Introduction WT1 mRNA levels assessed by q-PCR have been employed as a marker of remission in AML patients undergoing intensive chemotherapy (Nomdedeu J. et al. Leukemia 2013;27:2157) and hematopoietic cell transplantation (HCT) (Nomdedeu J et al. Biol Blood Marrow Transplant 2018;24:55). Patients with adjusted bone marrow WT1 mRNA levels greater than 100 copies showed a high probability of relapse. Despite the usefulness of WT1 quantification in this and other settings, this marker is scarcely employed as an MRD target because levels greater than >100 copies can also be observed in normal bone marrows. Furthermore, 10-20% of AML at diagnosis and most chronic myelomonocytic leukemia (CMML) cases lack WT1 upregulation. The high levels of bone marrow WT1 copy numbers observed in most AML, decline in responding patients, although the utility, if any, of WT1 levels in long-term survivors of AML is currently unknown. We investigated the frequency and biological findings of mCR AML samples showing WT1 upregulation(>100 copies) compared to that mCR AML with concordant low WT1 levels. Patients and methods The study included 717 bone marrow samples from 80 AML patients (23 patients with RUNX1-RUNX1T1, 18 pts with CBFb-MYH11, four pts with NPM1 mutations, and 35 pts with PML-RARa). Samples were obtained at diagnosis (58 samples) and the follow-up (659 samples, all in mCR and without relapses) from one month to 219 months after diagnosis. Chimeric transcripts and WT1 levels were investigated by standardized real-time methods (ELN). Hemoglobin and leukocyte counts were also recorded. Flow cytometry immunophenotyping was also performed using four-color flow cytometry. We analyzed the percentage of CD34+, CD117+, CD123+ cells, and the Myeloid/Lymphoid CD34 ratio. nCounter Stem cell panel (Nanostring) was used to investigate the differences between that AML in mCR, showing high (>100) and low (<100) WT1 levels. In an additional cohort (11 samples: 4 with WT1>100 and 7 with WT1 <100), we performed bulk RNAseq (Novogene) to investigate enriched genes in samples with high WT1 levels. Results Thirty-five patients had at least one sample with WT1 levels greater than 100 copies (44%) despite being in mCR. Twelve samples had WT1 levels greater than 300 copies. In one case, this level reached 770 copies. These WT1 "peak" samples were distributed as follows: 49/306 PML-RARa, 25/161 RUNX1-RUNX1T1, 10/174 CBFb-MYH11, and 1/18 NPM1. Patients showing WT1 peaks were older than those without elevations (47 years vs. 36 years, p:0.003) For PML-RARa samples, "peaking" was associated with low levels of hemoglobin (133 g/L for <100 copies vs. 125 g/L >100 copies, p:0.009). For RUNX1-RUNX1T1, "peaking" was linked to less platelets (132x109 vs. 197x109 p:0.05) , higher CD34 (1.5% vs. 1% p:0.05) and also higher CD117+ cells (1.6 vs.1%, p:0.01). Lastly, CBFb-MYH11+ samples with WT1 peaks had higher leukocyte counts that those showing low WT1 levels (4.5x109 vs. 2.9x109, p:0.006). Nanostring comparisons between mCR WT1 high vs. low samples showed upregulation of TOR1A, PIK3CB, PTPN6, and PGM1 with downregulation of HMGA1, HOXB2 and CDC6. RNA seq analysis detected an enrichment of SUZ12 targets in those samples with high WT1 levels. Conclusions Bone marrow mRNA WT1 levels commonly exceed the clinical relevant value of 100 copies in AML samples in mCR (13%). Peak samples were related to peripheral blood values and/or raised numbers of immature bone marrow cells (CD34, CD117). Bone marrow WT1 although unable to accurately predict relapses in long-term survivors of AML may be a non-invasive sensor of a yet undefined myeloid state.
Read full abstract