EHA Library - The official digital education library of European Hematology Association (EHA)

Red blood cell (RBC) membrane disorders constitute one of the major causes of chronic hereditary hemolytic anemia. Main RBC membrane disorders, namely hereditary spherocytosis (HS), hereditary elliptocytosis (HE) and hereditary stomatocytosis (HSt), alter membrane cohesion, membrane mechanical stability, and RBC volume, respectively. As a consequence, RBC deformability is compromised leading to their premature removal from circulation, manifested as hemolytic anemia. New generation osmotic gradient ektacytometry has become a powerful procedure for measuring red blood cell deformability and therefore for the diagnosis of red blood cell membrane disorders.

A total of 75 patients with chronic hemolytic anemia oriented as hereditary RBC membrane disorders (hemoglobin disorders discarded and negative Coombs test) were included during a period comprised between January 2015 and August 2016. Normal controls were obtained from blood donors. Osmotic gradient ektacytometry was performed using the osmoscan module of the Laser-assisted Optical Rotational Deformability Cell Analyzer: LoRRca MaxSis (RR Mechatronics). Evaluation of osmoscan parameters robustness for HS diagnosis was performed using the receiver operating characteristic (ROC) curve analysis. The optimal cut-off was determined as the one with the highest likelihood ratio. Statistical analysis was operated with GraphPad Prism.

Specific patterns of osmoscan LoRRca MaxSis were observed for each individual membranopathy. All HS curves were bell shaped but two different profiles were identified both presenting increased Omin, and decreased EImax and AUC. HE curves showed a characteristic trapezoidal shape with a decreased EImax, Omax and AUC. dHSt curve was bell shaped with a specific decrease in Ohyper and a slight increase in EImin. Reference ranges for each osmoscan parameter were established with 171 healthy subjects and compared with the values of the parameters obtained from the different RBC membrane disorders. ROC curve analysis was performed for HS and each one of the non-HS groups separately. The results determined that EImax was the parameter that better separated HS from normal controls and dHSt, while the Omin was the best to separate HS from HE. The optimal EImax cut-off to differentiate HS from normal controls was <0.5975 (sensitivity 98.46%, specificity 99.42%), while the optimal Omin cut-off to differentiate HS from HE was >159.0 (sensitivity 95.38%, specificity 85.71%). Expressing the results as % of variation in relation to the mean of our normal controls, the best combination of parameters for HS diagnosis would be EImax < -3% and Omin > 5.2%. This combination of parameters (EImax <-3% and Omin >5.2%) was used as criteria to classify all the 246 samples included in the present study, and the result showed 62 samples detected as HS and 184 as no-HS. Of the 62 patients identified as HS, 61 were real HS (specificity 98.38%) and 1 was an HE. On the other hand, 4 HS patients were identified as non-HS (sensibility 93.85%).

We can conclude that, the inclusion of LoRRca osmoscan as a screening test in RBC membrane diagnostic workflow will signify an important advance for the accurate diagnosis of HS patients, as well as for the identification of HE and specially dHSt patients.