PB2344 EVALUATION OF THE PREDICTIVE VALUE OF HEMATOPOIETIC PROGENITOR CELL COUNT IN SYSMEX XN ANALYZERS FOR CD34+ STEM CELL MOBILIZATION IN PERIPHERAL BLOOD PREVIOUS TO THE APHERESIS

Background: In peripheral blood stem cell (PBSC) mobilization after daily G‐CSF ± chemotherapy we need to check the presence of CD34+ stem cells in peripheral blood (PB) on days 4‐6 to start their collection. CD34+ cells measurement is done by flow cytometry. Sysmex XN hematology analyzers can count hematopoietic progenitor cells (HPC). This analyser is cheaper, faster and requires minimal training. Aims: To evaluate the predictive value of HPC count for CD34+ stem cell mobilization. Methods: Stem cells were mobilized in autologous patients by G‐CSF ± chemotherapy. We also used Plerixafor the day before the apheresis in poor mobilizers. Allogeneic donors were mobilized by G‐CSF without Plerixafor. HPC and CD34+ cells were quantified in samples collected from PB the day before the apheresis and the day of the apheresis before starting it. Good mobilization was defined as CD34+ cell count in the PB was ≥ 10 × 10 6 /L in autologous patients who will receive one transplant (Auto1 group), ≥ 20 × 10 6 /L in autologous patients with mieloma diagnosis who will receive two transplants (Auto2 group) and ≥ 20 × 10 6 /L in allogeneic donors (Allo group). ROC curve was generated to visualize the sensitivity (SE), specificity (SP), positive predictive value (PPV) and negative predictive value (NPV) for various measures of HPC values to predict the number of CD34+ cells mobilized, and the area under the curve (AUC) was measured. Positive HPC cut‐offs were calculated for predicting a CD34+ cell count ≥ threshold with 100 % PPV and 100 % SP and negative HPC cut‐offs were calculated for predicting a CD34+ cell count < threshold with 100% NPV and 100 % SE. Then we calculate correctly classified patients with the positive and negative HPC cut‐offs. Results: 80 consecutive patients were enrolled (50 Auto and 30 Allo). 105 apheresis harvests were done: 72 Auto (50 first, 21 second and 1 third apheresis) and 33 Allo (30 first and 3 second apheresis). In the autologous group the diagnosis were 25 lymphoma, 36 mieloma, 4 Hodgkin, 6 AML and 1 germ cell tumor. Correlation between HPCs and CD34+ cells . In the Auto group the correlation between HPC and CD34+ cells was good in the samples pre‐apheresis (r: 0.9666) and poor in the samples collected the day before (r: 0.5744). In the Allo group the correlation was moderate for the samples collected pre‐apheresis (r: 0.7874). Predictive value of HPC count in the mobilization phase . The AUC was 0.7694 and 0.8304 for the samples collected the day before the apheresis in the Auto1 and Auto2. See table 1 for positive and negative HPC cut‐offs to predict a good or a bad mobilization. Globally with the samples collected the day before the apheresis we could correctly classified 17/50 patients (34 %). The AUC was 0.8690 and 0.8833 with the samples obtained just pre‐apheresis in the Auto1 and Auto2. Globally in the autologous group with the samples collected pre‐apheresis we could correctly classified 30/50 patients (60 %). In the Allo group all the healthy donors had a CD34+ cell count ≥ 20 × 10 6 /L in the pre‐apheresis sample, so we could not generate ROC curve. Summary/Conclusion: In the autologous group the HPC count in samples collected the day before the apheresis is not useful because it predicts mobilization of CD34+ cells only in a third of the patients. In the autologous group the HPC count in samples collected pre‐apheresis was useful to predict the mobilization of CD34+ cells in 60 % of the patients. In the allogeneic group with healthy donors all of them mobilized well.