Red Blood Cell Mechanical Fragility Test For Clinical Research Applications

Abstract

Red blood cell (RBC) susceptibility to mechanically induced hemolysis, or RBC mechanical fragility (MF), is an important parameter in the characterization of erythrocyte membrane health. The rocker bead test (RBT) and associated calculated mechanical fragility index (MFI) is a simple method for the assessment of RBC MF. Requiring a minimum of 15.5 mL of blood and necessitating adjustment of hematocrit (Ht) to a ‘standard’ value (40%), the current RBT is not suitable for use in most studies involving human subjects. To address these limitations, we propose a 6.5 mL reduced volume RBT and corresponding modified MFI (MMFI) that does not require prior Ht adjustment. This new method was assessed for i) correlation to the existing test, ii) to quantify the effect of Ht on MFI, and iii) validation by reexamining the protective effect of plasma proteins on RBC MF. The reduced volume RBT strongly correlated (r = 0.941) with the established large volume RBT at matched Hts, and an equation was developed to calculate MMFI: a numerical estimation (R² = 0.923) of MFI if performed with the reduced volume RBT at “standard” (40%) Ht. An inversely proportional relationship was found between plasma protein concentration and RBC MF using the MMFI-reduced volume method, supporting previous literature findings. The new reduced volume RBT and modified MFI will allow for the measurement of RBC MF in clinical and preclinical studies involving humans or small animals.

Introduction

Red blood cell (RBC) mechanical fragility (MF) is a measure of erythrocyte susceptibility to mechanical stress. While a number of methods for the measurement of RBC MF have been proposed, a comparison of many existing methods has suggested that the rocker bead test (RBT) first described by Kameneva et al (1) is the most practical test for common use (2). An important parameter in the characterization of erythrocyte membrane health, RBC MF tests have proven useful in assessing donor blood quality during in vitro hemolysis testing of blood-contacting medical devices, sublethal cell damage induced by cell salvage suction devices, and in characterizing RBC storage lesion (3–5).

The original RBT mentioned earlier requires a minimum of 15.5 mL of RBC suspension. This large volume requirement has limited the use of the RBT to large animals, healthy adults, or banked blood, precluding assessment from adult and pediatric patients, as well as small animals. However, preliminary investigations have indicated certain disease states are associated with higher RBC MF compared to healthy control groups (6–8). Additionally, neonates have proven to have higher RBC MF than adults, potentially implicating a yet-to-be-investigated higher degree of sublethal cell trauma in pediatric patients from supraphysiological stresses like those found during mechanical circulatory support or extracorporeal membrane oxygenation (ECMO) (9).

In addition to sample volume requirements, the existing RBT is hematocrit (Ht) dependent. Comparisons between RBC MFIs can only be made if the RBT was performed at equal Hts. Blood obtained directly from patients would have varied Ht values, and manual adjustment requires additional blood volume and time. Therefore, development of an easily employed method or algorithm to calculate a standardized MF value despite sample measurement at varied Hts is certainly justified.

Materials and Methods

Experimental Design

Comparison of the traditional large volume RBT and our new, reduced volume RBT was accomplished by running these tests in parallel at the Ht values of 25, 30, 35, and 40% using matched bovine blood pools.

Human blood (n = 4) was used to characterize the effect of Ht on MFI and create a Ht-independent modified MF index (MMFI) for the reduced volume RBT. Each single-donor unit was assessed (n = 6) at four different Ht values (range 25–50%). Finally, validation of the reduced volume RBT’s ability to ascertain differences in RBC fragility was achieved by the variation of total plasma protein, previously shown to significantly affect RBC MF (1), using human RBCs from a single blood pool.

Blood Preparation

All blood samples used in these studies were purchased from certified biological supply companies. Venipuncture-obtained, bovine whole blood anticoagulated in K2-EDTA was purchased (Lampire no. 7200807; Lampire Biological Laboratories, Pipersville, PA, USA) and used 2 days following draw date. Adult donor human whole blood anticoagulated in K2-EDTA was purchased (Valley Biomedical no. HB1051K2; Valley Biomedical Products and Services, Inc., Winchester, VA, USA) and used within 4 days following withdrawal. Upon arrival, each unit was filtered (SQ40S blood transfusion filter; Pall Medical, Portsmouth, UK) to remove platelet microaggregates before storage at 0–4°C when not in use.

