Article
Author(s):
Andrew Ip, MD, delivers a keynote on the role for liquid biopsy and NGS in hematologic malignancies, including the barriers to implementing seamlessly into clinical practice.
The predictive value of next-generation sequencing (NGS) and liquid biopsies hold the potential to revolutionize practice across the continuum of care. However, oftentimes the seemingly limitless information provided with genomic analysis, can be overwhelming. The goal of harnessing the power of these approaches in hematologic malignancies, whether through the identification of predictive response signal and surrogate prognostic biomarkers, or detecting cancerous cells on a deeper molecular level, remains at the forefront of research in hematology.
During the 27th Annual International Congress on Hematologic Malignancies®: Focus on Leukemias, Lymphomas, and Myeloma, Andrew Ip, MD, chief of the Outcome and Value Care Division and member of the Division of Lymphoma at John Theurer Cancer Center at Hackensack University Medical Center, in New Jersey, delivered a keynote on the role for liquid biopsy and NGS in hematologic malignancies including the barriers to implementing seamlessly into clinical practice.
“The whole purpose of using NGS is to be able to detect our patients’ disease burden at a very high sensitivity and accuracy, hopefully at the molecular level,” Ip said. “As we learn more about the tumor microenvironment, and the machinery and the interplay of different protein expression, epigenetics, methylation, and all the things that are going on in our individual cells that we can apply it to our actual patients, and the different interplay between different diseases.”
Adapting the use of liquid biopsies for hematologic malignancies has demonstrated the most applicability in the molecular profiling of hematopoietic neoplasms. In a retrospective analysis of 65,700 liquid biopsy samples collected for NGS, there was a 73.8% positive percent agreement with variants detected using buffy coat, bone marrow aspirate, or tissue sequencing.1
The highest median somatic allele frequencies were in myeloid neoplasms at 20% to 45% compared with 2% to 10% in lymphoid or plasma cell neoplasms, which Ip noted was expected as “myeloid neoplasms are mainly living in the peripheral blood.” Further, Ip highlighted that the data showed that over 50% of the liquid biopsies harbored genomic mutations not detected in tissue samples. “[This] highlights that we need better testing overall and [understand] the role of these liquid biopsies [so they] can detect more information,” he said.
As a prognostic tool, Ip turned his attention to the detection of clonal hematopoiesis, the subtypes of which are categorized by somatic mutations, and include clonal hematopoiesis of indeterminate potential (CHIP). In an analysis presented at the 64th American Society of Hematology (ASH) Annual Meeting and Exposition investigators categorized approximately 11,000 for whom whole-exome sequencing data showed CHIP (92.4%) or unexplained cytopenias (7.6%) into risk groups.3 The investigators leveraged genetic mutations, blood cell parameters, and clinical outcomes in conditional probability-based recursive partitioning analysis to create the criteria for risk of myeloid neoplasms. “What they found is that you could have a low risk, intermediate risk, and high risk. And you can see in the curves, it's basically a tenfold difference with each risk level—for low risk, the risk was a little bit under 0.3%, and then intermediate risk was approximately 2% risk. [For] high risk, the risk of going as high as 25%. The high-risk group only had approximately 100 patients, so it’s a smaller sample size but if you have a test that can predict [this] in an [individual] without a cancer that 25% of them in the next 5 to 10 years will develop a myeloid neoplasm, that’s very powerful.”
Gaining a clearer picture of minimal residual disease (MRD) with liquid biopsy and NGS is also a rapidly expanding area for research. “Why do we use liquid biopsies? The bottom line is we’re trying to see below the surface more, we’re trying to detect their burden of disease at the molecular level,” Ip said. “Traditionally, when we started out with certain kinds of circulating tumor DNA, we were looking at maybe numbers of 1 in 10,000, detecting cancer cells and cell-free DNA. That could maybe tell us that the patient is in a complete response or remission [CR], and going down further to stringent CR and potentially molecular or flow CR. Now, we’re talking about looking at parts per million, or 10–6 or even greater.”
