Background: CD20 expression is still a controversial issue regarding its prognostic value for therapy outcome. While up to 50% of patients relapse within four of years after anti-CD20 directed treatment and constitute a need for a salvage or relapse/refractory treatment, only few publications address the longterm influence on CD20 expression assessed by flow cytometry. The technical complexity of the method and factors affecting the variability of measurements of samples at different time points and from different sources/patients are well recognized by the field and addressed by different approaches as standardization of instruments, automated gating and alignment of population or normalization. Aims: In order to evaluate long term influence of anti-CD20 directed therapy on CD20 expression in patients with chronic lymphocytic leukemia who showed only weak MRD responsiveness to therapy in different clinical trials, we aimed to establish a normalization approach based on the instrument standardization to reduce the proportion of technical variance within data. Methods: Mean fluorescence intensities (MFI) of peak seven of standardized fluorescence beads from daily quality control and instrument standardization of flow cytometers according to the EuroFlow protocol were used to establish a normalization approach. Proof of prinicple was performed using the residual peaks covering the MFI range provided by the beads. To assess different sources of variance ANOVA of a 3x3x3 experiment evaluating marker expressions on different cell populations in the blood of healthy donors using Quantibrite-PE beads was performed addressing technical variation (day, instrument) and biological variation. CD20 and CD19 MFI on CLL cells were retrospectively assessed from MRD measurements of peripheral blood samples of patients enrolled in the phase II GCLLSG trials CLL2-BIG (n=57), -BAG (n=58), -BIO (n=58), -BCG (n=22), combining facultative Bendamustin (B) debulking with either Obinutuzumab (G), Ofatumumab (O), Venetoclax (A), Ibrutinib (I) or Idelalisib (C). Results: Long term shifts of fluorescence intensities, and coefficients of variation were reduced 2 to 10 times across the MFI range covered by the fluorescence beads without distorting the longitudinal course of instrument performance. Most of the variance in assessments of molecules equivalent soluble fluorochrome from MFIs of NK cell and B cell markers in peripheral blood of healthy donors were related to inter-donor variations (p<0.002). Normalized MFIs for CD19 and CD20 on CLL cells do not significantly differ from the measured values in the Mann-Whitney-U test (p>0.85). Higher CD20 expression at therapy start seems to be correlated to strong MRD response solely in the CLL2-BIO trial (p=0.036, Fig. 1). Prognostic factors TP53-, IGHV-, and NOTCH1- mutation did not seem to correlate with the observed CD20 expression differences. Therapy strata showed differences (statistically not significant) in the proportion of firstline and r/r treated patients. Figure 1 Pre-treatment CD20 MFI on CLL cells of patients categorized according to treatment and variable (strong, medium, weak, and no) MRD response. Image:Summary/Conclusion: Following standardized staining and instrument monitoring, it should be possible to robustly assess longitudinal biological variations of marker expression based on MFI values. In a cross trial comparison Obinutzumab showed highest proportion of patients with strong MRD response independent from initial CD20 expression, whereas strong response to Ofatumumab seems to correlate with higher CD20 expression levels.
Read full abstract