Experiments were performed using autologous donor pools. Ht variations were achieved by dilution with autologous plasma. Variation of total plasma protein was performed by dilution of plasma with phosphate-buffered saline. Upon resuspension at the desired parameters, Ht (SurePrep 75 mm heparinized self-sealing capillary tubes; BD Clay Adams, Sparks, MD, USA) and total hemoglobin (tHb) (OSM3 Hemoximeter; Radiometer, Copenhagen, Denmark) concentration were measured in every blood aliquot.

MF Measurement

The rocker bead experiments were set up and performed as described previously (10), differing only in tube preparation for the proposed method. Each traditional large volume RBT utilized five 7.0 mL silicone-coated, glass collection tubes (Red-Top Serum Vacutainer ref: 366431; Becton-Dickinson, Franklin Lakes, NJ, USA) filled with 3.0 mL of whole blood. Three experimental tubes contained five 1/8″ stainless steel ball bearings (BBs) each (BNMX-2, Type 316 balls; Small Parts, Inc., Miami Lakes, FL, USA), and the remaining two tubes without BBs served as controls. In contrast, the modified reduced volume method consisted of three tubes (two experimental with BBs, one control) using 3.0 mL plastic collection tubes (No Additive Vacutainer ref: 366703; Becton-Dickinson) filled with 2.0 mL of whole blood and only three BBs per experimental tube.

The 7.0 mL tubes were placed on a rocker (Type M79735 Platform Vari-Mix rocker; Barnstead Thermolyne Corp., Dubuque, IA, USA) with their tops touching the edge of the platform, and 3.0 mL tubes were placed with their bottoms aligned to the bottoms of the 7.0 mL tubes. Rocking was performed for 1 h at 18 cycles/min at a rocking angle ±17° from horizontal on a platform mixer (Figure 1A). The tubes were centrifuged for 15 min at 2750 g (CR412; Jouan, Inc., Winchester, VA, USA) at which point the supernatant was transferred to 1.5 mL microcentrifuge tubes and spun at 20,800 g for 20 min (Eppendorf 5417R, Westbury, NY, USA). Free hemoglobin (fHb) in the final supernatant was measured via spectrophotometer (Genesys 5; Spectronic Instruments, Inc., Columbus, OH, USA) using the Wintrobe method (absorbance at 540 nm) (11). All of the above procedures were carried out at room temperature (22 ± 1°C).

Figure 1.

(A) Schematic of reduced volume and large volume rocker bead tests running simultaneously and (B) comparison of MFI values obtained from reduced volume and large volume tests run simultaneously at matched hematocrits between 25 and 40% using bovine blood (Pearson’s correlation coefficient r = 0.941).

The dimensionless MF index (MFI) value for each experiment was calculated as follows:

$$\mathit{MFI}=\frac{\overline{{\mathit{fHb}}_{\mathit{BBs}}}-\overline{{\mathit{fHb}}_{\mathit{Control}}}}{\overline{\mathit{tHb}}\ast 1000-\overline{{\mathit{fHb}}_{\mathit{Control}}}}\ast 100$$ (1)

where tHb is the total hemoglobin concentration of the blood sample in g/dL, and fHb is the average free hemoglobin concentration of the BB containing or control tubes in mg/dL.

Statistical Analysis

Descriptive statistics were expressed as mean ± one standard deviation. Linear correlation (Pearson’s correlation coefficient) was used to examine the relationship between the reduced volume and large volume methods, while multivariate linear regression (SPSS; IBM, Armonk, NY, USA) was used to establish the relationship of MFI with respect to Ht and create the reduced volume MMFI. One-way analysis of variance was applied to test MMFI significance between plasma protein concentrations, and a P value <0.05 was considered statistically significant. Post-hoc testing was conducted using the Tukey method.

Results

Comparison of Rocker Bead Methods

While the new reduced volume RBT displayed lower MFI values relative to the older large volume method at each bovine blood matched Ht, a very strong linear relationship was observed between the two experimental protocols (r = 0.941) as shown in Figure 1B.

The Modified Mechanical Fragility Index (MMFI)

A strong nonlinear relationship was observed between sample Ht of the reduced volume RBT and the resultant MFI using human blood as shown in Figure 2A. Linear regression analysis on the MFI residuals (difference in MFI from the paired MFI value at 40% Ht) as a function of Ht deviation multiplied by the test Ht yielded a highly correlated fit (r = 0.965) at a forced y-intercept of zero (Figure 2B). The coefficient of determination predicting MFI from the experimental value at 40% Ht using the generated correction factor was R² = 0.941. Defining MMFI as the MFI if run at a standard Ht of 40%, the MMFI equation for the reduced volume RBT became:

$$\mathit{MMFI}=\mathit{MFI}-\frac{Ht(Ht-40)}{3266}$$ (2)

where 3266 is the fitted correction constant derived from the slope of the linear regression analysis in Figure 2B.

