Raman spectroscopy for the assessment of acute myeloid leukemia: A proof of concept study

Acute myeloid leukemia (AML) is a proliferative neoplasm, that if not properly treated can rapidly cause a fatal outcome. The diagnosis of AML is challenging and the first diagnostic step is the count of the percentage of blasts (immature cells) in bone marrow and blood sample, and their morphological characterization. This evaluation is still performed manually with a bright field light microscope. Here we report results of a study applying Raman spectroscopy for analysis of samples from two patients affected by two AML subtypes characterized by a different maturation stage in the neutrophilic lineage. Ten representative cells per sample were selected and analyzed with high-resolution confocal Raman microscopy by scanning 64x64 (4096) points in a confocal layer through the volume of the whole cell. The average spectrum of each cell was then used to obtain a highly reproducible mean fingerprint of the two different AML subtypes. We demonstrate that Raman spectroscopy efficiently distinguishes these different AML subtypes. The molecular interpretation of the substantial differences between the subtypes is related to granulocytic enzymes (e.g. myeloperoxidase and cytochrome b₅₅₈), in agreement with different stages of maturation of the two considered AML subtypes. These results are promising for the development of a new, objective, automated and label-free Raman based methods for the diagnosis and first assessment of AML.