Molecular Characterization of Chromosomal Abnormalities In Myelodysplastic Syndrome and Acute Myeloid Leukemia: Validation of An MLPA Protocol and Analysis Method for Use In a Diagnostic Setting

Current diagnostic screening strategies for copy number variations (CNVs) in myelodysplastic syndrome (MDS) and acute myeloid leukemia (AML) include fluorescence in situ hybridization (FISH) or karyotyping, both of which are time-consuming, costly, laborious, and lacking in resolution. Multiplex ligation-dependent probe amplification (MLPA) can be used to detect copy number changes in multiple loci simultaneously in a single PCR reaction, and boasts a resolution down to single exons. To adapt MLPA for use in routine clinical diagnostics, we have developed and validated a protocol for automatic data analysis and interpretation of common chromosomal abnormalities in MDS/AML.

The study used a training set of 45 healthy subjects to establish a normal reference range for each individual probe. Using these ranges we built an automated Excel spreadsheet-based analysis system, which included multiple quality checks, and flagged samples failing these quality controls. Each probe was given a call of “no mutation detected,” “deletion,” or “gain,” based on whether the normalized ratio fell within or outside of the empirically-determined normal range for that probe. We then analyzed over 100 leukemia cases tested by FISH, including both suspected myeloid leukemia samples and suspected chronic lymphocytic leukemia (CLL) samples. Documented chromosomal abnormalities in CLL include 11q-, 17p- (loss of TP53), and trisomy 12, all of which had the potential to be detected by the probes in the MDS MLPA probemix. The greater prevalence of CLL and its associated CNVs provided additional positive controls for the validation of the MDS MLPA probemix and our analysis method.

The empirically-determined normal ranges demonstrated that some probes varied widely (3 standard deviation [3SD] normal range of 0.46–1.54), while others were extremely reliable (3SD normal range of 0.84–1.16). The MLPA assay demonstrated excellent overall accuracy (>90%) and specificity (>93%) for both suspected myeloid and CLL samples when compared to FISH. The sensitivity of the MLPA assay is somewhat lower than that of FISH, requiring a probe-dependent 20–40% positivity for a given CNV to be detected. However in several cases, the MDS MLPA assay was able to detect additional lesions too small to be seen by FISH.

For MLPA, the total process-to-report time, including data analysis, is 2–3 days, versus the 7–10 days required for FISH analysis. In addition, the MLPA assay is substantially cheaper and considerably less labor-intensive than FISH. Our improved MLPA assay protocol and analysis method provides a clinically robust, multiplexed, high-throughput, high-resolution, and low-cost solution for detection of copy number changes in MDS/AML, and can therefore be used as a first-line screening test in a clinical laboratory.

Donahue: Quest Diagnostics Inc.: Employment. Abdool: Quest Diagnostics Inc.: Employment. Wohlgemuth: Quest Diagnostics Inc.: Employment. Yeh: Quest Diagnostics Inc.: Employment.