Correlation of FLT3 Mutations with Expression of CD7 in Acute Myeloid Leukemia
Junaid Baqai, Domnita Crisan. William Beaumont Hospital, Royal Oak, MI
Background: FLT3 mutations are common in acute myeloid leukemia (AML), particularly in cytogenetics negative cases. Internal tandem duplication (ITD) and also point mutations affecting aspartic acid 835 (D835) are reported. A previous study demonstrated aberrant expression of CD7 on blasts in 11 of 15 de novo AML cases with FLT3/ITD mutations. The goals of the study are: (i) to expand the evaluation of this association to a larger group of patients, and (ii) to evaluate the association of CD7 aberrant expression in AMLs with D835 mutation, not previously done, (iii) to evaluate if CD7 aberrant expression may serve as a surrogate market for predicting FLT3 mutational status.
Design: Patients were retrospectively identified by diagnosis of AML, FLT3 mutation analysis performed, and flow cytometry (FC) evaluation including CD7 expression on blasts, between February 2004 and December 2011. The FLT3 mutation analysis was performed on DNA extracted from 149 patients (32 peripheral bloods, 117 bone marrows), using PCR amplification and analysis of amplicons by gel electrophoresis for ITD mutation, and by restriction endonuclease digestion fragment analysis for the D835 point mutation. Chart reviews were used for finding AML types, cytogenetic findings and FC evaluation. The statistical analysis was performed using SAS version 9.3 for Windows, R version 2.15.1 for Windows, and StatXact9.
Results: Out of 149 patients, 28 positive for FLT3 (20 ITD positive, 6 D835 positive and 2 ITD and D835 positive) were identified. CD7 was positive in 19 of 28 positive cases (67.9%). Out of 20 ITD positive cases, 13 were positive for CD7 (65%). Out of 6 D835 positive cases, 5 were positive for CD7 (83.3%). The association of CD7 positivity and FLT3 positivity was found to be significant using Pearson chi-square test (Χ2=9.24, p= 0.0024). Using FC, CD7 has a low positive predictive value (PPV) of 30.2% but a negative predictive value (NPV) of 89.5%.
|Sensitivity (%)||Specificity (%)||PPV (%)||NPV (%)||Accuracy (%)|