Red blood cell salvage analysis from clotted blood

Abstract

Background

Blood clots discovered within body cavities intra-operatively are often manually broken up and placed in an autotransfusion device to recover autologous blood cells. This study evaluated the efficiency at which these red blood cells can be recovered from clot and to determine if these cells would be free of fibrin and clumping which might pose a risk of micro-emboli.

Materials and methods

Whole blood was aliquoted into 25 mL volume samples. The blood was then allowed to clot, and after 24 hours the clotted blood was manually kneaded by hand for 1, 2, 3, or 5 minutes. One mL of the harvested blood was fixed and processed for scanning electron microscope imaging. Plasma from the rest of the sample was then separated and underwent spectrophotometry for analysis of relative free haemoglobin.

Results

Blood recovered from the clotted blood ranged from 60 to 80% as time increased from 1 to 5 minutes of kneading. Volume of erythrocytes recovered from 1 minute compared to 2 minutes was statistically significant but not significant between 2 minutes or any longer period of time. Imaging did not show any evidence of fibrin strands or significant cell fragmentation. Spectrophotometry showed a steady increase of observed absorption at 540 nm, indicative of free haemoglobin, as manual kneading time increased.

Discussion

Red blood cells were able to be efficiently recovered from clotted blood. Imaging studies did not show any evidence of red blood cells trapped within fibrin mesh.

Introduction

In many emergent surgical procedures where the patient is haemorrhaging, significant amounts of blood clot can be encountered upon entering a body cavity. Clotted blood is frequently encountered during thoracic and abdominal trauma surgeries, ectopic pregnancy, and many situations where the patient must return to the operating room^1–4. Many autotransfusionists advocate for this clot to be manually dissected and processed by an autotransfusion system to recover transfusable autologous erythrocytes. However, little validation of this practice has been performed. Valeri et al.⁵ have demonstrated that baboon erythrocytes that were stored for up to 72 hours in a clot could be recovered when the clot was broken up by manual kneading. After autotransfusion system processing, these erythrocytes retained 2,3 DPG concentrations that were appropriate for their length of storage, i.e. the longer the storage time, the lower the intracellular 2,3 DPG concentration, as expected. Survival of the erythrocytes that were recovered from clots appeared to be similar to the lifespan of erythrocytes that were stored in the blood bank. Thus, it appears feasible to recover and reinfuse erythrocytes from clots using an autotransfusion system. The aim of this study was to assess the efficiency of the capture and return of erythrocytes derived from clots. We also attempted to determine if erythrocytes, once freed from a clot, were clumped into groups of erythrocytes bound by fibrin, which might act as micro-emboli upon reinfusion.

Materials and methods

Blood clot formation

Human whole blood anticoagulated with sodium citrate was purchased from Lampire Biological Laboratories (Pipersville, PA, USA) or acquired from the Institute for Transfusion Medicine (Pittsburgh, PA, USA). The whole blood acquired from the latter source was anticoagulated with citrate phosphate dextrose (CPD) solution. Consent for use of the blood for experimental purposes was given by the donor at the time of donation.

From each of the whole blood units, a baseline sample of 0.5 mL was drawn immediately from the erythrocyte unit using an aseptic technique and was fixed for imaging with a scanning electron microscope. Each unit of whole blood was divided into four groups containing 125 mL; these units were further subdivided into five 25 mL aliquots using an aseptic technique. From each 25 mL aliquot, the baseline haemoglobin concentration was measured using a Hemocue Hb 201+ Analyzer (HemoCue America, Brea, CA, USA). One sample from each group acted as the control (Figure 1). To the other four samples, 1.25 mL of 10% calcium chloride (CaCl₂) was added to achieve an approximate concentration of 10 mM CaCl₂. The CaCl₂ was added to reverse the anticoagulation and to stimulate clot formation. The control and clotted samples were then stored for 24 hours at room temperature.

Figure 1.

Showing how each whole blood unit was fractionated. The whole blood unit was broken into four 125 mL parts. From each 125-mL fraction, further fractionation occurred into five 25 mL parts.

min: minutes.

