Technological advances in flow cytometry and real-time PCR are improving detection of low levels of minimal residual disease (MRD) in bone marrow, including in acute myeloid leukemia (AML). In AML, more precise detection of MRD would allow the patient's response to therapy to be gauged in order to pinpoint more precisely who might benefit from transplant.
“This is a pertinent time for monitoring of low levels of residual disease—it is moving higher up on the agenda of clinical trial groups,” said David Grimwade, MD, PhD, Professor of Molecular Hematology in the Department of Medical & Molecular Genetics at King's College London School of Medicine.
Dr, Grimwade and colleagues Paresh Vyas and Sylvie Freeman wrote an article on the topic in the November issue of Current Opinion in Oncology (2010;22:656–663), reviewing recent studies highlighting the potential for novel and standardized assays to provide independent prognostic information and develop more personalized treatment approaches.
The article makes the case that there is a strong rationale for prospective trials investigating the merits of extending MRD detection to alter therapy and potentially improve outcome in other AML subtypes.
Major advances include establishment of optimized real-time quantitative PCR assays for Wilms' tumor suppressor (WT1) gene, commonly overexpressed and assumed to be a therapeutic target; and mutant nucleophosmin (NPM1) genes.
Also, the authors note, sequential MRD monitoring using an internationally standardized real-time PCR assay for PML-RARA fusion transcripts has been successfully used to guide molecularly targeted therapy in acute promyelocytic leukemia.
And there have also been significant advances in multiparameter flow cytometry to detect MRD, with introduction of 6-10 color technology and improved understanding of the immunophenotype of leukemic stem cells, the authors said.
Improve Risk Stratification
“Monitoring for MRD in leukemic cells is going to improve risk stratification in AML compared with conventional pre-treatment factors,” Dr. Grimwade said, in a telephone interview. “This is likely to be used for more reliable information as to which patient is more appropriate to transplant in first remission or which could be potentially cured with chemotherapy alone.”
At the moment, he said, decisions on transplant in AML are based mainly on upfront prognostic factors including patient age, presenting white cell count, and cytogenetics.
“The problem is, you identify full groups of patients but not the individual patient who is going to do well or badly,” he said. “You are not pinpointing the patient who is going to be cured with front-line therapy or the one who is likely to relapse.”
He said these two technologies are not especially new but have been steadily improving over time. Real-time PCR, for example, was invented in the 1990s, but it is only in the last decade that it has been investigated for use in directing a patient's treatment approach.
He said MRD assessment is now a well-established tool to determine risk-stratification in childhood acute lymphoblastic leukemia (ALL), which can be monitored in a more consistent way because all the lymphoid neoplasms have a clonal marker and each patient has his or her own fingerprint that can be followed.
“AML is much more heterogeneous, so you can't apply a common approach to all patients,” he explained. “You have to identify the different molecular types before even starting the monitoring, so there's no overarching molecular approach.”
Dr. Grimwade said that most technologies can detect only between a 1-log and 2-log reduction in leukemia using that sort of technique.
“But if you consider that the patient might have 10-to-the-12th leukemic cells, when diagnosed, and if they go into remission—defined as less than 5% leukemic cells—there could be 10-to-the-10th leukemic cells still in the patient, and yet you think they are in remission,” he said.
At this time it is not possible to know whether there are no cells left in the patient because no technique at the moment can go down to that resolution—“but it is certainly better than morphology.”
Multiparameter Flow Cytometry vs Real-Time PCR
Multiparameter flow cytometry defines patterns of proteins on the surface of individual leukemic cells that are different from normal bone marrow cells and define the diagnosis, Dr. Grimwade explained. The sensitivity and specificity depend on whether the patient has cells with an identifiable pattern.
Another limiting factor is that leukemic cells can change their characteristic patterns over time—”they can change their spots,” Dr. Grimwade said—and the assays might miss the impending relapse.
