WO2021118375A1

WO2021118375A1 - Non–invasive and label–free method for evaluation of biochemical state of packed red blood cells

Info

Publication number: WO2021118375A1
Application number: PCT/PL2020/050093
Authority: WO
Inventors: Katarzyna MARZEC; Jakub DYBAS; Aleksandra WAJDA; Katarzyna BULAT; Magdalena KACZMARSKA; Marek GROSICKI
Original assignee: Uniwersytet Jagiellonski
Priority date: 2019-12-12
Filing date: 2020-12-04
Publication date: 2021-06-17
Also published as: PL432173A1

Abstract

A non–invasive and label–free method for evaluating the quality of packed red blood cells (pRBCs), where the pRBCs is largely red blood cells (RBCs) and a supernatant mixture (SM), along with trace amounts of other blood cell components including leukocytes and blood plasma. The measurement comprises at least one Raman spectrum of the pRBCs and a subsequent chemometric analysis. In such a way, at least one Raman spectrum is collected from the SM and a chemometric analysis of the collected spectrum is carried out for at least one ratio of the integral intensities of the chosen Raman marker bands within this spectrum. Subsequently, the result that is obtained for at least one ratio of the integral intensities is compared with a defined reference value, that is, either the reference value of the pRBCs' expiration (RV-Exp) or restricted reference value (RV-Res) of the pRBCs' quality, to assess the utility of the blood for transfusion to a chosen group of patients.

Description

Non-invasive and label-free method for evaluation of biochemical state of packed red blood cells

This invention describes a non-invasive and label-free method for evaluating the quality of packed red blood cells (pRBCs), particularly pRBCs intended for blood transfusion. The method is based on a measurement of the vibrational spectra of a pRBC supernatant mixture (SM) using Raman spectroscopy (RS) and conducting a chemometric analysis of its chosen marker bands, which are compared with the corresponding reference values.

The composition of pRBCs is largely erythrocytes, or red blood cells (RBCs), along with variable amounts of other blood cell components that include leukocytes and blood plasma, along with the SM, the composition of which is described on the manufacturer’s label of a pRBC storage container. There are various pRBC types, including leukoreduced pRBCs (LRBCs) and washed or irradiated pRBCs. The pRBC storage container is a bag made from polyvinyl chloride (PVC) foil that is suitable for erythrocyte storage. In these PVC foil bags, pRBCs are stored for up to 42 days at a temperature in the range of +2°C to +6°C, in most cases. During this time, the erythrocytes undergo biochemical alterations, the intensity and rate of which depend on variables including the age, sex, or physical condition of the blood donor. The biochemical alterations that occur during the erythrocytes’ storage may have negative impacts on the blood transfer efficiency, potentially causing complications for the blood recipient (Orlov et al., 2015). Before a transfusion, blood tests are carried out to ensure the compatibility of the blood group of the donor with that of the intended recipient, for serological control of the ABO system and the D antigen of the Rh system, to test the blood count, the hematocrit, and the hemoglobin (Hb) level, as well as to check for the presence of the markers of infectious agents that may be transmitted by the blood. The latter comprises tests for the presence of the HBs antigen, HCV antibody, HIV-1/2 antibody, and syphilis, as well as elevated A1AT activity; further tests also detect HBV, HCV, and HIV genetic material (Michalewska et al., 2009). A macroscopic evaluation of the erythrocyte hemolysis level is also performed based on the blood smear. During the pRBCs’ storage, the erythrocytes undergo hemolysis, which cannot exceed 0.8% (i.e., 0.8% hemolyzed RBCs compared to the total weight of the RBCs). Nonetheless, there is a lack of an analytical method for evaluating the RBCs have not undergone hemolysis, though their impairment can lead to a deterioration in the expected therapeutic effect. The application of a quick analytical method that does not require labeling or qualified staff to determine the pRBCs’ quality is, therefore, highly desirable. Despite it being commonly known that biochemical RBCs deteriorate during their storage, there is currently no method allowing for the determination of the pRBCs’ impairment before a transfusion (Jordan et al., 2016; Tzounakas et al., 2016). The advent of a quick and repeatable diagnostic method for assessing the biochemical state of pRBCs could contribute significantly towards minimizing the complications that can occur after a transfusion and improving the transfusion efficiency. Moreover, such a method could be useful for validating newly-designed supernatant mixtures. The biochemical alterations that occur inside RBCs are especially pronounced after the first week of their storage (Flatt et al, 2014). A decrease in glycolysis efficiency and adenosine triphosphate production can be observed and the pH drops, as well as the activities of the antioxidant enzymes that facilitate the oxidation of lipids and proteins. Consequently, these alterations evoke cytoskeleton damage and an ionic gradient imbalance that leads to changes in the rheology of the RBC membrane. This results not only in an elevated hemolysis level, but also a permanent impairment of the elasticity and deformability of the RBC. Moreover, due to the decreased efficiency of glycolysis, the amount of 2,3-bisphosphoglycerate (2,3 -DPG) drops and, thus, impairs the ability of the Hb for carrying out gas exchange. The next step of the RBC functionality deterioration is an elevation in the level of methemoglobin (metHb), which is the adduct unable to bind oxygen. Finally, Hb, free heme, and iron ions are released to the extracellular environment, greatly contributing to an elevated level of oxidative stress (Orlov et al., 2015). Together, the above- mentioned biochemical alterations decrease the functionality and viability of an RBC, which is not without significance for blood transfusions.

One of the best-characterized RBC metabolites is lactic acid, the product of glucose conversion during the glycolysis pathway, which is released to the extracellular environment. A determination of increasing lactic acid concentration, which leads to a deterioration in the efficiency of glycolysis, is a well-established marker of RBC metabolism (Flatt etak, 2014; Pagliaetal., 2016). The Hb, free heme, and iron ions released to the extracellular environment may, in turn, contribute to the oxidation of membrane lipids and proteins and their release from the cell, elevating their levels in the SM. Moreover, the auto-oxidation of Hb and free heme leads to an accumulation of metHb, hemichrome, and hemin, thus, greatly increasing their contribution towards the oxidative damage of RBCs. During hemolysis, the above-mentioned forms of Hb are released to the SM. RS is a non-invasive, label-free technique that can be utilized for matter studies. Thanks to technological developments and the emergence of modem lasers and sensitive detectors, RS can now be successfully applied to the study of complicated biological samples. There are significant advantages to this technique, such as its non-destructive effect on the studied material, the possibility of measurements in solutions, and the opportunity for solutions themselves to allow for analysis without the need for prior preparation or sample labeling (Dybas et al. , 2016) . Moreover, as has previously been shown, this technique can be successfully applied in the study of the erythrocytes in pRBCs.

In the field, there are known works — of, for instance, Gautam, Buckley, Vardaki, and Atkins (Gautam et al., 2018; Buckley et al., 2016; Vardaki et al., 2018; Atkins et al., 2017) — that describe studies on whole pRBCs (mixed fractions of the SM and RBCs), in which the quality of the pRBC is monitored based only on variations in the concentrations of various forms of Hb, as in such measurement conditions the RS spectra are dominated by the Hb signal. In the above-mentioned works, the authors use only the bands originating from the Hb for the chemometrics. They highlight how there is no possibility of differentiating between the RBC-encapsulated Hb and the free Hb related to the hemolysis. These works indicate processes related to the Hb oxidation that occurs due to the passing of gas through the PVC bag. The innovative approach of the herein presented patent application, therefore, in regard to the above-mentioned works, adds a measurement of the SM alone based on the collection of Raman spectra. This method allows for the analysis not only of the bands connected with Hb, but also, more importantly, the bands connected with various RBC metabolites that are released to the SM (i.e., lactates, glucose, proteins, and lipids). Moreover, the method presented herein comprises spectra measurements in the broader spectral range, which are analyzed with the use of the integral intensity ratios of the given bands. The analysis of the SM Raman spectra allows for a determination of the Hb amount that originates from the hemolysis itself and eliminates the impact on the obtained results of the gas that is passed through the PVC bag.

