 different types of DNA, mutation profiling, and up to approximately 1200 RNA expression detection, so it is much more comprehensive than some of the other tests that are usually done.”

The retrospective study included samples from March 2020 to September 2021 at the John Theurer Cancer Center and Hackensack University Medical Center in a sample size of 144 patients with a median age of 68.5 years (range, 24-96). There were 2821 plasma cell-free DNA (cfDNA) samples from patients with confirmed or suspected myeloid or lymphoid neoplasm, which were sequenced for chromosomal gain or loss. The diagnoses at baseline included acute myeloid leukemia (AML; 22%), myelodysplastic syndrome (MDS; 34%), and myeloproliferative neoplasms (44%).

Describing the study Ip said, “the idea here is to use liquid biopsies, essentially peripheral blood testing [and] for this study we looked at copy number variation [to identify] cytogenetic differences.”

Of the 2821 examined samples, 54.5% had mutations consistent with the presence of neoplastic clones. Among these (n = 1539), 59% of samples had abnormalities were associated with myeloid neoplasms, 16% of which had chromosomal structural abnormalities, and 41% of samples were associated with lymphoid neoplasms, 12% of which had chromosomal abnormalities. The abnormalities were shown in mutations of at least 1 gene in all samples except 2 which had no demonstrable mutations.

Further, the median variant allele frequency (VAF) was 8.09% (range, 0.002%-99.55%). Investigators noted that when VAF was detected at 13% or higher samples were “associated with clear, detectable chromosomal structural abnormalities.”

Ip described the results having, “a pretty high sensitivity,” adding that “within the the allele frequency, there is the VAF, this is detecting how much of the mutation is within the blood sample. For our study, 13% was the cut off to detect copy number variations or chromosome structure abnormalities that would define AML or MDS.” Ip explained that other NGS liquid biopsy tests could get slightly different results, and that comparing different tests or using many tests in conjunction with each other is a next step to validate the research.

For example, in this study, findings from 89 liquid biopsy samples from patients with MDS or AML were compared with available bone marrow samples. The cytogenetic data were obtained within 2 weeks of the liquid biopsy sample. To conduct the comparison, investigators grouped the cytogenetic and chromosomal findings into 3 risk groups: poor, intermediate, and complex. In terms of classification the NGS chromosomal structural analysis and the bone marrow samples had 100% concordance.

“That means we were able to classify these patients 100% of the time based off their leukemia risk, as defined by the European Leukemia Network and other organizations,” Ip said. “That showed the power of this test. I think that is important for clinical implications.”

Other noted advantages of the NGS test included the turnaround time of 5 to 7 days, ability to allow for less invasive and serial monitoring of patients, as well as the ability to circumvent hurdles with other tests including lack of circulating tumor cells and availability of bone marrow samples.

Gene selection for the NGS panel was also a noted advantage by the investgators who wrote, “genes were selected because they are involved in oncogenesis rather than because they are relevant to covering chromosomal regions. Therefore, these genes provided an important mutation profile in addition to detecting chromosomal abnormalities. Smaller panels may not cover all chromosomal regions that are important in hematologic neoplasms. The ability of such a panel to detect both mutations and chromosomal abnormalities is a highly cost effective and clinically useful approach in evaluating molecular abnormalities in hematologic neoplasms.”

Ip discussed future steps of a possible pairing with Hackensack University Medical Center’s partner, Georgetown University in Washington, DC, to conduct more research with a larger sample size.

Limitations of the study included the isolated chromosome abnormality, deletion 5q, which is a form of MDS that is not always detected by testing. Ip explained, “deletion 5q is a rare case and [approximately] 95% to 99% of these patients will have some sort of molecular mutation profile that you can detect here.” Another drawback of liquid biopsies is that most insurance companies do not pay for the test. However, Ip explained that is quickly changing “Every month, at least in the state of New Jersey, there seems to be more understanding that these tests are important for our patients.”

In discussing disadvantages of liquid biopsies, Ip said that the abundant amount of data can be overwhelming for clinicians to sort through. “I think this paper hopefully helps to [clarify] that patients could be risk stratified with this sort of test… The top [takeaway] is how do you interpret the wealth of information, because if you have 275 potential genes that are mutated, and 20 of them come back as positive, what does that mean?” There are also many different liquid biopsy tests available and “there’s no set standard for some of this and that’s frustrating for the clinician, and even for the research side, [as investigators] try to come together and ask: Is there 1 test that we can use and study?”

Reference

Ip A, Della Pia A, Kim GYG, et al. Reliability of cell-free DNA and targeted NGS in predicting chromosomal abnormalities of patients with myeloid neoplasms. Front Oncol. Published online June 14, 2022. doi:10.3389/fonc.2022.923809