New guidelines in laboratory assessment of micro-angiopathic hemolysis create cautious optimism.
Accurate identification of schistocytes is a primary responsibility of the hematology laboratory. The presence of schistocytes in the peripheral blood raises the possibility of thrombotic microangiopathic anemia (TMA), including thrombotic thrombocytopenic purpura (TTP) and hemolytic uremic syndrome (HUS).
Despite the clinical significance of schistocytes, the published literature regarding their formal identification and quantification is relatively scant. To address this situation, the International Council for Standardization in Hematology (ICSH) has recently published new consensus recommendations.1 According to this group, all of the following RBC forms should be considered “schistocytes” (Fig 1):
- small fragments with sharp angles or spines (triangular) and straight or distorted cytoplasmic borders; usually staining darkly, occasionally pale;
- microcrescents, which are similar to triangular forms but with a round contour on one side;
- helmet cells, which show an amputated zone of cytoplasm with a straight border and sharp angles;
- keratocytes, which show prominent spicules surrounding a concave border; and
- microspherocytes (small darkly-staining spheres without central pallor) should only be counted as schistocytes in the presence of the other schistocyte shapes.
ICSH guidelines also call for schistocytes to be reported as a percentage of total RBCs, based on a formal count of 1,000 erythrocytes, if they represent the dominant morphologic RBC abnormality. In contrast, if fragments are noted as part of a spectrum of RBC abnormalities, then a qualitative report is preferred.
The guidelines recommend 1% schistocytes as a threshold for suspecting a diagnosis of TMA. Certain types of automated hematology analyzers are capable of enumerating schistocytes. A normal result via this method has a high negative predictive value, though cases flagged for increased fragments must be confirmed by morphologic review. Automated fragment analysis may give a false negative result in cases with macrocytosis >105 fl due to the larger size of the fragments.
These guidelines are a useful resource for the laboratory professional charged with the evaluation of schistocytes. However, we would add several observations relating to their practical implementation. First, the guidelines were specifically developed for improving the accuracy and reproducibility of counting schistocytes in TMA and not for other causes of RBC fragmentation (Fig. 2). In routine practice, specialized laboratory technologists evaluate peripheral blood smears for a variety of abnormalities of which schistocytes are included. Most laboratories receive few smears from patients with TTP and, thus, need to base criteria for identification and quantitation of schistocytes on their patient population.
It is important that technicians do not “overcall” schistocytes in common disorders, such as iron deficiency, in which small hypochromic red cell fragments are common. To avoid this problem, laboratories in the United States usually comply with the College of American Pathologists definition and do not call fragments with central pallor as schistocytes (for proficiency testing or in routine practice); such cells are best considered non-specific poikilocytes (Fig. 3). The ICSH recommendations address this issue only in that “very small rounded, irregular crenated or distorted poikilocytes, and RBC microvesicles should also not be included in the [schistocyte] count in the context of anisopoikilocytosis.” The recommendation to include microspherocytes in some schistocyte counts is also likely to confuse the bench technologist, who often does not know the associated clinical history when assessing a blood film. Given these constraints, the ICSH recommendations may be best suited for pathologists.
Second, the recommendation to formally report a quantitative percentage of schistocytes (if these are the dominant RBC abnormality) raises operational questions. Should a normal range also be given, as is customary when reporting quantitative laboratory findings? If so, it would seem logical for each laboratory to establish its own reference range for schistocytes.
Alternatively, because the ICSH recommends 1% as the threshold value for clinical significance in schistocyte percentage, perhaps “<1%” could be an appropriate “normal” value. Either strategy, however, would likely necessitate quality control procedures to ensure that individual technologists adhere to the reference range. But even with the best quality control procedures, interobserver variability may arise due to dissimilarity in schistocyte numbers in different areas of some smears (Fig. 4). The ICSH does not provide data to evaluate the reproducibility of its recommendations in practice.
The potential impact on clinical decision-making must be evaluated. A numerical report of a minimally elevated schistocyte percentage could be an added temptation for clinicians to initiate plasma exchange in inappropriate situations, such as DIC. A schistocyte percentage that falls within or only slightly above a numerical normal range on a laboratory report might obscure the important point that TMA remains in the differential diagnosis even when schistocytes are not markedly increased.
- Zini G, et al. International Journal of Laboratory Hematology 2012;34:107-16.