How many cells with CLL-like phenotype (that is essentially CD19+ CD5+) are there in normal peripheral blood? Quite important question for understanding the cellular origin and for proper estimation of MRD in CLL. Lets not further deepen into human B-cell biology, consider PB CD5+ B-cells "circulating marginal zone B-lymphocytes" and try to make multiple independent estimations. See joint GEIL-GOELAMS trial by Durrieu et al. in Cytometry Part B: Clinical Cytometry early view for details.

Two antibody panels were tested in 49 apparently healthy volunteers.

1. CD19+/CD5+/CD43int/CD79blo showed 1.17 × 10−4 (95% CI: 0.63–2.03 × 10−4) of WBC

2. CD19+/CD5+/CD81lo/CD22lo showed 0.76 × 10−4 (95% CI: 0.41–1.48 × 10−4) of WBC

Notably these 2 panels are used in the gating strategy for MRD evaluation as per 2007 standardized approach. However 2 other panels in Rawstron recommendations (those with CD38/CD20 and CD45/CD14/CD19/CD3) were not included. Light chain specific Abs were also not used because of presumably polyclonal nature of studed cells.

In partial dilution experiments both 4-colored combinations were able to discriminate CLL cells from normal cells in dilutions up to 10-4. Using ROC curves authors also showed sensitivity of 87.5% and a specificity of 91% for the threshold of 4x10-4 for discrimination between normal CD19+CD5+ cells and CLL cells. In clinical experiments there were 9,8% of pts with CLL-like cells in the gray zone between 1x10-4 and 4x10-4 - those cases where clinical correlation and repeated testing (or perhaps following the 2007 recommendations) might help in elucidating MRD status.

However all of this might become uninteresting with the advent of ROR1-specific Abs.