Method and device for generating a metric representing a property of red blood cells in a sample

Abstract

The present disclosure relates to a method and device for generating a metric representing a property of red blood cells in a sample. In one implementation there is provided a method for generating the metric. The method comprises illuminating the red blood cells of the sample with a laser beam, measuring a first scatter of the laser beam by the red blood cells to obtain first scatter data, measuring a second scatter of the laser beam by the red blood cells to obtain second scatter data, and generating a metric representing a property of red blood cells in the sample using the obtained first scatter data and the obtained second scatter data. The first scatter corresponds to scattering in a first angular range and the second scatter corresponds to scattering in a second angular range. The first angular range and second angular range are different.

Description

[0001] METHOD AND DEVICE FOR GENERATING A METRIC REPRESENTING A

[0002] PROPERTY OF RED BLOOD CELLS IN A SAMPLE

[0003] The present disclosure relates to methods and devices for obtaining information about blood.

[0004] Red blood cells (RBCs) in storage undergo progressive run-down of metabolites and changes to shape and membrane properties, which can affect the speed with which oxygen is exchanged at capillaries. These changes may negatively affect transfusion outcomes. Certain inherited and acquired diseases of RBCs can also affect metabolic state (e.g. pyruvate kinase deficiency, glucose 6-phosphate dehydrogenase deficiency), shape and membrane properties (e.g. hereditary spherocytosis, HbH thalassemia, sickle cell disease), causing changes in oxygen exchange. The rate of oxygen unloading from individual RBCs, which can be described with a time constant, affects tissue oxygenation. The time constant is challenging to measure directly because of the resources needed for single-cell oxygen saturation imaging.

[0005] It is an object of the invention to at least partially address one or more of the issues mentioned above, or other issues.

[0006] According to an aspect of the invention, there is provided a method for generating a metric representing a property of red blood cells in a sample, the method comprising: illuminating the red blood cells of the sample with a laser beam; measuring a first scatter of the laser beam by the red blood cells to obtain first scatter data; measuring a second scatter of the laser beam by the red blood cells to obtain second scatter data; and generating a metric representing a property of red blood cells in the sample using the obtained first scatter data and the obtained second scatter data, wherein the first scatter corresponds to scattering in a first angular range and the second scatter corresponds to scattering in a second angular range, and the first angular range and second angular range are different.

[0007] Measurements of light scatter data can be obtained easily using readily available machines, such as flow cytometers. Such measurements can be used to generate a useful metric of a property of RBCs that can be generated cost-effectively and in a widely accessible manner. This facilitates rapid and inexpensive evaluation of RBCs held in blood banks, diagnostic purposes, or for scientific research, as well as supporting efficient retrospective meta-analyses of previously recorded datasets. The inventors have found that the approach of generating the metric using both of the first scatter data and the second scatter data provides improved performance against a directly measured property of RBCs.

[0008] Optionally, the generation of the metric comprises normalising one of the first scatter data and second scatter data using the other of the first scatter data and second scatter data. Normalising a measure of scatter data with another measure of scatter data when generating the metric may improve the accuracy of the metric.

[0009] Optionally, the generation of the metric comprises evaluating a term of the form <z(SSC + ?) / (FSC + y) + <5, where SSC represents a measured amount of second scatter derived from the second scatter data, FSC represents a measured amount of first scatter derived from the first scatter data, a, y and 5 are constants, and ?, y and 5 may be zero. The inventors have found that generating the metric in this way provides particularly high performance. The provision of the constants provides freedom for the metric to be calibrated to different combinations of measurement techniques used to obtain the SSC and the FSC. Known statistical methods such as multi-variate regression may be used to evaluate suitable values for the constants and thereby perform the calibration.

[0010] Optionally, the measuring of the first scatter corresponds to measuring a forward scatter of the laser beam by the red blood cells and the measuring of the second scatter corresponds to measuring a side scatter of the laser beam by the red blood cells. Using a measure of forward scatter as the first scatter measurement and a measure of side scatter as the second scatter measurement allows the metric to be generated using measurements that are commonly understood in the field and readily obtainable from commonly available machines, such as flow cytometry machines.

[0011] Optionally, the method further comprises aligning a characteristic axis of the red blood cells perpendicular to a common direction. Aligning the RBCs in this way allows the RBCs to have a consistent orientation when illuminated by the laser beam. This promotes more consistent contributions to the first scatter and second scatter and thereby improves a performance of the metric.

[0012] Optionally, the method further comprises measuring one or more further scatters of the laser beam by the RBCs, and generating the metric using the obtained first scatter data, the obtained second scatter data and the obtained further scatter data, wherein each further scatter corresponds to scattering in a further angular range, and each further angular range is different to the first angular range, second angular range, and any other further angular range. Using further measurements of scatter to generate the metric may further improve the accuracy of the metric.

[0013] According to an aspect of the invention, there is provided a device for generating a metric representing a property of RBCs in a sample, the device comprising: a laser beam arrangement configured to illuminate the red blood cells of the sample with a laser beam; a first optical detector configured to measure a first scatter of the laser beam by the red blood cells to obtain first scatter data; a second optical detector configured to measure a second scatter of the laser beam by the RBCs to obtain second scatter data; and the data processing system configured to generate a metric representing a property of RBCs in the sample using the obtained first scatter data and the obtained second scatter data, wherein the first scatter corresponds to scattering in a first angular range and the second scatter corresponds to scattering in a second angular range, and the first angular range and second angular range are different.

[0014] