Real-Time Quantitative PCR (RQ-PCR) Detection of Minimal Residual Disease (MRD) by Optimized WT1 Assay To Enhance Risk Stratification in Acute Myeloid Leukemia (AML): A European LeukemiaNet Study.

Relapse remains the major cause of treatment failure in AML; however, RQ-PCR assays to detect leukemic fusion transcripts have been shown to identify reliably those patients at highest risk of relapse, allowing development of a more individualized treatment approach. In cases lacking a leukemia-specific MRD marker, quantification of genes over-expressed in AML e.g. Wilms’ Tumor gene (WT1) could provide more a precise measurement of disease response and quality of remission, potentially enhancing risk scores such as the one developed by the MRC used to identify those patients most and least likely to benefit from allogeneic transplant in first CR. WT1 is over-expressed in at least 75% of AML cases. Within the European LeukemiaNet we systematically evaluated 9 published and “in house” WT1 assays. Assays were excluded due to demonstrated lack of RNA-specificity or location within the 3′ region of the gene which has been shown to be subject to deletion or mutation in AML. An assay located within the 5′ region associated with superior sensitivity was ultimately selected following parallel testing in 11 labs and evaluated in 238 diagnostic peripheral blood (PB) and 386 bone marrow (BM) samples. WT1 was over-expressed in the majority, with comparable levels in PB and BM (PB - median 4637 WT1 copies/10⁴ABL copies, range 0–1132709; BM - median 7212, 0–750571), as compared to normal BM (median 19.8, 0–213), PB (median 0.01, 0.01–47.6) and PBSCs (median 6.1, 0–39). In cases over-expressing WT1, kinetics of transcript reduction were evaluated following induction. A greater response was associated with a significantly reduced risk of relapse (hazard ratio 0.65 per log reduction (95% CI 0.43–0.96), p=0.03), although this failed to remain significant when adjusted for age, presenting WBC and cytogenetic risk group, which are key variables in the MRC risk index. Indeed, there was a highly significant correlation between larger log reduction in normalized WT1 transcript level and better risk score (p=0.0001). Sequential analysis of PB and BM samples from 15 AML cases with low WT1 expression (<250 copies) showed no significant modulation in transcript level on regeneration after chemotherapy, indicating that in WT1+ AML, transcript levels detected in follow-up samples reliably reflect disease status. This study provides evidence that recognized pre-treatment risk factors for relapse correlate closely with kinetics of response to induction therapy and lend support to the evaluation of early assessment of MRD to develop more robust risk scores, to enhance risk stratification and identify those patients most suited to proceed rapidly to allogeneic transplant.