CD123 expression and its correlation with mutation profile in acute myeloid leukemia.

Abstract

e18524 Background: CD123, the interleukin-3 receptor α-chain, is frequently overexpressed by acute myeloid leukemia (AML) blasts and may represent a potential target for chemo-immunotherapy. Methods: We retrospectively studied consecutive patients (pts) with AML to assess CD123 expression by blasts, its correlation with the 2022 World Health Organization (WHO-5) and International Consensus Classification (ICC), and related mutational profiles. Results: 99 pts were included, with a median age of 66 yrs (range, 26–90). 59% had newly diagnosed (ND) AML, 24% had relapse/refractory (RR) AML, 12% had secondary AML (S-AML), and 5% had therapy-related AML (t-AML). The median percentage of blasts expressing CD123 was 85% (range, 13–100), with 57 pts (58%) having ≥80% expression. The median CD123 fluorescence intensity (MFI) was 1189 (interquartile range (IQR), 606–2235). The median percentage of blasts expressing CD123 was 84% (range, 14–99) in ND AML, 87% (range, 61–100) in RR AML, and 82% (range, 13–99) in S-AML and t-AML, with median CD123-MFI 1216 (IQR, 581–2058), 1128 (IQR, 803–2008), and 1006 (IQR, 391–3482), respectively. Age (66 vs. 67 yrs), blast percentage (40% vs. 53%), median LDH (380 vs. 294), and adverse karyotype (64% vs. 53%) were similar between pts with <80% and >80% CD123 expression. The most frequent AML categories for ND AML, based on the WHO-5 and ICC classifications, are shown in the Table. 5 pts with acute erythroid leukemia, as per WHO-5, were classified as TP53 mutated by ICC and had a median CD123 expression of 32% (range, 13-37). The most common mutations in the entire cohort were TP53 (32%), DNMT3A (20%), RUNX1 (18%), TET2 (16%), FLT3 -ITD (10%), NRAS (10%), U2AF1 (10%), IDH1 (10%), SRSF2 (10%), and NPM1 (9%). FLT3 -ITD was more common in pts with CD123 expression ≥80% (16% vs. 2%, p=0.04), whereas NRAS was more frequent in those with CD123 <80% (19% vs. 4%, p=0.02). Trends showed higher frequencies of TP53 (38% vs. 28%), SRSF2 (14% vs. 7%), and ASXL1 (12% vs. 5%) in pts with CD123 <80%. RUNX1 (21% vs. 14%) tended to be more common in those with CD123 ≥80%. Similar trends were seen using the third quartile CD123-MFI cutoff, although the difference in NRAS was not statistically significant. Conclusions: CD123 expression is consistent across AML subtypes; shows a positive correlation with FLT3 -ITD, a negative correlation with NRAS , and a trend towards TP53 , suggesting potential predictive markers for anti-CD123 therapy response. Genes (entire cohort) CD123 <80%n=42 (%) CD123 ≥80%n=57 (%) p-value Classification (ND AML) CD123 <80%n=25 (%) CD123 ≥80%n=33 (%) p-value TP53 16 (38) 16 (28) 0.3 WHO 0.9 DMT3A 7 (17) 13 (23) 0.5 Myelodysplasia-related (MR) 12 (48) 17 (52) RUNX1 6 (14) 12 (21) 0.4 Defined by differentiation 5 (20) 4 (12) TET2 8 (19) 8 (14) 0.5 NPM1 3 (12) 6 (18) FLT3-ITD 1 (2) 9 (16) 0.04 ICC 0.4 NRAS 8 (19) 2 (4) 0.02 TP53 mutated 8 (32) 5 (15) U2AF1 4 (10) 6 (11) 1.0 MR 7 (28) 13 (39) IDH1 3 (7) 7 (12) 0.5 NPM1 3 (12) 6 (18) SRSF2 6 (14) 4 (7) 0.3 NOS 2 (8) 3 (9)