After 24 hours, the samples were visualized by eye to ensure that clotting had occurred. The control, unclotted samples were diluted with 50 mL of normal saline or lactated ringer’s solution. The clotted samples were emptied into a basin containing 50 mL of normal saline or Ringer’s lactate solution. The samples were evenly divided between the two solutions. Each of the four clotted samples were then kneaded by hand for 1, 2, 3, or 5 minutes. The kneading process consisted of the clot being gently squeezed between the palms of gloved hands allowing the freed blood to collect in a plastic basin underneath. Any smaller clots that fell into the collection basin were then recovered and kneaded again for the allotted time. A plastic Pasteur pipette was then used to collect the disintegrated clot into a separate container. The volume and haemoglobin concentration of each sample was measured. The extent to which residual clot was still present post kneading was determined by subtracting the total haemoglobin concentration in the unclotted control sample from the plasma haemoglobin concentration in the processed sample. The percentage of volume recovered from each sample was calculated using the following equation:

$$\begin{array}{c}\text{Percentage\hspace{0.17em}volume\hspace{0.17em}recovered}=(\text{Volume\hspace{0.17em}of\hspace{0.17em}recovered}\\ \text{blood}/\text{Volume\hspace{0.17em}of\hspace{0.17em}sample\hspace{0.17em}pre-clot})\times 100\%\end{array}$$

Measurement of relative free haemoglobin

After the volume and haemoglobin concentration of both the post-kneading clotted samples and the control samples were measured, 3 mL of each sample was placed into a test tube and centrifuged at 3,600 rpm for 20 minutes at room temperature. The supernatant was then transferred to a spectrophotometer cuvette (1.5 mL semi-micro UV methacrylate cuvette, Fisher Scientific, Nazareth, PA, USA) and the plasma free haemoglobin was determined by measuring the light absorbance at 540 nm (Spectronic Genesys 5 Spectrophotometer, Spectronic Instruments, Inc., Columbus, OH, USA). The absorption measurement was used to calculate the relative free haemoglobin concentration between each sample within each kneading group using the following calculation:

$$\begin{array}{c}\text{Relative\hspace{0.17em}free\hspace{0.17em}Hb\hspace{0.17em}concentration}=(\text{Absorbance\hspace{0.17em}of\hspace{0.17em}clotted}\\ \text{sample}-\text{Absorbance\hspace{0.17em}of\hspace{0.17em}control\hspace{0.17em}sample})/\text{Absorbance\hspace{0.17em}of}\\ \text{control\hspace{0.17em}sample.}\end{array}$$

Evaluation of erythrocyte morphology by electron microscopy

To assess the morphology of the erythrocytes in both the control and clotted samples by electron microscopy, 0.5 mL of each sample was fixed using phosphate buffered 2.5% glutaraldehyde for 1 hour and subsequently washed three times in 0.1 M phosphate buffered solution. The samples were then immersed in 1% osmium tetroxide for 1 hour followed by an additional three washings with phosphate buffered solution and dehydrated in graded ethanol (30%-50%-70%-90%-100%). Hexamethyldisilazane was added to each sample and then each sample was loaded onto glass slides and allowed to air dry overnight. The glass slides were placed in a vacuum and coated with a gold-palladium layer for scanning electron microscope analysis. A JSM 6335F scanning electron microscope (JEOL, Peabody, MA, USA) was used for evaluation of the red blood cell (RBC) morphology.

Statistical analysis

The control and the kneaded sample times were compared using a one-way analysis of variance. A Tukey’s multiple comparisons test was used between groups. p<0.05 was considered statistically significant.

Results

Two 500 mL units of whole blood were analysed. Pre-clot haemoglobin concentrations were all similar (p=0.96). Figure 2 shows the percentage of the original 25 mL aliquot which was recovered from each sample in each group. As kneading time increased, so too did the volume of recovered erythrocytes, but it is evident that there is a significant degree of diminishing returns after 1 minute of manually kneading the clot. Significant volume recovery maximised at 2 minutes of kneading.

Figure 2.

The volume of the original 25-mL aliquot of blood that was recovered over time of kneading. No difference in blood recovery was seen after 2 minutes of manipulation.

Figure 3 demonstrates that the relative free haemoglobin concentration generally increased as the clot kneading time increased for samples suspended in both normal saline and ringer’s lactate. No difference was seen in recovery volume and free haemoglobin between crystalloid solutions, so all data were pooled.

Figure 3.

Absorption at 540 nm which reflects the amount of free haemoglobin or damage to the erythrocytes which resulted from the manual kneading.