Atkins et al. (2016) have presented study results on the SM alone, which allowed for the monitoring of changes, e.g., in the amounts of lactates. Based on US patent no. US2017219568A, there is a known method and device for the quality control of blood-related products, such as pRBCs, platelets concentrate, granulocytes concentrate, leukocytes concentrate, whole blood, or blood plasma. The method that is presented therein comprises the collection of the Raman spectrum of blood-related products and the evaluation of whether or not this product can be employed for blood transfusion. The analysis of the obtained Raman spectrum allows for a quantification of the total number of altered RBCs, along with the presence of bacteria or other contaminants in the blood-related product. The collection of the Raman spectrum is carried out by trapping at least one erythrocyte in an optic trap produced using an excitation beam in the Raman device. The Raman device is coupled with an evaluation device for the performance of a cluster analysis, particularly, a principal component analysis, to identity alterations to at least one cell type in the blood-related product. The invention proposes the application of the excitation line of a wavelength from 700 to 1064 nm and a spectral analysis in the range of up to 1650 cnr¹. The evaluation device identifies RBCs and their alterations based on the analysis of the spectral ranges: 1650-1600, 1350— 1250, 1180-1120, 1100-1050, 930-890, and 700-650 cnr¹, in which the bands’ assignments and meanings are not explained. For this invention, the sample could be a dried or freshly-acquired blood drop of the blood-related product. The invention is let down by its lack of a method for sample acquisition and an absence of definitions for the data treatment and marker bands’ analysis.

As a large number of patients worldwide are in need of transfusions of pRBCs we aim do desig an efficient, non-invasive and label-free technology providing information about the biochemical state of pRBCs, which may affect their quality. During our research we were able to unexpectedly developed such Raman spectroscopy based method for the pre-transfusion quality check of RBCs where the proper measurement, analysis and comparison with reference values allow for non-invasive and label-free method for evaluation of biochemical state of pRBCs.

During conducted here studies, a new non-invasive and label-free method was invented for pRBCs’ quality evaluation when intended for blood transfusion, based only on a measurement of the vibrational spectra of the pRBC SM, utilizing RS and a suitable chemometric analysis of the chosen marker bands. In such a way, the sample comprises the SM and not the whole blood-related product, as is the case in the US patent mentioned above. The intervention is focused on the analysis of the RBC metabolites that are formed due to RBC biochemical alterations and released to the extracellular environment. When following the approach presented in this patent application, there is no need to restrain the excitation line to a wavelength of 700 to 1064 nm (as we have reported and presented in Example 3 using the 488 nm excitation line). In the spectral analysis presented herein, a high wavenumber region (2800-3050 cm ¹) was also taken into consideration. For the first time, we have also stated the exact ranges of the marker bands that need to be analyzed. As well as that, we have defined the reference value of the pRBCs’ expiration (RV-Exp) and the restricted reference value of the pRBCs’ quality (RV-Res) for the ratios of the integral intensities of the chosen marker bands.

The herein presented patent application employs the below definitions:

‘Marker band’ is the band present on the Raman spectrum originating due to the vibrations of the functional groups that are characteristic of a given compound or group of compounds. The presented herein patent application gives the marker band position as a wavenumber range within which it occurs on the Raman spectra, as presented in the Examples. The given wavenumber ranges are used for calculations of the integral Raman intensity ratios, which are then correlated with the results obtained using reference techniques.

‘Packed red blood cells’ (pRBCs) are the blood-related product used in transfusion medicine comprising mainly erythrocytes and the remnants of other whole blood components. There are, for instance, variable amounts, depending on the centrifugation parameters, of leukocytes, platelets, and blood plasma. There is also the SM, with a composition that depends on the manufacturer. The term, pRBCs, includes all variations, including leukoreduced pRBCs (LRBCs) and washed or irradiated pRBCs.

‘Supernatant mixture’ (SM) is the mixture of preservatives and additive solutions, comprising various chemicals, which allows for the storage of blood and its components. In the SM, the preservatives are sterile and the apyrogenic solution prevents blood clotting. The SM is composed of a citrate -phosphate -dextrose (CPD) solution, where trisodium citrate is an anticoagulant, citrate acid is a preservative, glucose is a nutrient, and sodium dihydrogen phosphate is a buffer. The SM is enriched with a saline-adenine-glucose- mannitol (SAGM) solution that increases RBCs’ viability due to the presence of sodium chloride, adenine, glucose, and mannitol. The composition of the SM can be additionally enriched with other components.

‘Reference value of pRBCs’ expiration’ (RV-Exp) is the value for the ratios of the integral band intensities, as shown in the patent, above which the pRBCs are unsuitable for transfusion. The RV-Exp value is determined independently for each of the ratios of the integral band intensities presented in this application, based on a data analysis performed at week six of pRBC storage for 30 pRBCs collected from both female (N = 15) and male (N = 15) donors.

'Restricted reference value of pRBCs’ quality’ (RV-Res) is the value for the ratios of the integral band intensities, as shown in the patent, above which the pRBCs are unsuitable for a transfusion to a patient with an increased risk of post-transfusion complications. The RV-Res value is determined independently for each of the ratios of the integral band intensities presented in this application, based on data analysis carried out at week three of storage, at which time the greatest increase in hemolysis was observed based on the results of flow cytometry, for 30 pRBCs collected from both female (N = 15) and male (N = 15) donors.

This invention details a method for evaluating the quality of pRBCs, where the pRBCs are composed mainly of erythrocytes and the SM, possibly also containing trace amounts of other whole blood components, including leukocytes and variable amounts of blood plasma. The method comprises the collection of at least one Raman spectmm of the pRBCs; chemometric analysis is then carried out for at least one marker band. The latter part of the method is characterized by a mathematical operation, including machine learning algorithms and utilizing the marker band spectral ranges. The obtained result is compared with a defined reference value, either the RV-Exp or RV-Res of the pRBCs’ quality, which is defined based on the results of a weekly analysis of 30 pRBCs obtained from female (N=15) and male donors (N=15).

For the method presented herein, it is preferred if the Raman spectrum is collected from a separated SM fraction. It is particularly preferred if this is done through the pRBC storage bag using Spatially Offset Raman Spectroscopy (SORS), and/or a dried separated SM fraction placed on a calcium fluoride (CaF₂) slide taken from a pilot tube after centrifugation, and/or a dried separated SM fraction placed on a calcium fluoride (CaF₂) slide taken directly from the pRBC storage bag.

In addition, it is preferred if the Raman spectrum is collected with the use of a 785 nm excitation wavelength and an analysis of the 200-3050 cm ¹ spectral range and, additionally, with the use of a 488 nm excitation wavelength and an analysis above the 2800 cm ¹ spectral range.

It is also preferred, when the Raman spectmm is collected with the use of 785 nm excitation wavelength, if the chemometric analysis is conducted for at least one ratio of any two marker bands observed in the following spectral ranges: 472±3-578±3 cnr¹, 390±3-478±3 cnr¹, 830±3-870±3 cnr¹, 870±3-910±3 cnr¹, 1520±3-1695±3 cnr¹, and 2867±3-2964±3 cnr¹. Or, if the Raman spectmm is collected with the use of 488 nm excitation wavelength, for at least one ratio of any two marker bands to be observed in the 2800±3-2900±3 cm ¹ and 2900±3-3040±3 cm ¹ spectral ranges.

It is particularly preferred that the chemometric analysis is conducted for at least one ratio of any two marker bands observed in the following spectral ranges:

472± 3 -578± 3 c nr ¹ : 390± 3 -478± 3 c nr ¹ ( 785 nm excitation wavelength)

870±3-910±3cm ¹ : 830±3-870±3cnr ¹ (785 nm excitation wavelength)

1520±3-1695±3cm ¹ : 2867±3-2964±3cnr^l (785 nm excitation wavelength) 2800±3cm ¹-2900±3cm ¹ : 2900±3cm ¹-3040±3cm ¹ (488 nm excitation wavelength) or the reciprocals of these marker band ratios.

Preferably, the chemometric analysis of at least one ratio of the marker bands of the Raman spectmm is to be collected based on an evaluation of the marker band integral intensities, or on a chemometric analysis comprising a mathematical operation, including machine learning algorithms and utilizing the given marker band spectral ranges.

It is preferred that at least one of the integral intensity ratios of the marker bands of the Raman spectmm is compared with the relevant RV-Exp, the value for which are defined as follows: not higher than 2.0 for 472±3-578±3cnr¹ : 390±3-478±3cm not lower than 0.8 for 870±3-910±3cnr¹ : 830±3-870±3cm not higher than 6.0 for 1520±3-1695±3cnr¹ : 2867±3-2964±3cm ' not higher than 0.7 for 2800±3cnr¹-2900±3cnr¹ : 2900±3cnr¹-3040±3cnr¹.

It is preferred that at least one of the integral intensity ratios of the marker bands of the Raman spectmm is compared with the relevant RV-Res, the values for which are defined as follows: not higher than 1.5 for 472±3-578±3cnr¹ : 390±3-478±3cm not lower than 1.0 for 870±3-910±3cnr¹ : 830±3-870±3cm not higher than 3.1 for 1520±3-1695±3cnr¹ : 2867±3-2964±3cm not higher than 0.4 for 2800±3cm ¹-2900±3cm ¹ : 2900±3cm ¹-3040±3cm ¹

It is preferable for the SM used in the evaluation of the pRBCs’ quality to comprise glucose, mannitol, adenine, trisodium citrate, citric acid, and sodium dihydrogen phosphate.