Real-time PCR is most helpful if the patient's cells have a particular mutation or a fusion gene made by a chromosomal translocation or balance chromosomal rearrangement, which is the case in acute promyelocytic leukemia (APL) (Grimwade et al: JCO 2009; 27:3650–3658).
He said the leukemia-specific markers for some patients are not known, so there is more work to be done in terms of characterization.
Besides organizing clinical trials to determine whether MRD monitoring improves patient survival, quality-of-life issues must also be weighed, Dr. Grimwade said. “It's yet to be determined what the benefit is in terms of improved survival when the patient didn't think it was worth taking regular bone marrow punctures.”
St. Jude Study
Dr. Grimwade cited a study done at St. Jude Children's Research Hospital that used flow cytometry to guide patient management in pediatric AML(Rubnitz JE, et al. Minimal residual disease-directed therapy for childhood acute myeloid leukaemia: results of the AML02 multicentre trial. Lancet Oncol 2010;11:543–552).
He said this was the first randomized, multicenter study in which MRD assessed by flow cytometry was included in risk classification in order to guide consolidation therapy in AML. The 216 patients were randomized to receive high- or low-dose induction chemotherapy. Consolidation therapy was guided by post-induction MRD levels as well as by genetic abnormalities at diagnosis.
The three-year event-free survival rate was 63% and the overall survival rate was 71%. Both outcomes compared favorably with results obtained with comparable treatment protocols that did not use MRD-based risk adapted treatment, Dr. Grimwade said.
Mikkael A. Sekeres, MD, MS, agreed with Dr. Grimwade that these tests must be standardized.
Dr. Sekeres, OT's Clinical Advisory Editor for Hematology/Oncology, who is Director of the Leukemia Program at Cleveland Clinic Taussig Cancer Institute and Chair of the Hematology/Oncology Pharmacy & Therapeutics Committee there, said the two hurdles to using detection of MRD for treatment planning are standardizing the tests, and demonstrating that using the tests to detect MRD as a prospective intervention makes a difference.
“One center's flow cytometry is not the same as another's,” Dr. Sekeres said in a telephone interview, explaining that standardization of tests has occurred internationally for PCR in CML, and now needs to be adapted for other subtypes of leukemia.
Dr. Sekeres also echoed Dr. Grimwade on the potential value of testing for MRD in leukemias. He said that if remission induction achieves a 3 to 4 log kill of cells, the number is reduced to below the level of detection, but there could still be a million leukemic cells in the body that routine tests such as bone marrow biopsy looking for morphology would not detect.
“In a younger person, we would give three to four more subsequent cycles of therapy to reduce that number 2 to 4 log kill every treatment, to get the cell count down to where the native immune system is able to eliminate the rest,” Dr. Sekeres said. “But if we are able to detect cells on a much more sensitive level and detect persistent disease quicker, we could potentially intervene earlier when there is MRD, preparing the patient for bone marrow transplant, for example, instead of giving post-remission chemotherapy.”
One difficulty in detecting MRD in most leukemias is that the patient's tumor must have a recognizable molecular fingerprint, and not everyone has that, he said: “If a patient expresses MPM-1 or FLT3 mutations or has a t(8;21) translocation or inversion of chromosome 16, you can detect that and see if they have disease relapsing. But not everybody has a definable molecular lesion associated with their leukemia.”
Dr. Sekeres said he believes that eventually there will be a molecular fingerprint for every patient, but that could take decades.
Call to Arms
Leukemia is one of the easier diseases to monitor for molecular abnormalities because these can often be detected in blood, so early interventions in leukemia may have an impact on patients destined to relapse. “But this hasn't been studied prospectively, and that's the call to arms in the article by Dr. Grimwade,” Dr. Sekeres said.
“Right now we've been able to demonstrate that technologies such as PCR or sophisticated flow cytometry are able to detect MRD and that accurately predicts patients destined to relapse versus those who aren't,” he said, speaking of AML as well as ALL and CML.
“What we haven't yet shown is that detecting MRD, and then intervening on that, makes a difference. Now is the time to identify these patients prospectively and intervene prospectively to see if the interventions make a difference.”