The prognostic value of flow cytometry determination of CD38 and CD49d on leukemic cells of Argentinian CLL patients.

CORDINI, GREGORIO; GAMBERALE ROMINA

Edinburgh

Workshop; XVIII International Workshop on Chronic Lymphocytic Leukemia,; 2019

CLLshows a highly variable clinical course, which can be predicted by severalbiologic markers, including the mutational status of immunoglobulin heavy variable(IGHV) gene, genomic abnormalities, and expression of CD38 and CD49d. The aimof this study was to determine whether combined flow cytometry determination ofCD38 and CD49d can be used to predict time to first treatment (TTFT) in ourcohort of CLL patients. The study includes peripheral blood samples from 159patients affected by typical B-CLL. The median age at diagnosis was 73 years witha male to female ratio of 1.65. The median follow-up of our patients was 10.3years; 60 out of 159 required treatment (37.7%) with a median TTFT of 8.3 years(95% CI 5.7-11.8). Thedistribution of clinical stages at diagnosis according to Rai?s criteria was: 36.8%,stage 0; 26.5%, stage I; 14.7%, stage II; 5.9%, stage III; 16.2%, stage IV. Fluorescence in situ hybridization (FISH) testing was available for 69 patients: del13q14 26.9%, del11q227.2%, del17p13 15.9% and trisomy 12 15.2%. The IGHV mutational status,available for 82 out of 159 cases showed 63.4% of mutated- and 36.6% ofunmutated-IGHV gene configuration. The expression of CD38 and CD49d on CD19+cells was evaluated in our laboratory by three-color immunofluorescence (seeFigure 1). Patients with ≥7%of CD19+ cells expressing CD38 were considered CD38+ (1) and patients with ≥30% of CD19+ cells expressing CD49d wereconsidered CD49d+ (2). Figure 1A shows thepercentage of CD19+ cells expressing CD38 and CD49d for each of the159 CLL patients. We found that 46 patients (28.9%) were CD38+ and CD49d+, while 67 patientswere negative for both markers (42.2%). The remaining samples were discordantfor the expression of CD38 and CD49d: 28 patients (17.6%) were CD38+CD49d-and 18 patients (11.3%) were CD38-CD49d+. As previouslydescribed (3), the association between the two antigenswas highly significant (p=0.001). Figure1B shows that, in agreement with previous reports (3), both CD38+ and CD49+CLL patients had highly significant shorter TTFT, as compared tocases in which CLL cells expressed the antigens below the established cut-offvalue (CD38- or CD49-). CD38-CLL patients did not reach the median of TTFT, while CD38+ onesdisplayed a median of 5.1 years (95% CI 3.8-6.4). Regarding CD49d, the medianof TTFT was 12.8 years (95% CI 10.3-15.3) for CD49d- CLLpatients and 6.9 years (95% CI 4.7 - 9.0) for CD49+ samples.When our patients were segregated in threegroups according to concordant expression of CD38 and CD49d (CD38-CD49d-and CD38+CD49d+) or discordant expressionof both markers (CD38-CD49d+ or CD38+CD49d-), we found statistically significantdifferences in TTFT among the three groups (Figure2A). CD38-CD49d- CLL patients did not reach the median of TTFT and, at fifteen years,65% did not required treatment. On the contrary, CD38+CD49d+CLL patients had significantly shorter TTFT with a median of 4 years (95% CI 1.2?6.3).Discordant subgroup of CLL patients showed 11.3 years of TTFT (95% CI 5.5-17.2).Remarkably, patients with RAI II-IV clinical stages, unmutated, with trisomy 12and who required treatment, preferentially belonged to CD38+CD49d+subgroup (Table I). Nosignificant differences were observed in laboratory parameters except for beta-2-microglobulinand lactate dehydrogenase values, which are higher in CD38+CD49d+subgroup.As expected, unmuted-CLL patients hadshorter TTFT compared to mutated-CLL samples (5.3 vs 11.1 years, p<0.001). Itwas previously reported that high levels of CD49d is a valuable marker for TTFTin patients with mutated-IGHV status (4). When mutated-CLL patients weresegregated based on the expression of CD38 and CD49, we found statisticallysignificant differences in TTFT (Figure 2B).Thus, mutated-CD38+CD49d+CLL patients had a significantly shorter TTFT (4 years, CI 1-7)compared to mutated-discordant patients (11.4 years, CI 5.8-16.8) or mutated-CD38-CD49d-patients, which did not reach the median of TTFT.In conclusion, we have confirmed in ourCLL patients that co-expression of high levels of CD38 and CD49d are associatedwith advanced-stage patients and worse prognosis. We here also provide evidencethat the combined analysis of CD38 and CD49d expression can predict TTFT and evenmore important, was able to separate CLL patients with mutated-IGHV into groupswith different TTFT. Flow cytometry determination of CD38 andCD49d   1.          Br J Haematol. 2005;130(4):549-57.2.          JClinOncol.2014;32(9):897-904.3.          Leukemia.2006;20(3):523-5.4.          BrJ Haematol. 2016;172(1):48-55.