The invention relates to a quick and label-free method for evaluating the degradation of pRBCs based on a measurement of the overall biochemical SM properties using RS, accompanied by a suitable chemometric analysis of the chosen marker bands. Contrary to the methodologies that are known in the field (Atkins et al., 2017; Buckley et ah, 2016), in the invention presented herein, there are three possible approaches to the measurement of the SM fraction: (1) when pRBCs are stored, before the storage bag with the pRBCs is mixed, the RBCs are sedimented and the SM is separated as the upper fraction, thus, the SM can be measured directly with SORS; (2) when an SM fraction is taken from the pilot tube after its centrifugation and a fresh or dried SM drop is placed on the measurement slide; and (3) when an SM fraction is withdrawn directly from the pRBC storage bag using a syringe after its centrifugation and a fresh or dried SM drop is placed on the measurement slide.

Thereby, the invention facilitates the acquisition of general spectral profdes for the substrates and RBC metabolites in the SM. It does not rely on obtaining results from measurements of the RBCs alone, as has been presented in previously published methodologies (Atkins et al.,2017; Buckley et al., 2016; Vardaki et al., 2018; Gautam et al., 2018). The new approach, according to the invention presented herein, allows not only for an analysis of the various Hb forms in the SM (oxyhemoglobin, deoxyhemoglobin, and methemoglobin) but, more importantly, it allows us to track the glucose concentration, lactic acid derivatives, and protein and lipid amounts at the same time; that is otherwise impossible with measurements of pRBCs or isolated RBCs due to the dominant Hb signal. In the PVC foil bag, the pRBCs are stored vertically, enabling the formation of an RBC sediment and separated upper fraction of the SM. In such conditions, a Raman measurement can be carried out directly through the PVC storage bag using SORS.

Otherwise, the SM sample can be acquired from the pRBC pilot tube, or directly from the pRBC storage bag using a syringe. In such cases, centrifugation of the acquired sample is necessary to separate the RBC fraction. After centrifugation, the RBC sediment is clearly separated from the upper fraction of the SM. The SM sample, with a volume of approximately 10-50 mΐ, can be placed on a microscope slide and either measured immediately or dried. It is advantageous for the SM sample to be dried and placed on a calcium fluoride (CaF₂) slide.

The sample measurement of the SM is carried out by analyzing the Raman signal from the spectra acquired, using a suitable Raman spectroscope. The application of SORS allows the collection of the Raman spectrum directly through the PVC storage bag. Otherwise, the measurement must be carried out directly on the SM sample (when the SM sample is placed on the slide, preferably a slide made from CaF₂). Raman measurements can be acquired in a single spectrum mode or in a mode of spectra collection along a line (that is, a line scan mode). Preferably, at least 5 single spectra are collected from various places within the SM sample with the acquisition time of 3 s and 10 as the number of accumulations. To obtain a better signal to noise ratio, and to increase the analysis’ sensitivity, it is advantageous to collect at least 10-20 single spectra from various places within the SM sample with the above-mentioned parameters (3 s, 10 accumulations) need to be collected. Depending on the Raman spectroscope (system) that is used, the measurement parameters can vary. After the measurements are taken, the collected spectra should be preprocessed, including the removal of cosmic rays and a vector normalization in the whole spectral range. Preferably, a baseline correction should be conducted. Preprocessing can be carried out using various software, e.g., WITec Software 5.0, OPUS 7.2, or OriginLab 2019.

As is shown in the Examples, the measurements of the SM carried out using RS highly correlate with the data acquired using reference quantification techniques, including changes in the concentrations of lactates, glucose, total iron, triglycerides, and cholesterol, as well as correlating with the hemolysis results acquired from flow cytometry. Therefore, it can be understood that RS allows for a simultaneous, label- free, semi-quantitative analysis of changes in the concentrations of lactates, glucose, total lipid fraction, and proteins in the SM. Moreover, it was demonstrated that these concentrations found in the SM from pRBCs correlate with the hemolysis results acquired from flow cytometry, indicating a correlation of the Raman results from the SM measurements with the RBC state in pRBCs.

An evaluation of the pRBCs’ quality, based on an overall SM biochemical profile, is carried out with the 785 nm excitation line for the following spectral ranges: 390±3-478±3 cnr¹, 472±3-578±3 cnr¹, 830±3-870±3 cm ¹, 870±3-910±3 cnr¹, 1520±3-1695±3 cnr¹, and 2867±3-2964±3 cnr¹, as well as with the 488 nm excitation line for the 2800±3-2900±3 cm ¹ and 2900±3-3040±3 cm ¹ spectral ranges.

The integral intensity (that is, the area under the band) reflects the number of functional groups that participate in the origin of a band; this is calculated for each of the marker band spectral ranges. The integral intensity is proportional to the concentration of the given compound and relates to the vibrations of its characteristic functional groups. The chemometric analysis of at least one ratio of such marker bands of the Raman spectrum, collected according to the method presented herein, is conducted based on an evaluation of the marker band integral intensities or a chemometric analysis comprising mathematical operations, including machine learning algorithms and utilizing the given marker band spectral ranges. In the end, at least one integral intensity ratio of the marker bands of the Raman spectrum is compared with the RV-Exp or the RV-Res.

The below-presented band ranges, integral intensity ratios, and RV-Exp or RV-Res are defined in accordance with their best correlations with the quantitative data acquired using the reference techniques.

Band assignments and vibrational analysis in the presented herein patent application

The proper band assignments of the chosen spectral ranges of the pRBCs’ spectra are made based on the measured Raman spectra of the reference compounds: adenine, mannitol, and glucose (which all are components of the SAGM additive solution), the SAGM additive solution utilized for the pRBCs’ preparation, and sodium lactate and hemoglobin (which are the main RBC metabolites). The Raman spectra of above-mentioned components and the SAGM additive solution are compared with the pRBCs’ spectra after one and eight weeks of storage to assure the correct band assignments in the Raman spectra, as presented in Figs. 1 (standard pRBCs) and 2 (LRBCs).

The Raman spectmm of the pRBCs/LRBCs after one week of storage is highly comparable with the SAGM spectrum, i.e., the SAGM components including glucose, mannitol, and adenine in approximate concentrations as found in SAGM. The Raman spectrum of the pRBCs/LRBCs after eight weeks of storage is comparable with the spectra of the SAGM components with almost the same concentrations of glucose, mannitol, and adenine, but including new components connected with the pRBCs/LRBCs’ aging, as well as RBC metabolites, including Hb and lactates.

