Optimizing Wilms tumor 1 thresholds for measurable residual disease monitoring in acute myeloid leukemia: Improved sensitivity and concordance with nucleophosmin 1 in a single-center validation study.
In acute myeloid leukemia (AML), measurable residual disease (MRD) assessment is essential for predicting relapse and guiding therapy decision-making. Nucleophosmin 1 (NPM1) mutations are reliable MRD markers but apply to only ∼30% of patients with AML. Wilms tumor 1 (WT1) expression monitoring is applicable to a broader population but the European LeukemiaNet (ELN) threshold of 50 WT1 copies per 104 ABL copies (0.5%) may be too high, which limits sensitivity.
With WT1 expression data from 100 healthy controls, this study established a revised WT1 threshold of seven copies per 104 ABL copies (0.07%). Its performance was retrospectively validated against NPM1 in 308 paired follow-up samples from 63 patients with NPM1-mutated AML, and against the core binding factor (CBF) β-MYH11 fusion transcripts in 83 samples from 12 patients. Statistical analyses included concordance, sensitivity/specificity, survival estimates, and model comparison with the Akaike information criterion.
Compared to the ELN cutoff, the revised threshold showed higher concordance with NPM1 (78.6% vs. 73%) and sensitivity (54% vs. 24%; p < .0001) and acceptable specificity (91% vs. 98%; p = .002). In survival analyses, the seven-copy cutoff demonstrated stronger prognostic value than the ELN threshold, particularly after consolidation therapy. In the CBFB::MYH11 cohort, the two thresholds showed similar concordance but WT1 positivity with the revised cutoff preceded relapse in selected patients.
In AML, the revised WT1 threshold of seven copies per 104 ABL copies enhanced MRD sensitivity while maintaining specificity, improving concordance with NPM1, and showing prognostic relevance. These findings support the clinical value of locally optimized WT1 thresholds, and highlight the need for prospective multicenter validation and harmonization of WT1-based MRD monitoring.
With WT1 expression data from 100 healthy controls, this study established a revised WT1 threshold of seven copies per 104 ABL copies (0.07%). Its performance was retrospectively validated against NPM1 in 308 paired follow-up samples from 63 patients with NPM1-mutated AML, and against the core binding factor (CBF) β-MYH11 fusion transcripts in 83 samples from 12 patients. Statistical analyses included concordance, sensitivity/specificity, survival estimates, and model comparison with the Akaike information criterion.
Compared to the ELN cutoff, the revised threshold showed higher concordance with NPM1 (78.6% vs. 73%) and sensitivity (54% vs. 24%; p < .0001) and acceptable specificity (91% vs. 98%; p = .002). In survival analyses, the seven-copy cutoff demonstrated stronger prognostic value than the ELN threshold, particularly after consolidation therapy. In the CBFB::MYH11 cohort, the two thresholds showed similar concordance but WT1 positivity with the revised cutoff preceded relapse in selected patients.
In AML, the revised WT1 threshold of seven copies per 104 ABL copies enhanced MRD sensitivity while maintaining specificity, improving concordance with NPM1, and showing prognostic relevance. These findings support the clinical value of locally optimized WT1 thresholds, and highlight the need for prospective multicenter validation and harmonization of WT1-based MRD monitoring.
Authors
Barriere Barriere, Bourgne Bourgne, Tassin Tassin, Establie Establie, Ramos Ramos, Veronese Veronese, Ravinet Ravinet, Tchirkov Tchirkov, Berger Berger, Guieze Guieze, De Antonio De Antonio, Lebecque Lebecque
View on Pubmed