Optimized method to enhance the limit of detection for large FLT3 internal tandem duplications

Abstract

Abstract Internal tandem duplications (ITDs) in the fms-related tyrosine kinase 3 gene (FLT3) are recurrently detected in acute myeloid leukemia (AML) and are associated with a poor prognosis. Evaluation for the presence of ITDs in AML is considered standard of care; detection alters risk stratification and therapeutic selection. ITDs can range in size, being a minimum of three base pairs to over 300 base pairs. Larger ITDs are considered rare; however, it is unclear whether this represents true prevalence or a consequence of conventional assay design. The FDA-approved companion diagnostic test for the detection of FLT3 ITDs has a six times higher limit of detection for large ITDs (~300 bp) compared to medium or small ITDs. Further, the conventionally used short-read sequencing approaches to detect ITDs can also struggle to detect larger duplications. Therefore, it is unknown how many patients may have large ITDs that have gone undetected. We developed an approach to improve the detection limit for large ITDs that uses polymerase chain reaction (PCR) and capillary electrophoresis (CE) with a spiked-in reference. Larger ITDs undergo a predictable and reproducible negative PCR bias, which can be adjusted once the effect has been quantitatively characterized. FLT3 ITDs are reported as an allele ratio, the ratio of the ITD to the wild-type allele, as a biomarker for prognosis; the spike-in reference allows for quantification of an allele ratio despite distinct CE run conditions for the ITD and wild-type alleles. Using existing commercially available controls and manufactured plasmid controls, sensitivity was optimized, including temperature, injection time, polymerase, and PCR primers. With this method, the detection limit for large ITD variants is comparable to the limit of detection for medium and small ITDs in the FDA-approved companion diagnostic assay. We now have a robust tool to evaluate the accuracy, precision, linearity, and limit of detection at different ITD lengths. Enhanced identification of large FLT3 ITDs may modify the prevalence estimates of large ITDs in AML and ultimately alter patient management.