Figure 2.

(A) MFI values at hematocrit concentrations between 25 and 50% and (B) the resultant regression analysis of the MFI residuals after linearization using the reduced volume RBT with human blood.

The calculated MMFI for each experiment is shown in Figure 3 corresponding well to the MFI measured at 40% Ht for each unit of human blood (R² = 0.923).

Figure 3.

Comparison of the reduced volume MMFI to the MFI (mean ± 1.0 SD) across four different human blood units at hematocrits between 25 and 50%.

MMFI Reduced Volume Validation: Plasma Protein Protective Effect

There was a clear inverse linear relationship between total plasma protein concentration and RBC MF using the reduced volume RBT with human blood, with a statistically significant difference between 8.0 and 5.0 g/dL (P < 0.01) (Figure 4). There was a minor change in plasma viscosity associated with dilution, resulting in values of 1.74, 1.56, and 1.42 cP for the 8, 6.5, and 5.0 g/dL solutions, respectively.

Figure 4.

Effect of total plasma protein on red blood cell modified mechanical fragility index values as measured by the reduced volume RBT at 40% hematocrit using human blood. ( * - P < 0.01)

Discussion

The previously utilized large volume RBT of RBC MF, while an inexpensive, easily employable, and precise test, is limited in its scope of possible applications due to large blood sample volume requirements. For example, while it has been used in assessing blood quality during in vitro biocompatibility testing of blood pumps, researchers have been previously incapable of measuring the level of sublethal blood damage occurring in patients on mechanical circulatory support. Large sample volume requirement has also made it impossible to assess preoperative baseline RBC MF as a potential biomarker indicative of degree of hemolysis or anemia following left ventricular assist device (LVAD) implantation (12). Should this prove a relevant clinical biomarker, the new test may predict patient outcomes prior to LVAD implantation.

The same volume requirements that precluded the use of the original MFI on patients formed the basis for using bovine blood for comparison and correlation of the large volume and reduced volume RBT methods. Since the objective of these studies was to create a clinically acceptable and Ht-independent rocker bead MF measurement, the subsequent reduced volume experiments to develop the MMFI were performed using human blood.

As a function of MFI and a correction factor, MMFI remains dimensionless while retaining the familiar magnitude of scale. The same reasoning and principles applied with the selection of 40% Ht as the ‘standard’. While the Ht range examined (25–50%) represents typical clinical concentrations, MMFI values from reduced volume RBTs performed outside of this range may be less representative of RBC MF.

While previously published data investigated the effect of total plasma protein concentrations on RBC MF at 50 and 0% relative to baseline whole blood, we investigated concentrations of ~80 and ~60% relative to baseline whole blood (1). A similar inversely proportional relationship was found in this study with a statistically significant difference in RBC MF at ~60% from baseline.

The new reduced volume RBT has some limitations. The large volume RBT has been previously shown to have excellent intralaboratory reproducibility but poor interlaboratory reproducibility (5). While the phenomenon of interlaboratory reproducibility was not investigated in our development of the reduced volume method, we hypothesize that it will be similar. There are a number of factors such as minor variations in rocker speed, rocker amplitude angle, and blood collection method that can influence results when performed in different laboratories. This does not decrease the efficacy or meaningfulness of our test because in the context of a single-center clinical or preclinical study, it is the relative differences in MMFI values between samples, not the absolute magnitude of the MMFI values, which is significant.

The use of K2-EDTA, a dry-form calcium chelator, for these studies was to minimize hemodilution and possible plasma protein alternations. Since anticoagulation may affect RBC MF through plasma protein dilution, activity, and other direct or indirect mechanisms of actions, one anticoagulant should be used for all samples in a given study.

This new test was created to be of use in single-center studies involving patients and/or small animal experiments. The possible applications other than the above-mentioned mechanical circulatory support patient studies are plentiful; examples include assessing mechanical RBC damage in patients supported by ECMO or dialysis, testing the effects of pharmaceutical treatments on blood diseases or blood bank stored donor RBC transfusion, and assessing the viability of modified RBCs acting as intravascular drug delivery vehicles (13).

Conclusion

We have created a reduced volume rocker bead test of red blood cell mechanical fragility, and associated modified mechanical fragility index, that is easily employable, inexpensive, and requires a relatively small volume of blood. This method can be utilized in clinical and preclinical research studies as well as a part of diagnostics and/or treatment of patients.