The band assignments for each of the spectral ranges observed in the Raman spectra of the pRBCs and LRBCs after one and eight weeks of storage are presented below. The band assignments of the Raman spectra are based on measurements of the reference compounds with the use of 785 nm (a-h) or 488 nm (i) excitation wavelengths. The spectra are presented in Figs. 1 and 2. a) Spectral range <350 cm ¹ - after one week of pRBC/LRBC storage, the band located at 160 cm ¹ has a similar profile to the SAGM spectrum. After eight weeks of pRBC/LRBC storage, the integral intensity of this band has greatly increased in connection with the increase in the concentration of lactates and the amount of free Hb from RBC hemolysis. Therefore, an increase in the integral intensity of this band located in the spectral range 70-270 cm ¹ over the time of pRBC/LRBC storage is connected with increases in the lactates and free Hb concentrations, along with a negligible decrease in the adenine concentration (Sugita et ak, 1965). The integral intensity of this band over the time of pRBC/LRBC storage corresponds with an increase in the concentrations of both lactates and glucose. A band located at around 323 cm ¹ originates from the vibrations in calcium fluoride and, as presented in the Examples herein, is used as the substrate for the pRBCs/LRBCs and the reference compound measurements, and thus is not considered in the spectral analysis. b) Spectral range 390±3-478±3 cm ¹ - a band located at around 420 cm ¹ in the Raman spectra of the pRBCs/LRBCs after one week of storage corresponds mainly with the vibrations in the glucose, with a negligible influence of the mannitol. Over the time of pRBC/LRBC storage, the impact of the vibrations in the glucose decreases, while the vibrations connected with the lactates’ presence increase (band at around 425 cm ¹ originates from the rocking vibrations of the carboxylic ion, -COO ). The integral intensity of this band over the time of pRBC/LRBC storage corresponds with a decrease in the glucose concentration accompanied by an increase in the lactates’ concentration. c) Spectral range 472±3-578±3 cm ¹ - a band located at around 520 cm ¹ in the Raman spectra of the pRBCs/LRBCs after one week of storage is connected with the presence of glucose and mannitol. Over the time of pRBC/LRBC storage, the impact of the vibrations in the glucose decreases while the vibrations connected with the lactates’ presence increase (band at around 535 cm ¹ originates from the wagging vibrations of the carboxylate ion, -COO ). The integral intensity of this band over the time of pRBC/LRBC storage corresponds with a decrease in the glucose concentration accompanied by an increase in the lactates’ concentration, while the mannitol concentration remains constant. d) Spectral range 700±3-790±3 cm ¹ - the integral intensity of the band located in this range increases over the time of the pRBC/LRBC storage as it originates mainly from vibrations in Hb (750 cm ¹) and the lactates (775 cm ¹). e) Spectral range 830±3-910±3 cm ¹ - the broad band (with a high full width at half maximum) located at around 880 cm ¹ and observed in the pRBCs/LRBCs after one week of storage is a combination of two bands originating from two glucose modes (855 and 920 cnr¹) and a mannitol mode (intense band at 880 cnr¹). After eight weeks of pRBC/LRBC storage, this band also comprises a component from a mode located at around 855 cm ¹ connected with the lactates’ presence. This band originates from the stretching vibrations in the carboxylate ion and corresponds with the overall contribution of all of the lactic acid derivatives (Cassanas et al., 1991). The integral intensity of the band at about 880 cm ¹ over the time of pRBC/LRBC storage corresponds with a decrease in the glucose concentration accompanied by an increase in the lactates’ concentration, while the mannitol concentration remains constant. The integral intensity of the band within the 830-870 cm ¹ spectral range over the time of pRBC/LRBC storage corresponds with a decrease in the glucose concentration accompanied by an increase in the lactates’ concentration. The integral intensity of the band within the 870-910 cm ¹ spectral range corresponds mainly with a decrease in the glucose concentration. f) Spectral range 970±3-1500±3 cm ¹ - comprises modes originating from SAGM components; changes within this spectral range correspond mainly with a decrease in the glucose concentration accompanied by an increase in the lactates’ concentration. However, the correlations remain obscure due to bands overlapping. The increase in the integral intensity of a band located at around 1260 cm ¹ over the time of pRBC/LRBC storage corresponds with an increase in the glucose concentration. g) Spectral range 1520±3-1695±3 cnr¹ -the broad band (with a high full width at half maximum) located in this spectral range relates to the appearance of Hb in the SM over the time of pRBC/LRBC storage. The integral intensity of this band corresponds with an increase in the Hb concentration in the SM and, thus, the hemolysis level. h) Spectral range 2867±3-2964±3 cm ¹ - the modes observed in this spectral range originate from C-H stretching vibrations. These bands are present in the SAGM components, sodium lactate, and Hb, and are typical of all biological components (Movasaghi et al., 2007). The integral intensity of the bands in this spectral range corresponds with the overall content of all of the sample components. i) Spectral range 2800±3-3040±3 cm ¹ - the bands located within the 2800-2900 cm ¹ spectral range originate mainly from -CH₂ stretching vibrations and are characteristic of lipid-related compounds, whereas the bands located within the 2900-3040 cm ¹ spectral range originate mainly from -CH₃ stretching vibrations and are characteristic of protein-related compounds (Dybas et al., 2016; Movasaghi et al., 2007; Mo et al., 2009). The integral intensities of these bands are much higher on the Raman spectra obtained with a 488 nm excitation wavelength compared to 785 nm due to the lower quantum efficiency of the CCD detectors when the excitation wavelength approaches infrared (Krafft et al., 2016). Moreover, the integral intensities of Raman bands depend on the excitation wavelength (Czamara et al., 2015) and the integral intensities of lipid-related bands can be understood to increase with an increase in the excitation energy (i.e., an excitation wavelength decrease) (Jamieson et al., 2018). Therefore, to assess the lipids/proteins ratio, it is preferable to use either a 488 or 532 nm excitation wavelength. The integral intensities of the bands within the 2800-2900 cm ¹ spectral range over the time of pRBC/LRBC storage correspond mainly with the total lipid content. The integral intensities of the bands within the 2900-3040 cm ¹ spectral range correspond mainly with the total protein content. Spectral analysis of the chosen bands with the corresponding spectral ranges allowing for a pRBC quality evaluation based on an analysis of the SM

I. 785 nm excitation wavelength:

1) (472±3-578±3 cm-¹) : (390±3-478±3 cm⁴)

The ratio of the integral intensity of the 472-578 cm⁴ spectral range to the 390-478 cm⁴ spectral range corresponds with an increase in the concentration of lactates over the time of pRBC storage. This can be accounted for by the greater impact of the lactates’ vibration found in the 472-578 cm⁴ spectral range compared to the 390-478 cm⁴ spectral range (band located at around 535 cm⁴ is approximately three times more intense when compared to the band at around 425 cm⁴ in the sodium lactate reference compound). The change in this ratio is correlated with the lactates’ increase, as observed with reference techniques. The line correlation coefficient is 84% on average, which proves the above-mentioned analysis of the integral intensity ratio of the given spectral ranges.

The ratio of the integral intensity of the wagging vibrations of the carboxylate ion (band at about 535 cm⁴) to the rocking vibrations of the carboxylate ion (band at about 425 cm⁴) increases with an increase in the lactic acid esters compared to the lactic acid and lactate ion (Cassanas et al., 1991). The lactic acid esters increase over the time of pRBC storage, therefore indicating an impairment of the RBCs’ biochemical states.

A value for this ratio that is higher than 2.0 always relates to expired pRBCs. That value is defined based on an analysis of 30 pRBC samples (N=15 from female donors, N=15 from male donors). Such an RV-Exp is based on a calculation of the average ratios obtained at the 42^nd day of storage (value: 1.8 ± 0.2). Therefore, values for this ratio that are higher than 2.0 indicate expired pRBCs in a rapid and accurate way.

Different reference values can also be defined independently for the more restricted requirements of pRBCs’ quality, which could be applied, for example, to patients at a higher risk of complications after a transfusion. We define the RV-Res based on an average ratio observed in the third week of storage, as the greatest increase in apoptotic RBCs was observed between the third and fourth weeks of storage. The RV- Res is 1.5 (value: 1.4 ± 0.1).

2) (870±3-910±3 cm⁴) : (830±3-870±3cm-¹)

The ratio of the integral intensity of the 870-910 cm⁴ spectral range to the 830-870 cm⁴ spectral range corresponds mainly with a decrease in the glucose (G) concentration and the sum of the decrease in glucose (G) and increase in lactates (L) [G/(G+L)]. This ratio decreases over the time of pRBC storage and highly correlates with the biochemical data (R² > 0.92).

A value for this ratio that is lower than 1.0 always relates to expired pRBCs. That value is defined based on an analysis of 30 pRBC samples (N=15 from female donors, N=15 from male donors). Such an RV-Exp is based on a calculation of the average ratios obtained at the 42^nd day of storage (value: 0.9 ± 0.1). Therefore, the values for this ratio that are lower than 1.0 indicate expired pRBCs in a rapid and accurate way. Different reference values can also be defined independently for the more restricted requirements of pRBCs’ quality, which could be applied, for example, to patients at a higher risk of complications after a transfusion. We defined the RV-Res based on an average ratio observed in the third week of storage, as the greatest increase in apoptotic RBCs was observed between the third and fourth weeks of storage. The RV- Res is 1.0 (value: 1.1 ± 0.1).

3) (1520±3-1695±3cm^_1) : (2867±3-2964±3cm-¹)

The ratio of the integral intensity of the 1520-1695 cm ¹ spectral range to the 2867-2964 cm ¹ spectral range increases over the time of pRBC storage and corresponds with an increase in Hb concentration. This ratio is correlated with the concentration of free iron ions in the SM; the line correlation coefficient is 93% on average.

A value for this ratio that is higher than 6.0 always relates to expired pRBCs. That value is defined based on an analysis of 30 pRBC samples (N=15 from female donors, N=15 from male donors). Such an RV-Exp is based on a calculation of the average ratios obtained at the 42^nd day of storage (value: 4.4 ± 1.6). Therefore, the values for this ratio that are higher than 6.0 indicate expired pRBCs in a rapid and accurate way.

Different reference values can also be defined independently for the more restricted requirements of pRBCs’ quality, which could be applied, for example, to patients at a higher risk of complications after a transfusion. We defined the RV-Res based on an average ratio observed in the third week of storage, as the greatest increase in apoptotic RBCs was observed between the third and fourth weeks of storage. The RV- Res is 3.1 (value: 2.3 ± 0.8).

II. 488 nm excitation wavelength: (472±3-578±3 cm^-1) : (390±3-478±3 cm^-1)

The ratio of the integral intensity of the 2800-2900 cm ¹ spectral range to the 2900-3040 cm ¹ spectral range increases over the time of pRBC storage and corresponds with an increase in the lipids/proteins ratio. This ratio is correlated with the increase in lipids obtained with a biochemical analyzer; the line correlation coefficient is 89% on average.

