Regarding: automated red blood cell differential analysis on a multi-angle scatter/fluorescence hematology analyzer.

Abstract

In a recently published article, Kim et al. (1) describe a new method of analyzing red blood cells in human samples on the modified Abbott CELL DYN 4000 hematology analyzer. The article focuses on technical considerations related to Abbott technology. This new method is compared and contrasted through the article to red blood cell (RBC) methods available on Bayer hematology platforms (H* family and ADVIA 120 hematology analyzers). We found a significant number of inaccuracies throughout the article that should be pointed out and clarified. On page 44, Kim et al. (1) state: “More advanced hematology analyzers in terms of RBC morphology are the H*1, H*2, H*3, and ADVIA 120 hematology analyzers (Bayer, Tarrytown, NY). All these systems are designed to measure RBC [volume] V and [Hgb concentration] HC on [a] cell-by-cell basis using a two-dimensional (2D) matrix, a combination of two selected FSC signals…. Drawbacks of the method are that the 2D matrix does not provide any information on [abnormal RBCs] AbnRBCs or schistocytes. The scatter signals from AbnRBCs either fall on wrong locations or completely fall off the precalibrated 2D matrix for V and HC measurements because the 2D matrix was constructed based on the theory for perfect spheres.” The Bayer method, as well as the Abbott method presented in this article, uses isovolumetric sphering, studied extensively by Ponder (2), and adapted by Kim and Ornstein (3), to assure that all RBCs of a given size (V) and refractive index (HC) produce a unique scatter signal in a flow cytometer, independent of their orientation. Mie scattering theory predicts that the magnitude of monochromatic light scattered by a homogeneous sphere is a joint function of sphere size, refractive index, and absorption (4). Tycko (5) appreciated that a carefully selected pair of angular intervals produce a one-to-one correspondence between cell volume and refractive index on the one hand, and the associated pair of scattering intensities on the other hand (especially at wavelengths at which hemoglobin absorbs very weakly, e.g., at 633 nm and 670 nm). For the first time, this permitted quantitative measurement and plotting of cells in cytograms, with V and HC as linear axes (6). In most red cell diseases studied using the Bayer method, the cellular abnormality leads to a broadening of the height and width of the RBC cluster in the V/HC due to the changes in MCV and CHCM. This provides meaningful information about abnormal red cells (abbreviated as AbnRBCs in the cited article), including distributional abnormalities (RDW and HDW) and quantitative enumeration and flagging of samples with microcytic/macrocytic and hypochromic/hyperchromic cells outside the normal V/HC region. Kim et al.'s article states on numerous occasions that schistocytes are not identified with Bayer technology. Schistocytes are RBC fragments, smaller than typical red blood cells, and usually larger than platelets. In a sphering agent, these cells also sphere. Because of their hemoglobin content and concentration, their refractive indices are almost always higher than those of platelets and they are well-separated from platelets in the Bayer RBC cytograms. The RBC fragment count can be included in the ADVIA 120 report screen. Thus, the same information on schistocytes is reported by the Bayer method and the proposed Abbott method. On page 45, Kim et al. state: “When the system is perfectly standardized, all the sphered RBCs (all normal RBCs are perfectly sphered in this reagent) will be distributed very closely around the surface and abnormally shaped RBCs such as sickle cells, will generate signals that are more distant from the surface.” This statement implies that the 3-D spaces occupied by sphered and nonsphered RBCs do not overlap. The authors have not demonstrated this. Consequently, their assertion that the abnormal RBC count corresponds exclusively to signals at some distance from the sphered-cell surface is also undemonstrated. Unsphered abnormal cells occupy a range of points that may include the surface, depending upon how they are oriented with respect to incident radiation. Sphered abnormal cells must appear on the surface, and cannot be distinguished from normal cells by distance. It should also be pointed out that the cited article contains unsupported statements concerning the clinical use of reticulocyte hemoglobin content (CHr). The CHr parameter is used as a hematologic “gold standard” for iron-deficient states (7), and its clinical utility is well documented in the literature in a variety of clinical applications (8-10).