[1355] Minimal Residual Disease (MRD) Detection in Chronic Lymphocytic Leukemia (CLL): Flow Cytometry (FC) or Immunohistochemistry (IHC)?

Catalina Amador-Ortiz, David M Menke, Riccardo Valdez, Liuyan Jiang, Timm M Michael, William G Morice, Dragan Jevremovic, Tait D Shanafelt, Curtis A Hanson. Mayo Clinic, Rochester, MN; Mayo Clinic, Jacksonville, FL; Mayo Clinic, Scottsdale, AZ

Background: MRD detection in CLL is increasingly important as chemoimmunotherapy (CIT) regimens are effective in reducing tumor burden or eliminating detectable disease. Post-treatment MRD has been shown to predict durability of remission. Current multi-color FC can detect CLL cells to 0.005%. While FC is the current standard for MRD detection, IHC has not been studied in this setting. This study's goal was to compare IHC and FC in detecting CLL MRD.
Design: 66 patients with confirmed CLL that had been treated with CIT and had follow-up bone marrow (BM) biopsies for MRD assessment were selected. H&E BM sections and IHC for CD3, CD5, CD23 and PAX5 were assessed in all cases for % CLL involvement. High-sensitivity FC was done in BM aspirates in all cases. 500,000 events were collected and analyzed with a single tube, 6-color panel (CD45, CD19, CD20, CD5, kappa, lambda).
Results: Histologically, 36 cases (55%) had sufficient lymphoid infiltrates to diagnose CLL involvement; 13 had subtle, non-diagnostic interstitial lymphocytic infiltrates; 17 had no obvious infiltrate. 43 cases (65%) had diagnostic PAX5+ aggregates; 9 of the 23 PAX5- cases showed singly dispersed PAX5+ lymphocytes. 46 cases (70%) were CD23+: lymphoid aggregates only in 35, lymphoid aggregates with scattered CD23+ B cells in 10, and only dispersed CD23+ B cells in 1. CD5+ B-cells were seen in 32 (48%) cases; in 34 cases the stain was too weak to interpret or could not be separated from CD5+/CD3+ T-cells. When IHC stains were interpreted together, 50 cases (76%) were considered positive: ≤5% involvement in 26, 10% in 16, and ≥20% in 8. MRD detection by FC identified a clonal B-cell population (0.01%-42.10%; med=0.75%) in 50 cases (76%); in 16 cases no clonal B cells were found.
IHC and FC were concordant in 58 cases (88%): IHC+/FC+=46; IHC-/FC-=12. 4 BM were IHC+/FC-; PAX5+/CD23+ lymphoid aggregates were ≤5% of the BM cellularity in 3 and 10% in 1. The 4 IHC-/FC+ had flow MRD of 0.005%-0.02%.
Conclusions: There is excellent concordance between IHC and FC for MRD BM monitoring in CLL. However, neither IHC nor FC detected all cases with MRD. Although FC is regarded as the standard, 4 cases were IHC+/FC- which may be due to poor cell recovery or hemodilution. Similarly, the 4 IHC-/FC+ MRD were likely due to lack of BM aggregate formation by the clonal B-cells. MRD detection by IHC was enhanced by both CD23 and PAX5. A combination of high-sensitivity FC and CD23/PAX5 IHC may be the best approach for MRD detection in CLL patients.
Category: Hematopathology

Monday, March 19, 2012 1:00 PM

Poster Session II # 204, Monday Afternoon


Close Window