A value for this ratio that is higher than 0.7 always relates to expired pRBCs. That value is defined based on an analysis of 30 pRBC samples (N=15 from female donors, N=15 from male donors). Such an RV-Exp is based on a calculation of the average ratios obtained at the 42^nd day of storage (value: 0.6 ± 0.1). Therefore, the values for this ratio that are higher than 0.7 indicate expired pRBCs in a rapid and accurate way.

Different reference values can also be defined independently for the more restricted requirements of pRBCs’ quality, which could be applied, for example, to patients at a higher risk of complications after a transfusion. We defined the RV-Res based on an average ratio observed in the third week of storage, as the greatest increase in apoptotic RBCs was observed between the third and fourth weeks of storage. The RV- Res is 0.4 (value: 0.42 ± 0.02).

The subject of the herein presented patent application is given in the below-mentioned non-limiting Examples and the Figures, which show: Fig. 1 Average Raman spectra of the SAGM components: adenine (Sigma-Aldrich), mannitol (Sigma- Aldrich), glucose (D-glucose, Merck), and SAGM (Maco Pharma), and the exemplary Raman spectrum of the LRBCs after one week of storage. The spectra are recorded with a 785 nm excitation wavelength, with the laser power at the laser spot of approximately 130 mW, and are presented after a baseline correction.

Fig. 2 Average Raman spectra of the SM components: Hb, glucose (D-glucose, Merck), and sodium lactate (Sigma-Aldrich), and the exemplary Raman spectra of the LRBCs after one and eight weeks of storage. The spectra are recorded with a 785 nm excitation wavelength, with the laser power at the laser spot of approximately 130 mW, and are presented after a baseline correction.

Fig. 3 Schematic diagram of the method of the herein presented patent application, with letter designations indicating the possible embodiments of the method: A) application of SORS - measurement of the SM fraction directly through the PVC storage bag; B) measurement of the SM fraction taken from the pRBC storage bag using a syringe and placed on a substrate; C) aspiration of the pRBCs from the pilot tubes placed on a substrate after the centrifugation and the measurement of the SM fraction.

Fig. 4 Progression of changes to the pRBCs (A-C) and LRBCs (D-F) during their 56 days of storage in the PVC storage bags at 4 °C. All presented biochemical data is collected using the ABX Pentra 400 Analyzer (Horiba Medical).

Fig. 5 Average Raman spectra of the given SM samples obtained from the pRBCs. The spectra are recorded with a 785 nm excitation wavelength, with a laser power at the laser spot of approximately 130 mW, and are presented after a baseline correction.

Fig. 6 Set of box charts of the integral intensity ratios of metabolites found in the given SM samples obtained from the pRBCs based on the baseline-corrected Raman spectra recorded with 785 nm excitation wavelength (the ratios: 472-578/390-478 cnr¹, 870-910/830-870 cm ¹ and 1520- 1695/2867-2964 cm ¹ were acquired using A-type integration method). Box ends represent the standard deviation, horizontal line in the middle of the box represents the mean value and whiskers represent min and max values. Horizontal dotted and dashed lines passing through box charts represent RV-Res and RV-Exp values, respectively.

Fig. 7 Average Raman spectra of the given SM samples obtained from the LRBCs. The spectra are recorded with a 785 nm excitation wavelength, with a laser power at the laser spot of approximately 130 mW, and are presented after a baseline correction.

Fig. 8 Set of box charts for the integral intensity ratios of the metabolites found in the given SM samples, as obtained from the LRBCs based on the baseline-corrected Raman spectra recorded with a 785 nm excitation wavelength (the ratios: 472-578/390-478 cnr¹, 870-910/830-870 cnr¹, and 1520- 1695/2867-2964 cm ¹ were acquired using the A-type integration method). The box ends represent the standard deviation, the horizontal line in the middle of the box represents the mean value, and the whiskers represent the minimum and maximum values. Horizontal dotted and dashed lines passing through box charts represent RV-Res and RV-Exp values, respectively. Fig. 9 Correlation of the data obtained using RS and a 785 nm excitation wavelength (the ratio of various lactic acid derivatives based on the integral intensities of the 472-578/390-478 cm ¹ spectral ranges) with the reference data obtained using the ABX Pentra 400 Analyzer (concentration of lactic acid derivatives in mmol/L) within six weeks.

Fig. 10 Correlation of the data obtained using RS and a 785 nm excitation wavelength (the ratio of glucose content to the sum of the glucose and lactic acid derivatives is based on the integral intensities of the 870-910/830-870 cm ¹ spectral ranges) with the reference data obtained using the ABX Pentra 400 Analyzer (the ratio of glucose concentration to the sum of the glucose and lactic acid derivatives) within six weeks.

Fig. 11 Correlation of the data obtained using RS and a 785 nm excitation wavelength (the ratio of the Hb content to the quantity of C-H bonds present in all sample components based on the integral intensities of the 1520-1695/2867-2964 cm ¹ spectral ranges) with the reference data obtained using the ABX Pentra 400 Analyzer (concentration of free iron ions in mmol/L) within six weeks.

Fig. 12 Flow cytometry analysis of the apoptosis of human RBCs in pRBCs intended for transfusion. The RBCs stored in the pRBCs are analyzed every week with flow cytometry to assess their apoptosis. The analysis requires staining with annexin-V labeled with fluorescein isothiocyanate.

Fig. 13 Average Raman spectra of the given SM samples obtained from the pRBCs. The spectra are recorded with a 488 nm excitation wavelength, with a laser power at the laser spot of approximately 3 mW, and are presented after a baseline correction.

Fig. 14 Set of box charts for the integral intensity ratios of metabolites found in the given SM samples obtained from the pRBCs based on the baseline-corrected Raman spectra recorded with a 488 nm excitation wavelength (the ratio 2800-2900/2900-3040 cm ¹ was acquired using the A-type integration method). The box ends represent the standard deviation, the horizontal line in the middle of the box represents the mean value, and the whiskers represent the minimum and maximum values. Horizontal dotted and dashed lines passing through box charts represent RV- Res and RV-Exp values, respectively.

Fig. 15 Correlation of the data obtained using RS and a 488 nm excitation wavelength (the ratio of lipids to proteins based on the integral intensities of the 2800-2900/2900-3040 cm ¹ spectral ranges) with the reference data obtained using the ABX Pentra 400 Analyzer (total lipid concentration in mmol/L) within six weeks.

Example 1. Evaluation of standard (non-leukoreduced) pRBC quality based on an SM measurement using RS

To evaluate the pRBCs’ quality, RS measurements were carried out every seven days over a period of 56 days on samples acquired from three male donors: donor A (age: 27, blood type: A Rh+), donor B (age: 27, blood type: B Rh+), and donor C (age: 28, blood type: 0 Rh+). The pRBCs were stored in standard PVC storage bags at 4 °C. The SM for all of the above-mentioned samples comprised a CPD anticoagulant- preservative solution (citrate-phosphate-dextrose) and an SAGM additive solution (saline-adenine-glucose- mannitol). The composition of the CPD was as follows: trisodium citrate (26.30 g/1), citric acid (3.27 g/1), glucose (25.50 g/1), and sodium dihydrogen phosphate (2.22 g/1), while the SAGM contained sodium chloride, adenine, mannitol, and glucose.

To correlate the data obtained from the SM measurements using RS with the absolute concentrations of every SM component, the samples were additionally measured using quantitative reference techniques. Simultaneously to the RS, SM biochemical parameters were assessed including total cholesterol, glucose, free iron ions, and the total content of lactic acid derivatives and triglycerides. Moreover, to correlate the observed SM alterations with the RBC quality in the pRBCs, the hemolysis level was evaluated using flow cytometry.

The pRBC samples were aspirated using a syringe directly through the PVC storage bag and then centrifuged in plastic tubes with an acceleration of 500 xg, a time of 10 min, and without braking at room temperature. The separated upper fraction SM was aspirated and divided into two - the first part was used for reference studies using the ABX Pentra 400 Analyzer (Horiba Medical) and the second part was used for RS measurements. In the latter, 50 mΐ of SM was transferred to a CaF₂ slide and left for 2 h until completely dry.

The RS measurements were carried out using a confocal Raman imaging system, the WITec Alpha 300 with air-objective, with a 100 magnification (Olympus, MPlan, NA = 0,9). The spectra were recorded with a 785 nm excitation wavelength and a laser power at the laser spot of approximately 130 mW. Ten spectra were recorded from randomly -chosen places within the SM sample with an acquisition time of 3 s and 10 as the number of accumulations. All measurements were carried out after one, two, three, four, five, six, and eight weeks of pRBC storage (while the term of validity was 42 days (6 weeks), the last measurement after eight weeks of pRBC storage was carried out after their validity date).