In an abstract presented at ASH, finding that disease burden in patients with acute lymphoblastic leukemia (ALL) following allogeneic hematopoietic cell transplant showed that clonoSeq, an ultra-sensitive NGS sequencing assay, had the potential to predict disease recurrence at 2 years.4 MRD was evaluated and categorized as undetectable (0), low (< 10–4), high (≥ 10–4to ≤ 10–3) and very high (> 10–3). Among 122 patients, those who had very high MRD pretransplant, over half relapsed at 2 years at compared with approximately 10% of patients with low MRD pretransplant.4
As a surrogate end point, MRD has been tracked for patients following chimeric antigen receptor (CAR) T-cell therapy and postinduction treatment strategies to signal increased risk of relapse. For example, in follicular lymphoma, patients treated in the FIL FOLL12 study who had MRD positivity at any time during the follow-up had an increased risk of relapse within 6 months vs those who were MRD negative (HR, 2.82; 95% CI, 1.84-4.32; P <. 001).5 Patients in this study were randomly assigned to rituximab maintenance or PET/MRD response-based postinduction treatment in the second-line setting. “It proves the point that MRD monitoring for certain diseases, especially in follicular lymphoma, may be very helpful in determining patient outcomes.”
“Standardization is a very big thing that we’re missing as we move forward with a lot of the different technologies we’re studying and, and the many different companies that are out there,” Ip said. Efforts are underway to compare sensitivity rates of the panels, but Ip cautioned that no methodology is perfect and false positive rates are an inherent risk with each test.
In terms of applicability, Ip noted during a discussion that the difference in panels available will also play a role in standardization. “There is a difference between platforms, cloneSeq may focus on different specific B-cell mutations, genomic expression, and it may just be very targeted to a few different genes, vs some of the other broader NGS panels that give us kind of a larger breadth of genomic information and molecular understanding,” Ip said. “I think cloneSeq does have a role, and it has an approval to monitor for certain mutations for different key malignancies. But I think a lot of what [we’re discussing] is the value of adding more broader, but yet targeted sequencing platforms and how that can help us understand more about our patients’ disease burden.”
In terms of screening, moderator Andre Goy, MD, said that with more screening comes a higher incidence of cancer. “If we had a way to detect things earlier, the outcome would be better. I believe in our in my lifetime, in our lifetime, we will have a way to detect early and mitigate early cancer cells and properties with targeted therapy and immune therapy potentially, by but before we get there, just the ability to screen is something we [can achieve].” Goy is physician in chief of Hackensack Meridian Health Oncology Care Transformation Services, and chairman, chief physician officer, and lymphoma division chief at John Theurer Cancer Center at Hackensack University Medical Center.
Goy noted that only 3% of patients who have a heavy history of smoking are screened with a routine, low-resolution CT. “If you’re able to add NGS papers show that [we] are able improve the ability to diagnose potential nodules instead of waiting for them to grow.”
Ip noted that success in solid tumors with liquid biopsy is a promising signal for what is to come in hematologic malignancies. “We want to choose the best treatments and the best sequence of treatments and the best targeted drugs, if possible, for our patients, because we want to eradicate their disease,” Ip added. “And hopefully, NGS can become a value add to become more predictive in this sense, because I think that’s something that is sorely needed, especially from the patient perspective.”
“[NGS] is used being used across a very large number of trials [to identify] surrogate prognostic biomarkers, and just in general for exploratory discovery of predictive response signals,” Ip said. “That certainly will continue as we understand more about the cellular tumor microenvironment for a lot of these cancers. It is quickly moving to the bedside. And we already have certain tests that we use and providers are using these tests and making decisions based off of them. It’s something that is already adding value today.”
In adding value, Ip noted that these tests are not likely to replace imaging standard, but supplement them. As it stands, the possibility of NGS lowering costs would come down to uptakes in the clinic, but for now pose the same insurance battles as accepted standards such as PET scans. “I’m on the phone all the time with insurance companies arguing for a PET scan,” Ip said. “Standard of care certainly weighs in on my response right now [as to whether NGS will replace imaging]. PET scans or imaging techniques are highly trusted by the medical community. There’s data that [show] liquid biopsies and NGS certainly could replace them, but I don’t think that's going to be the case. More likely, it’s going to augment the tools that we have to diagnose and treat.”