Photo courtesy of NIH
NEW YORK—Combining 2 technologies—flow cytometry and high-throughput sequencing (HTS)—produces a very sensitive approach to detecting minimal residual disease (MRD) in chronic lymphocytic leukemia (CLL), according to a speaker at Lymphoma & Myeloma 2016.
The approach is both reproducible and widely accessible, added the speaker, Peter Hillmen, MB ChB, PhD, of St James’s University Hospital in Leeds, UK.
“PCR [polymerase chain reaction] is more sensitive than flow cytometry,” he noted, “but it is probably not necessary to assess response by that criteria.”
Features to consider when choosing a technology for detecting MRD include sensitivity, specificity for the patient and/or the disease, applicability to all patients or to an individual, the platform—flow cytometry, conventional molecular PCR, or next-generation sequencing—and which tissue to use, blood or bone marrow.
In the era of immunotherapy, Dr Hillmen said, assessment should be performed at least 2 months after completion of immunotherapy to get a reliable assessment of MRD, particularly after treatment with alemtuzumab, rituximab, and other antibodies targeting CLL.
History of MRD analysis
The definition of MRD hasn’t changed since 2008, when the International Workshop on CLL updated National Cancer Institute guidelines. It is still 1 single cell in 10,000 (10-4) leukocytes, regardless of the tissue used.
Prior to the mid-1990s, there were limited options for assessing MRD, Dr Hillmen said.
Based on the profound remissions patients experienced with alemtuzumab, investigators began to develop assays to assess MRD.
“[W]e started standardizing these assays around 2007,” Dr Hillmen said, and a standardized assay was used prospectively in clinical trials beginning in 2012.
Several technologies can be used to assess MRD.
Flow cytometry using 6 colors or 8 colors—to simplify the assay and to make it more sensitive—is a multiparameter assessment of CLL phenotype that is not clonality-based.
Allele-specific oligonucleotide PCR (ASO-PCR) is laborious to perform, Dr Hillmen said, but it’s very sensitive.
“[I]t probably shouldn’t be considered as an MRD test,” he said, since it uses patient-specific primers, not consensus primers.
HTS provides an increasing amount of information on B-cell sequences and enumeration of the CLL-specific immunoglobulin gene, “and I would move it towards being approved as a regulatory endpoint,” Dr Hillmen asserted.
Flow cytometry and HTS
A consensus document by the European Research Initiative on CLL (ERIC) identified and validated a flow cytometric approach to MRD assessment in parallel with HTS.
According to the ERIC investigators, flow cytometry had to utilize a core panel of 6 antigens used by most labs—CD19, CD20, CD5, CD43, CD79b, and CD81. And the markers used had to quantitate cells to a level of 0.01% (10-4).
Assays had to be independent and compatible with older, established therapies as well as newer treatments.
For example, 6-color flow had to be effective with fludarabine, cyclophosphamide, and rituximab (FCR) regimens, as well as effective with the novel agents ibrutinib and venetoclax (ABT-199).
Compared to PCR, multiparameter flow cytometry is more convenient, Dr Hillmen noted.
And while PCR is more sensitive than flow cytometry (sensitive to 10-5 to 10-6), it is more difficult to apply to large clinical trials because the assay must be validated for each patient.
The investigators validated the flow cytometric approach in 450 patients on FCR-type therapy enrolled in the ADMIRE and ARCTIC trials.
They assessed MRD in patients’ bone marrow 3 months after the last course of treatment and presented the data at EHA 2015 (Rawstron AC, abstract S794).
They found that all patients who were MRD negative, including 9 patients with partial responses, achieved a significantly better progression-free survival (PFS) than patients who had achieved a complete response but were still MRD positive.
The parallel analysis of HTS showed good concordance with flow cytometry at the 0.01% (10-4) level.
Peripheral blood or bone marrow?
“[T]he blood is a tissue which can be used, but it’s certainly not as sensitive as the bone marrow,” Dr Hillmen said. “And depending upon what we are using MRD for, the marrow is probably a better tissue, with some exceptions.”
Data from the ADMIRE and ARCTIC trials confirmed that 177 patients on FCR-based therapy who were negative in the bone marrow were always negative in the blood. However, a quarter of patients negative in the blood were positive in the bone marrow.
Investigators followed the same patients on FCR-based therapy for 3 years and found no difference in outcome in terms of PFS for patients negative in peripheral blood and positive in the bone marrow (PB-/BM+) and those negative in peripheral blood and negative in the bone marrow (PB-/BM-) (Rawstron abstract S794).
But for patients on alemtuzumab, with the same analysis, those who were PB-/BM+ did less well and had similar PFS to those who were PB+/BM+.
And at a follow-up of 4 years or longer, patients on FCR-based therapy who were PB-/BM- had superior outcomes than those who were PB-/BM+.
“So as a predictive marker, the bone marrow is a better tissue to look at, but peripheral blood negativity also can predict with FCR but not with agents such as alemtuzumab,” Dr Hillmen summarized.
Prognostic value of MRD assessment
Multivariate analysis of a 10-year follow-up of 133 CLL patients revealed that MRD level and adverse cytogenetics were the only significant parameters in terms of PFS.
And in terms of overall survival (OS), MRD level, prior treatment, Binet stage, and age were significant.
Sixty-seven of these patients had been treated with chemoimmunotherapy, 31 with single-agent chemotherapy, 7 had autologous stem cell transplants, and 28 had prior exposure to alemtuzumab.
In terms of survival beyond 10 or 15 years, previously untreated patients who were MRD negative after their first therapy had significantly better PFS and OS than previously treated patients who were MRD negative and patients with or without prior treatment who were MRD positive (P<0.001).
“Consistently, MRD, regardless of therapy, is the most important prognostic marker,” Dr Hillmen said.
Data from MD Anderson Cancer Center showed that 75% of patients treated first-line with FCR who achieved a complete response were MRD negative.
And patients who achieved MRD negativity had significantly better PFS (P<0.001) and OS (P=0.006) than patients who remained MRD positive.
In the ADMIRE and ARCTIC trials mentioned earlier, patients who achieved MRD negativity in the marrow at 3 months post-therapy also had significantly better PFS (P<0.0001) and OS (P=0.0002) than those who were MRD positive.
For every log of positivity, Dr Hillmen said, patients have a worse survival. Conversely, for every log reduction in MRD level, there is a 33% reduction in risk for disease progression.
“MRD is probably the most important prognostic marker we have,” he said. “We need to look at MRD levels with novel agents and use it to define duration of therapy, maybe use it to define additional therapy if patients are stalled in their response.”