Validation of blood counters for quality control of platelet concentrates with high platelet counts

Introduction

Platelet (PLT) transfusion is the standard treatment in thrombocytopenia or thrombocytopathia. PLTs are transfused prophylactically or therapeutically in hemato-oncology patients while PLTs are given to treat coagulopathy in trauma patients [1], [2], [3], [4]. Clinicians rely on a standard PLT dose of the PLT products released from the blood transfusion service. The quality of the PLT products has to be proven in at least 1% of the collected units according to the current guidelines [5], [6].

Electronic blood counters are used in quality control of PLT concentrates [7], [8], [9], [10], [11], [12]. However, the devices were developed for counting peripheral blood samples and the old generation of apparatuses showed limitations in counting very low and high PLT numbers. As a consequence, strategies had to be developed in the quality control of PLT concentrates [7], [9], [10], [12], [13], [14]. The aim of the present study is to report the validation of different counting methods for PLT concentrates with respect to precision.

Materials and methods

Plateletpheresis donors fulfilling current national and European eligibility criteria [5], [6] underwent PLT collection using the Trima Accel (Terumo BCT, Lakewood, CO, USA) or the MCS+ (Haemonetics, Braintree, MA, USA) blood cell separators. Minimum hemoglobin levels of donor eligibility were 13.5 g/dL for men and 12.5 g/dL for women. Donors were monitored for body temperature, blood pressure, heart rate and weight prior to plateletpheresis. All donors signed informed consent. After PLT collection, sample aliquots were drawn from the platelet concentrates (PCs) under aseptic conditions at day 1, 4, 5 and 6 of storage. PCs were stored under constant conditions at 22±2°C on a flad bed agitator (Helmer Inc., Noblesville, IN, USA).

Samples were measured using four different blood counter devices. Cell Dyn Emerald (Abbott, Wiesbaden, Germany) and Sysmex K-4500 (Sysmex, Norderstedt, Germany) use the impedance counting method while Cell Dyn Ruby (Abbott, Wiesbaden, Germany) uses the scatter-light optical method. The Sysmex X-550 can be run using the impedance or the fluorescence optical method.

For precision performances of the apparatuses, samples were measured at least 5 times serially and the coefficients of variation were calculated and compared with manufacturers’ requirements.

Additionally, 50 peripheral blood samples were compared and the normal hemogram parameters (red blood cells [RBC], white blood cells [WBC], hemoglobin [HGB], hematocrit [HCT], PLT) and mean corpuscular volume (MCV) were analyzed.

Data are presented as mean±standard deviation if not stated otherwise. Statistical analysis was done using GraphPad software (San Diego, CA, USA). The Shapiro-Wilk test was used to test for normal distribution. An analysis of variance (ANOVA) was used to test for statistical significance. A p-value of <0.05 was considered as statistically significant.

Results

The comparison showed significant differences between the blood counter devices in measuring high PLT counts over the storage period. The CELL-DYN-Emerald, the Sysmex K-4500 and the Sysmex XN-550 with the optical counting method measured significantly higher PLT counts compared to the CELL-DYN-Ruby and the Sysmex XN-550 with the impedance counting method (Figure 1; p<0.0001) independent of their principle of measurement. The manufacturers provide comparable coefficients of variation. We achieved similar results for all counters (data not shown). Results of the peripheral blood count parameters were only comparable for the parameter WBC. Statistical analysis of the other hemogram blood parameters showed specific significant differences as shown in Figure 2. Highest PLT counts were measured using the XN-550 with the impedance counting technology. This trend was not observed in PCs where Sysmex K-4500, CELL-DYN-Ruby and Sysmex XN-550 with optical counting method showed higher PLT counts.

Figure 1:

Blood counter comparison of platelet concentrates (n=18).

Platelet count (PLT) in 10⁹/L measured on day 1 (day after collection), day 4, 5 and 6 (2 days after shelf life).

Figure 2:

Hemogram parameters obtained with different blood counters.

Discussion

Quality control of blood components is mandatory according to the current guidelines [5], [6]. The required number of quality control is 1% of all manufactured PCs per month. Counting of high PLT numbers is crucial because of the different counting techniques of various apparatuses. Blood count analyzers use either a fixed or a floating discriminator to distinguish RBC from PLT, and use a device-specific formula to correct for large PLT that are counted as RBC, and for small RBC that are counted as PLT [15], [16], [17]. In the absence of red cells (as is the case in PCs), this often leads to wrong platelet numbers. Improved PLT discrimination uses two light-scatter angles of PLT and non-PLT particles as they pass through a laser beam 17]. The volume and refractive index of each PLT and particle are derived from the light-scatter measurements. Together, these two measurements provide improved PLT discrimination compared with one-dimensional methods.

The first generation of electronic blood counters showed poor linearity for very low and high PLT counts. As a consequence, alternative counting methods such as microscopic counting had to be established in severely thrombocytopenic patients. On the other hand, quality control of PCs with high PLT counts was implemented in transfusion medicine using a dilutional step prior to PLT counting [7], [9], [10], [11], [12], [13], [14]. Today, modern blood count devices such as the Sysmex XN-550 measure PLTs in the range of 0–5.000×10³/μL with a linearity of ±5–6% making dilution of the PC unnecessary 18]. However, the current guidelines of the German Medical Association on quality assurance in medical laboratory examination define only permissible relative deviations for PLTs in a range of >150–700×10³/μL [19]. As the counting technique of the blood counters may have an impact on the PLT count, the present study was undertaken to validate blood counters with different measurement methods with respect to PLT counts. It is the responsibility of blood transfusion services to validate apparatuses used in quality control of blood components. Validation is an objective proof that requirements are met. Objective proof can be done through observation, testing or measurement.

Our results showed significant differences among the studied counters in analyzing high PLT counts over the storage period (Figure 1). The scatter-light optical count of CELL DYN Ruby and the impedance count of Sysmex XN-550 showed the lowest PLT counts. The impedance count of CELL DYN Ruby and the optical fluorescence count of Sysmex XN-550 resulted in higher PLT counts. Thus, the counting method had no impact on the PLT count but the peculiarity of the manufacturer. Daves et al. reported on a decrease of PLT counts in stored peripheral blood samples using the Sysmex XN hematological analyser [20]. After 24 h, a meaningful bias was observed for both impedance and optical PLT counting methods compared to the baseline sample. The decrease of PLT counts of our PCs may also be biased by dilution due to daily sampling.

When standard hemogram parameters of peripheral blood were analyzed, comparable WBC counts were obtained by the different blood counters. However, significant differences were found for RBC, HBG, HCT, MCV and PLTs (Figure 2). This may be due to the fact that the samples were not balanced for gender and age. Nebe and co-workers reported on age-, gender- and analyzer-specific analytes [21] in 1158 healthy males and females. Based on this study, reference ranges for the complete blood count, the differential blood count and reticulocytes were obtained.

Conclusions

Our study showed the importance of counter validations focusing on PCs with a high PLT count. Not only the technique of measuring, but also the manufacturers’ peculiarities seem to play an important role.