All collected Raman spectra were preprocessed, including the removal of cosmic rays (using WITec Software 5.0), vector normalization in the whole spectral range (400-3050 cnr¹, using OPUS 7.2 software), and additionally baseline correction (using OriginLab 2019 software). To assess the integral intensity ratios of the metabolites found in the SM samples, OPUS 7.2 software and the A-type integration method (Applied integration method was based on calculation of the integral value represented as the area bounded by the bandshape, abscissa and the wavenumbers limits defined as local minima of the given band) were used (suitable for integration of bands on baseline-corrected spectra). Analyses of the integral intensities of the bands found the following ranges were taken into consideration: 390-478, 472-578, 830-870, 870-910, 1520-1695, and 2867-2964 cnr¹. Box charts were constructed using OriginLab 2019 software to enable a graphical representation of the distribution of statistical features.

Subsequently, the data obtained from RS were correlated with the data obtained using the reference technique. In Fig. 4 (A, B, C), the reference data are presented, which were obtained using the ABX Pentra 400 Analyzer. According to the obtained information, over the time of pRBC storage, the lactates’ total content increased in a manner that was inversely proportional to the glucose concentration. There was an observable increase in the free iron ions’ concentration, indicating ongoing hemolysis. Moreover, increases in the cholesterol and triglycerides were observed, which were especially pronounced in the final weeks of the experiment. The above-mentioned parameters can be concluded as pRBC degradation markers. The data obtained using RS were correlated with the absolute concentrations of each metabolite assessed using the biochemical analyzer (Fig. 9 A, B, and C - Fig. 11 A, B, and C) over the six weeks of pRBC storage. The ratio of the various lactic acid derivatives assessed using RS (based on the integral intensities of the 472-578/390-478 cm ¹ spectral ranges) was correlated with the total concentration of the lactic acid derivatives assessed using the biochemical analyzer with a compliance of greater than 83%. A precise correlation of the RS data with the biochemical reference technique was also available for the ratio of glucose content to the sum of glucose and overall lactic acid derivatives, with a compliance of greater than 95%. Moreover, the increase in Hb concentration in the SM, obtained using RS, was correlated with the released iron ions found in the SM when assessed using the biochemical analyzer, with a compliance of greater than 91%. The line correlation coefficients differ between the donors (A, B, and C) in relation to the different rates of RBC deterioration. As previously mentioned, this depends on various factors including the age, lifestyle, and physical condition of the donor.

In the next step, the data obtained using RS were correlated with pRBC quality. The information obtained using RS (Fig. 6) indicated that, over the time of pRBC storage in the SM, there was an observable increase in the ratio of the integral intensity of the wagging vibrations to the rocking vibrations of the carboxylate ion (the ratio of integral intensities of 472-578/390-478 cm ¹ spectral ranges). The most pronounced change occurred between the fourth and fifth weeks of pRBC storage, indicating the most pronounced increase in lactic acid esters compared to free lactate ions (lactic acid). On the contrary, the biggest decrease in the ratio of the glucose content to the sum of the glucose and lactic acid derivatives (the ratio of integral intensities of 870-910/830-870 cm ¹ spectral ranges) was observed between the first and second weeks of pRBC storage. This ratio corresponds with a simultaneous decrease in the glucose and increase in the lactates in the SM. Moreover, based on the ratio of the integral intensities of the 1520- 1695/2867-2964 cm ¹ spectral ranges, an increase in Hb content was observed; this could be addressed to monitor the RBCs’ hemolysis progression.

Most of the data obtained with non-invasive and label-free SM measurements using RS highly correlated with the data obtained for the SM using quantitative reference techniques. This indicates that the biochemical composition of the SM can be evaluated using RS.

Alterations to the metabolite concentrations in the SM, observed both with RS and the reference technique, relate to RBC metabolism and aging. In the course of the metabolic pathway, RBCs transform glucose into lactic acid. The oxidation of proteins and peroxidation of lipids, which occurs during pRBC storage, leads to a decrease in the total lipid content in the RBC membrane and, thus, the release of RBCs into the SM. Consequently, Hb, free haem, and iron ions are released to the extracellular environment, as well, greatly contributing to an increase in oxidation stress and an impairment of the proper RBC function.

Flow cytometry studies were carried out to confirm a correlation of the metabolites’ alteration, as found in SM, with the pRBCs; quality deterioration; the results are presented in Fig. 12. The alterations to the glucose, lactates, iron ions, cholesterol, and triglycerides observed in SM were accompanied by a gradual deterioration in the RBCs’ viability. Increasing apoptosis (as the RBCs raptured) over the time of the RBCs storage is presented in Fig. 12 and highly correlates with the biochemical parameters found in the SM. The above-mentioned biochemical parameters monitored in the SM can be concluded to be the pRBCs’ degradation markers.

This Example clearly indicates that an analysis of the chosen ratio of the Raman marker bands of the SM vibrational spectra allows for a determination of the concentration of the chosen components and RBC metabolites found in the SM. The correlation of this data with the biochemical quantitative techniques proves the usefulness of RS for the analysis of the composition of the SM. The correlation of the biochemical, quantitative, and RS data with the hemolysis level assessed using flow cytometry proves the usefulness of pRBCs’ quality evaluation based on the chosen SM biochemical parameters.

Moreover, as is marked in Fig. 6, at each point of the pRBCs’ spectroscopic quality measurement, the chosen ratios of the integral intensities of the chosen marker bands can be compared to the specific reference values — defined here as the RV-Exp or the RV-Res of pRBCs’ quality — given for each of the integral intensities ratios presented here.

In conclusion, Example 1 clearly proves that an RS measurement of the SM, according to the method in the herein presented patent application, and an analysis of the chosen band spectral ranges (390-478. 472-578, 830-870, 870-910, 1520-1695, and 2867-2964 cnr¹), as well as a comparison of the values of the chosen ratios of the integral intensities of such bands to the RV-Exp or RV-Res of pRBCs’ quality, is a suitable method for evaluating the quality of pRBCs.

In this example, the procedure designated as C on the schematic diagram (Fig. 3) was applied. However, the procedures designated as A and B (Fig. 3) allowed identical results to be obtained. Identical results can also be acquired using different Raman spectrometers or using various Raman-related techniques (e.g., SORS, SERS, and others).

Example 2 Evaluation of leukoreduced pRBCs’ (LRBCs) quality based on measurements of the SM using RS

Similarly, as explained in Example 1, to evaluate LRBCs’ quality, RS measurements were carried out every seven days over 56 days on samples acquired from three male donors: donor D (age: 28, blood type: A Rh+), donor E (age: 26, blood type: 0 Rh+), and donor F (age: 29, blood type: 0 Rh+). The LRBCs were stored in standard PVC storage bags at 4 °C. The SM for all of the above-mentioned samples comprised a CPD anticoagulant-preservative solution (citrate-phosphate-dextrose) and an SAGM additive solution (saline-adenine-glucose-mannitol). The composition of the CPD was as follows: trisodium citrate (26.30 g/1), citric acid (3.27 g/1), glucose (25.50 g/1), and sodium dihydrogen phosphate (2.22 g/1), while the SAGM contained sodium chloride, adenine, mannitol, and glucose.

The acquisition and preparation of the SM samples from the LRBCs was carried out in the same way as explained in Example 1 for the pRBCs. Similarly, the RS measurements were carried out in the same manner as was explained in Example 1 using the same confocal Raman imaging system and the same measurement parameters. All measurements were carried out after one, two, three, four, five, six, seven, and eight weeks of LRBC storage (while the term of validity was 42 days (6 weeks), the last measurements after seven and eight weeks of LRBC storage were carried out after their validity date). The chemometrics and preprocessing of the Raman spectra obtained for the SM samples from the LRBCs were conducted in the same way as was explained in Example 1 for the SM samples from pRBCs.