Scanning electron microscopy analysis of erythrocyte morphology was conducted on the control and clotted samples. Extensive cellular crenellation was seen, likely due to the storage age of the whole blood used in these experiments. Figure 4A is a scanning electron microscope image from 400× magnification of a control sample, while Figure 4B is a scanning electron microscope image from 950× magnification of a sample that underwent 5 minutes of kneading. The image shows a collection of normal appearing blood cells without any evidence of significant cell damage, as well as some debris of unknown aetiology. Other images did not demonstrate a significant number of erythrocytes trapped in fibrin mesh.

Figure 4.

Electron micrographs showing a typical view of the erythrocytes following clotting and manual kneading of the clot in saline.

(A) Control sample that did not have the CDP reversed and was not allowed to clot. (B) Sample that underwent 5 minutes of kneading time.

Figure 5 shows a ratio between the cellular debris and normal cells across kneading times. No significant differences were seen in fragmentation across all groups (p=0.09).

Figure 5.

Amount of cellular debris that resulted from the manual kneading. No statistical significance was seen but this may be the result of a small sample size.

Discussion

This series of experiments demonstrated that blood clots can be a viable source of erythrocytes that would increase the efficiency of erythrocyte recovery during cell salvage. Mathematical models have previously shown that even minor increases in cell salvage efficiency can have a significant impact on the blood loss that a patient can sustain before requiring an allogeneic transfusion^6,7. Thus, using clotted blood to augment the salvaged liquid blood might help to increase the efficiency of blood return. It was found that manual kneading of these clots maximised erythrocyte return at approximately two minutes. During our experiments, the kneading process was performed by the same individual to reduce any variability in RBC recovery. The kneading process would ideally consist of minimal force to release the blood from the clot for approximately two minutes. However, there will be an element of variability between technicians including force, rate, and technique, which would subsequently alter RBC recovery and mechanically induced haemolysis.

To assess the safety of this practice, electron microscopy was used to assess whether the RBC returned would present a risk of micro emboli. No evidence across multiple micrographs showed that this would be a potential risk. In 1919, a simple method for the defibrination of blood was published⁸. In this study, a flask plus vigorous shaking was used to defibrinate blood. Given that manual agitation probably replicates this shaking, it is not surprising that the erythrocytes were readily freed.

In addition, an assessment was made of how much erythrocyte destruction would occur during the manual kneading of the clot. Increasing degrees of free haemoglobin resulted from increasing kneading time; however, the cell salvage washing process should easily reduce free haemoglobin levels⁹. Cellular fragments were also assessed via electron microscopy and there were no differences in number of fragments seen across kneading time. However, further studies to assess the biochemical changes and properties that RBCs undergo after this manipulation would be beneficial to assess the optimum kneading time. It would be advantageous to analyse the function of RBCs after they have been recovered from clots. Hamasaki et al.¹⁰ and Valeri et al.⁵ both suggest and use 2,3 bisphosphoglyceric acid (2,3 DPG) as a correlate for RBC oxygen transport function. A similar model may be beneficial in future experiments to ensure that the overall function of RBCs is not significantly impaired considering the fact that the recovered RBC may have undergone molecular damage that was not visible via electron microscopy.

This study had several limitations. As the blood used was purchased from a commercial source, the blood was collected six days prior to the time that we had received it, whereas blood salvaged from clots in a patient would likely be much fresher. When imaging this blood, it was noted that a significant amount of the erythrocytes, even non-treated baseline cells, had already undergone significant morphological changes, such as crenellation, which can be seen early in the storage process¹¹. Thus, the ability to analyse changes in erythrocyte shape with increasing kneading times was limited; however, it was evident that there was not a significant amount of schistocytes or other evidence of cell damage compared to the control erythrocytes caused by the kneading process. The debris of unknown aetiology that was observed in the electron microscopic analysis was not increased in the kneaded samples compared to control samples, and it would have been removed through the autotransfusion system’s processing before reinfusion to the patient. Furthermore, it is also important to note that this was an in vitro recreation of clotted blood and that the blood was stored at room temperature for 24 hours. It is unclear as to whether the change in temperature could influence cell structure and potentially make it more susceptible to RBC membrane breakdown.

Conclusion

These experiments demonstrate that intact erythrocytes can be collected from clotted blood. A more detailed biochemical analysis of the collected erythrocytes should be performed prior to their reinfusion. Further study could assess whether the circulating time of these cells is equivalent to a non-clotted erythrocyte.