3.2 Evolution of High-Sensitivity, Multi-Color Flow Cytometric Immunophenotyping for Minimal Residual Disease Detection in Chronic Lymphocytic Leukemia: Peripheral Blood versus Bone Marrow?

Abstract

Background: Detection of minimal residual disease (MRD) in chronic lymphocytic leukemia (CLL) patients has become increasingly important as modern chemoimmunotherapy (CIT) regimens have become effective in reducing tumor burden or even eliminating detectable disease. From a clinical perspective, MRD status posttreatment has been shown to predict durability of remission. Current 6and 8-color flow cytometry assays can easily detect CLL cells to the 0.005% level (1 in 20,000 events). Historically, peripheral blood (PB) has been the standard source for evaluating CLL, while bone marrow (BM) studies have been used primarily to evaluate patients with unexplained cytopenias. A previous study based on 4-color flow cytometry analysis suggested a high concordance (87%) of MRD analysis between PB and BM as long as patients had not received monoclonal antibody therapy in the preceding 8-12 weeks. The goal of the present study was to compare MRD detection in PB and BM from CLL patients after first-line CIT using a highly sensitive 6-color flow assay to determine whether there were any differences in MRD levels. Methods: Thirty-one paired PB and BM specimens were obtained from 25 patients with a confirmed diagnosis of CLL. Twenty-one paired samples were obtained from 15 previously untreated patients on a phase II clinical trial evaluating lenalidomidebased consolidation therapy after 6 cycles of pentostatin, cyclophosphamide, and rituximab induction therapy. Ten paired samples were from previously untreated patients on a phase II clinical trial evaluating pentostatin, cyclophosphamide, and ofatumumab. Twentyeight paired specimens were obtained 3 months after rituximab/ ofatumumab and the remaining 3 specimens were obtained 8-12 weeks following rituximab therapy. Individual patient samples were processed and analyzed on the same day. High-sensitivity flow cytometric immunophenotyping was performed in all specimens. A total of 500,000 events were collected in all cases and analyzed on a Becton-Dickinson Canto II flow cytometer. The standard single-tube, 6-color antibody panel included CD45, CD19, CD20, CD5, and kappa and lambda immunoglobulin light chains; additional antibodies were added as needed. Positive events were based on the identification of lymphoid cells by light scatter followed by gating on CD19positive B-cells, then dual expression of CD20(lo)/CD5 cells, and finally evaluation for kappa/lambda monoclonality. The level of MRD was assessed by percentage MRD calculated as (CLL events)/ (total live, non-aggregated WBCs) 100. This assay was developed in a clinical laboratory and validated by comparing it to a previous standard 4-color assay in 562 specimens (unpublished data). Validation studies confirmed a higher level of sensitivity with the 6-color assay and the ability to consistently detect MRD or monoclonal B-cell lymphocytosis events to the 0.005% level (25/500,000 events). Results: PB and BM MRD analysis was concordant in 20/31 (65%) paired samples using the single-tube, 6-color antibody analysis; 15/31 paired samples had detectable MRD in both PB and BM, and 5 had no detectable MRD in either PB or BM. Eleven of 31 (35%) studies showed detectable MRD in BM, but no CLL cells in PB. Considering BM MRD analysis as the standard, the sensitivity of the PB MRD assay in this cohort was 58% (15/26). The amount of MRD was on average 5-fold greater in BM as compared to PB. The level of MRD in BM (range 0.006-6.008%; median 0.765%) was greater than in PB (range 0.035-38.726%; median 0.161%) in all cases, with a median difference of 0.343% (p 0.0054). In the 15 paired specimens with MRD in both PB and BM, the level of residual disease was 1% in 12/15 PB and 10/15 BM specimens. Among the 11 paired samples in which MRD was only detected in BM, 10 (91%) had 1% clonal B-cells present on the BM MRD assay, with positivity ranging from 0.005% to 3.231% (median 0.061%); this was 10 times lower than the median value of BM MRD for samples in which both PB and BM were MRD-positive, indicating a lower level of disease burden in the group with discordant MRD results on BM and PB analysis. Conclusion: A single-tube, 6-color flow assay is a highly sensitive approach for the detection of MRD in CLL. The level of MRD following effective CIT was 5-fold higher in BM as compared to PB. A negative PB MRD study cannot be used as a surrogate for the absence of BM disease in CLL patients.