Subsequently, the data obtained from RS were correlated with the data obtained using the reference technique. In Fig. 4 (D, E, and F), the reference data are presented that were obtained using the ABX Pentra 400 Analyzer. According to the obtained information, similar to that presented in Example 1, over the time of the LRBCs’ storage, the lactates’ total content increased in a manner that was inversely proportional to the glucose concentration. There was an observable increase in the free iron ions’ concentration, indicating ongoing hemolysis, which was confirmed by flow cytometry results (Fig. 12). In Example 2, increases in the cholesterol and triglycerides were observed, which were especially pronounced in the final weeks of the experiment. The integral intensity ratios of the spectral ranges 472-578/390-478 cnr¹, 870-910/830- 870 cnr¹, and 1520-1695/2867-2964 cm ¹ were correlated with the absolute concentrations of each metabolite assessed using the biochemical analyzer (Fig. 9 D, E, and F - Fig. 11 D, E, and F). The ratios of various lactic acid derivatives, assessed using RS, were correlated with the total concentration of the lactic acid derivatives, as assessed using the biochemical analyzer with a compliance of greater than 75% (over the six weeks of storage). On the contrary, the ratio of the glucose content was correlated to the sum of the glucose and lactic acid derivatives with a compliance of greater than 91%. The ratio of the integral intensities of the 1520-1695/2867-2964 cm ¹ spectral ranges was correlated with the free iron ions’ concentration found in the SM with a compliance of greater than 87%. The line correlation coefficients were noticeably lower for the SM samples from LRBCs, compared with those from the pRBCs. This could be related to the more pronounced hemolysis of the LRBCs in the final weeks of the experiment, resulting in Raman spectra with a lower signal to noise ratio; that could be improved through the collection of an additional number of spectra.

In the next step, the data obtained using RS were correlated with the LRBCs’ quality. The information obtained using RS is presented as box charts (Fig. 8), which indicate that over the time of the LRBCs’ storage in the SM there was an observed increase in the lactic acid esters compared to the free lactate ions (lactic acid), with the biggest change observed between third and fourth weeks of LRBC storage (the biggest increase in the ratio of the integral intensity was of the bands located at about 532/428 cnr¹). A decrease in the glucose content accompanied by an increase in the lactates’ concentration (based on the ratio of the integral intensities of the 870-910/830-870 cm ¹ spectral ranges) was characterized by a linear progression. On the contrary to the SM samples from the pRBCs presented in Example 1, the big drop between the first and second weeks of LRBC storage was not observed. A gradual increase in Hb content (based on the ratio of the integral intensities of the 1520-1695/2867-2964 cm ¹ spectral ranges) was observed over the whole time of the LRBCs’ storage. An analysis of the chosen ratio of the Raman marker bands of the SM vibrational spectra allowed for a determination of the concentration of the chosen components and RBC metabolites found in the SM. The correlation of this data with the biochemical quantitative techniques proves the usefulness of RS for analysis of the composition of the SM obtained from LRBCs. The correlation of the biochemical, quantitative, and RS data proves the usefulness of LRBCs’ quality evaluation, based on the chosen SM biochemical parameters. Moreover, as is marked in Fig. 8, at each point of the pRBCs’ spectroscopic quality measurement, the chosen ratios of the integral intensities of the chosen marker bands can be compared to the specific reference values — defined here as the RV-Exp or RV-Res of pRBCs quality — given for each of the integral intensities ratios presented here.

In conclusion, Example 2 clearly proves that RS measurement of the SM, according to the method in the herein presented patent application and an analysis of the chosen band spectral ranges (472-578/390- 478 cnr¹, 870-910/830-870 cnr¹, and 1520-1695/2867-2964 cnr¹), as well as a comparison with the values of the chosen ratios of the integral intensities of such bands to the RV-Exp or the RV-Res of the LRBCs; quality, is a suitable method for evaluating the quality of LRBCs.

In this example, the procedure designated as C in the schematic diagram (Fig. 3) was applied. However, the procedures designated as A and B (Fig. 3) allowed identical results to be obtained. Identical results can also be acquired using different Raman spectrometers or using various Raman-related techniques (e.g., SORS, SERS, and others).

Example 3 Evaluation of pRBCs’ quality based on measurements of the SM using RS and a 488 nm excitation wavelength

As explained in Example 1, to evaluate pRBCs’ quality, RS measurements were carried out every seven days over 56 days on samples acquired from three male donors: donor A (age: 27, blood type: A Rh+), donor B (age: 27, blood type: B Rh+), and donor C (age: 28, blood type: 0 Rh+). The pRBCs were stored in standard PVC storage bags at 4 °C. The SM for all of the above-mentioned samples comprised a CPD anticoagulant-preservative solution (citrate-phosphate-dextrose) and an SAGM additive solution (saline-adenine-glucose-mannitol). The composition of the CPD was as follows: trisodium citrate (26.30 g/1), citric acid (3.27 g/1), glucose (25.50 g/1), and sodium dihydrogen phosphate (2.22 g/1), while the SAGM contained sodium chloride, adenine, mannitol, and glucose.

The acquisition and preparation of the SM samples from pRBCs were carried out in the same ways as explained in Example 1. The RS measurements were carried out using a confocal Raman imaging system, WITec Alpha 300 with air-objective, with 100 ^c magnification (Olympus, MPlan, NA = 0,9). The spectra were recorded with a 488 nm excitation wavelength and a laser power at the laser spot of approximately 3 mW. Ten spectra were recorded from randomly -chosen places within the SM sample with the acquisition time of 3 s and 10 as the number of accumulations. All measurements were carried out after one, two, three, four, five, six, and eight weeks of pRBCs storage (while the term of validity was 42 days (6 weeks) the last measurement after eight weeks of pRBC storage was carried out after their validity date).

All of the collected Raman spectra were preprocessed, including the removal of cosmic rays (using WITec Software 5.0), vector normalization in the whole spectral range (400-3050 cnr¹, using OPUS 7.2 software), and additionally baseline correction (using OriginLab 2019 software), and are presented in Fig. 13. To assess the integral intensity ratios of the metabolites found in the SM samples, OPUS 7.2 software and the A-type integration method were used (suitable for the integration of bands on baseline-corrected spectra). The following integral intensity spectral ranges were taken into consideration during the analysis: 2800-2900 and 2900-3040 cm f Box charts were constructed using OriginLab 2019 software to enable a graphical representation of the distribution of statistical features.

Subsequently, the data obtained using RS were correlated with data obtained using the reference technique. The information obtained using RS is presented as box charts (Fig. 14), indicating that over the time of pRBC storage in the SM there was an observable increase in the lipids to proteins ratio with the biggest change observed from the fourth week of the pRBCs’ storage. The ratio of lipids to proteins was calculated using the integral intensities of the 2800-2900 and 2900-3040 cm ¹ spectral ranges, and was then correlated with the total lipid content assessed using the biochemical ABX Pentra 400 analyzer, where the total lipid content related to the sum of the cholesterol and triglyceride concentrations (expressed in mmol/L). The correlation, as presented in Fig. 15, proved that an increase in the lipids to proteins ratio could be observed as an increase in the integral intensities of the above-mentioned spectral ranges (Raman marker bands) on the Raman spectra collected with a 488 nm excitation wavelength.

Moreover, as is marked in Fig. 13, at each point of the pRBCs’ spectroscopic quality measurement, the chosen ratios of the integral intensities of the chosen marker bands can be compared to the specific reference values — defined here as the RV-Exp or RV-Res of pRBCs quality — given for each of the integral intensities ratios presented here.

In conclusion, Example 3 clearly proves that RS measurement of the SM, according to the methods in the herein presented patent application and an analysis of the chosen band spectral ranges (2800-2900 and 2900-3040 cnr¹), as well as a comparison of the values of the integral intensities ratio of such bands to the RV-Exp or RV-Res of pRBCs’ quality, is a suitable method for evaluating the quality of pRBCs.

Example 4 Definition and exclusion from transfusion of the pRBCs that do not meet the corresponding reference values based on examples 1-3

Our procedure allows for a definition of the pRBCs that do not fulfill the RV-Exp or RV-Res of pRBC quality. Different approaches to comparing the values of ratios measured for specific pRBCs with RV-Exp and RV-Res can be applied, depending on the applied laser excitation (for example, 788 nm in Examples 1 and 2 or 488 nm in Example 3) and the number of ratios taken into consideration in the analysis.

The content of the SM at the specific time of pRBC analysis depends on the donor profile. For example, in some pRBCs, we observed high hemolyzes by the third week of storage, though the lactates’ levels at the same time were very low. Sometimes, the situation was the opposite; though the hemolyzes were low, the level of lactates had been significantly exceeded. Therefore, to properly analyze the pRBCs, all possible parameters should be measured and any obtained value out of the range should serve as a basis for the rejection of the specific pRBCs (determining quality parameter). As each spectroscopic ratio defined in this invention depends on the concentrations in the SM of different chemical molecules, the appearance of which in the SM depends on the donor profile. All should be compared with the appropriate RV-Exp or RV-Res values. The basis for the rejection of specific pRBCs should be defined preferably on this ratio, which represents a failure to meet the necessary condition (determining quality ratio).

