Abstract
Background In patients diagnosed with AML, the implementation of leukemia gene testing in accordance with the ELN 2022 guidelines is of paramount importance for the stratification of prognoses, thereby facilitating the determination of optimal treatment strategies. Presently, comprehensive NGS-based methods are extensively employed for these classifications; however, implementing them globally, including in Japan, is constrained by several factors in daily clinical practice, such as cost per sample, turn-around time, and technical hurdles in quality control.
Aims The objective of this study was to develop a cost-effective screening system capable of simultaneously screening multiple genes necessary for prognostic stratification of known targets. In the previous report, we selected three specific mutations: NPM1 mutations, FLT3-ITD mutations, and in-frame indel mutations in the bZIP domain of the CEBPA and developed assay panel. Since these mutations can be detected based on the length of DNA fragments, we adopted a multiplex fragment analysis method using a capillary electrophoresis sequencer (CES), which is cost-effective, highly sensitive and competent for various length of fragments.
Methods We assessed the clinical performance of our assay panel measured 53 de novo AML patient samples (median age, 66 years; 36 males and 18 females). We employed the Sanger sequencing method as a control method. We detected the presence or absence of each mutation in the patient samples and calculated the percentage of agreement with the Sanger sequencing analysis.
Clinical samples were obtained from Bioresource Center, Kobe University Hospital in accordance with the Declaration of Helsinki and under an approved Kobe University institutional research protocol (B240242).
Results We confirmed 100% accuracy in comparison with the Sanger sequencing, which is the control method. The breakdown of mutations contained in clinical samples was as follows: 12 samples with NPM1 mutations, 11 samples with FLT3-ITD mutations, and 1 sample with in-frame indel mutations in the bZIP domain of the CEBPA, although the number of positive results includes duplicate calculations for samples containing multiple genetic mutations. For samples with a low mutation allelic ratio, the presence of mutations was confirmed, but they were below the sensitivity of the Sanger sequencing and could not be sequenced. Also, it was found that some FLT3-ITD mutation-positive samples contained ITDs of multiple sizes, with up to four different sizes (63, 69, 84, and 175 bp) of ITDs identified.
Discussion In our previous report, it was demonstrated that the assay panel achieved 100% accuracy in agreement with Sanger sequencing when using simulated samples created by mixing synthetic genes with healthy human peripheral blood mononuclear cells, and detected mutations at a rate of 2.9%.
In this clinical evaluation, utilizing real patient's cells, the method exhibited substantial agreement with the control method, thereby reinforcing its efficacy.
It is noteworthy that even low-frequency variants were effectively identified. This case suggests that the assay panel may be useful for classifying the disease appropriately and determining treatment strategies in MDS with a tumor content of less than 20%, based on the 5th edition of the WHO classification. Furthermore, the test demonstrated an adequate capability to detect samples with multiple FLT3-ITDs of multiple sizes. Due to the inherent characteristics of NGS technology, the detection and annotation of these types of mutation remain intricate, and the design of primers that are specific to each mutation in qPCR is an arduous task.
Conclusion The three-gene assay panel using our developed fragment analysis system in CES was able to detect the mutations of three key genes that can be used for minimum decision-making to perform prognosis stratification of de novo AML in accordance with current ELN guidelines in a manner that is rapid, highly sensitive, and cost-effective in comparison to NGS. Additionally, it can handle a wide range of mutation patterns without customization, a feature that distinguishes it from qPCR.
We consider that this will enable rapid risk stratification prior to standard treatment to be provided to clinical settings in diverse regions around the world.