In the case of donor A, in Example 1 (Fig. 6), a comparison of each ratio with the RV-Exp suggests that the pRBCs should not be used after six weeks of storage. On the other hand, a comparison with RV- Res, based on two ratios, suggests that the pRBCs could be transfused to the patients at a high risk of complications after a transfusion during the first two weeks of storage, while one ratio indicates that only in the first week of storage are the RV-Res quality conditions met. An analysis of the additional ratio, which is possible when a 488 nm excitation is applied (Fig. 13), confirms the results and suggests the use of pRBCs for up to two weeks for restricted transfusions and for up to almost six weeks until the expiration time. Altogether, based on the determining quality ratio, the analysis of the pRBCs of donor A — on the basis of the Raman spectrum obtained with a 785 nm excitation, as is presented in Example 1 — suggests that this pRBC could be transferred to the patients at a high risk of complications after a transfusion only in the first week of storage, while it will expire for standard use after the sixth week of storage.

The same analysis was carried out for all donors, as is presented in examples 1-3, and the results are presented below based on the determining quality ratio.

Table 1. Summary of the results of the weeks of exclusion from transfusion for the pRBCs (donors A-F, as presented in Examples 1-3) that do not meet the corresponding reference values, as based on the label-free spectroscopic methodology presented in this invention.

As is presented in Table 1, the pRBCs of donor B had the worst quality and should never have been transferred to patients at a high risk of complications after a transfusion; their expiration length, based on this quality assessment, is five weeks. The pRBCs of donor D show a great quality until the fourth week of storage, but then demonstrate a rapid decline in quality, which determines their short length of expiration (four weeks). The best quality is observed in the pRBCs of donor F, which could be transferred to the patients at a high risk of complications after a transfusion up until five weeks of storage. Their quality maintains a sufficient level for use even in the seventh week of storage (so, these PCRBs could be used one week longer than is defined by the formal use date).

Such results clearly prove that the types of changes, as well as their kinetics, are different for each of the donors presented in the Examples. As well as that, the results prove that our methodology allowed for the label-free evaluation of the biochemical state of the pRBCs. The analysis has proven the utility of the invented methodology for indicating pRBCs’ that are of a quality that is or is not sufficient for transfusion.

Literature

Atkins, C. G. et al. Analyst 142, 2199-2210 (2017).

Atkins, C. G. et al. Analyst 141, 3319-3327 (2016).

Buckley, K. et al. Analyst 141, 1678-1685 (2016).

Cassanas, G., Morssli, M., Fabregue, E. & Bardet, L. J. Raman Spectrosc. 22, 409-413 (1991). Czamara, K. et al. J. Raman Spectrosc. 46, 4-20 (2015).

De Gelder, J. et al. J. Raman Spectrosc. 38, 1133-1147 (2007).

Dybas, J. et al. TrAC - Trends Anal. Chem. 85, 117-127 (2016).

Gautam, R., Oh, J.Y., Patelb, R. P., & Dluhy, R. A. Analyst 143, 5950-5958 (2018).

Flatt, J. F., Bawazir, W. M. & Brace, L. J. Front. Physiol. 5 JUN, 1-12 (2014).

Jamieson, L. E., Li, A., Faulds, K. & Graham D. R. Soc. Open Sci. 5, 181483 (2018).

Jordan, A. et al. Vox Sang. Ill, 8-15 (2016).

Krafft, C. et al. J. Chem. Soc. Rev., 45, 1819-1849 (2016).

Mathlouthi, M. & Luu, D. V. Res. 78, 225-233 (1980).

Michalewska, B. et al. Med. 175-242 (2009).

Mo, J. et al. Anal. Chem. 81, 8908-8915 (2009).

Movasaghi, Z., Rehman, S. & Rehman, I. U. Appl. Spectrosc. Rev. 42, 493-541 (2007).

Orlov, D. & Karkouti, K. Anaesthesia 70, 29-37 (2015).

Paglia, G. et al. Blood 128, e43-e50 (2016).

Perez-Guaita, D. et al. Chem. Rev. 118, 5330-5358 (2018).

Sugita Y. & Simon E.R. J Clin Invest. 44, 629-642 (1965).

Tzounakas, V. L. et al. Proteomics - Clin. Appl. 10, 791-804 (2016).

Vardaki, M.Z. et al. Analyst 143, 6006-6013 (2018).

Claims

1. A method for evaluating the quality of packed red blood cells (pRBCs) — where the pRBCs are composed mainly of red blood cells (RBCs) and a supernatant mixture (SM), as well as trace amounts of other blood cell components, including leukocytes and blood plasma — comprising the measurement of at least one Raman spectrum of the pRBCs and the execution of a chemometric analysis, wherein at least one Raman spectrum is collected for the SM and chemometric analysis of the collected spectrum is conducted for at least one ratio of Raman marker bands: a) 390±3M78±3 cm-¹, 472±3-578±3 cm-¹, 830±3-870±3 cm-¹, 870±3-910±3 cm-¹, 1520±3-1695±3 cm ¹, and 2867±3-2964±3 cm^-1 regarding the Rayleigh band located at 0 enr¹ and using an excitation of a wavelength above 600 nm; b) 2800±3-2900±3 enr¹ and 2900±3-3040±3 enr¹ regarding the Rayleigh band located at 0 enr¹ and using an excitation of a wavelength below 600 nm; and the obtained integral intensities ratio is compared with the corresponding reference value.

2. The method for evaluating the quality of pRBCs according to claim 1, wherein at least one Raman spectrum is collected from: a) the separated SM fraction, directly, through the PVC storage bag that is intended for pRBCs’ storage, using spatially offset Raman spectroscopy (SORS), and/or b) a dried SM fraction taken from the pilot tube after its centrifugation and placed on a calcium fluoride measurement slide, and or c) a dried SM fraction taken directly from the pRBC storage bag, from the separated fraction of the SM, and placed on a calcium fluoride measurement slide.

3. The method for evaluating the quality of pRBCs according to claim 1, wherein the Raman spectrum is collected using a 785 nm excitation wavelength for the bands at 390±3-478±3 cm ¹, 472±3-578±3 cm ¹, 830±3-870±3 cm ¹, 870±3-910±3 cm ¹, 1520±3-1695±3 cm ¹, and2867±3- 2964±3 enr¹ in regard to the Rayleigh band located at 0 cm ¹.

4. The method for evaluating the quality of pRBCs according to claim 1, wherein the Raman spectrum is collected using a 488 nm excitation wavelength for the bands at 2800±3-2900±3 enr 1 and 2900±3-3040±3 cm^-1 in regard to the Rayleigh band located at 0 cm ¹.

5. The method for evaluating the quality of pRBCs according to claims 1, 3 and 4 wherein chemometric analysis is conducted for at least one ratio of any two marker bands observed in the following spectral ranges: a) 472±3-578±3 enr¹ : 390±3-478±3 cm ¹; b) 870±3-910±3 cm-¹ : 830±3-870±3 cm ¹; c) 1520±3-1695±3 cm-¹ : 2867±3-2964±3 cm ¹; d) 2800±3-2900±3 cm ¹ : 2900±3-3040±3 cnr¹ or reversals of these marker band ratios.

6. The method for evaluating the quality of pRBCs according to claim 5, wherein the integral intensity ratios are compared with the reference values of expiration, which for the pRBCs intended for a transfusion are: a) not higher than 2.0 for 472±3-578±3cnr¹ : 390±3-478±3cm ¹; b) not lower than 0.8 for 870±3-910±3cnr¹ : 830±3-870±3cm ¹; c) not higher than 6.0 for 1520±3-1695±3cnr¹ : 2867±3-2964±3cm ¹; d) not higher than 0.7 for 2800±3cnr¹-2900±3cnr¹ : 2900±3cnr¹-3040±3cnr¹.

7. The method for evaluating the quality of pRBCs according to claim 5, wherein the integral intensity ratios are compared with the restrict reference values, which the pRBCs intended for a transfusion for a patient with a high risk of complications after the transfusion are: a) not higher than 1.5 for 472±3-578±3cnr¹ : 390±3-478±3cm ¹; b) not lower than 1.0 for 870±3-910±3cnr¹ : 830±3-870±3cm ¹; c) not higher than 3.1 for 1520±3-1695±3cnr¹ : 2867±3-2964±3cm ¹; d) not higher than 0.4 for 2800±3cm ¹-2900±3cm ¹ : 2900±3cm ¹-3040±3cm ¹.

8. The method for evaluating the quality of pRBCs according to claims 1, 6 or 7, wherein the integral intensity ratios are compared with the reference values of expiration calculated at the 42^nd day of pRBC storage and the restrict reference values calculated at the 21^st day of pRBC storage.

9. The method for evaluating the quality of pRBCs according to claims 1 or 5, wherein the results of the chemometric analysis are further processed with machine learning algorithms.

10. The method for evaluating the quality of pRBCs according to claim 1, wherein the measurement and chemometric analysis are carried out on pRBCs comprising an SM composed of glucose, mannitol, adenine, trisodium citrate, citric acid, and sodium dihydrogen phosphate.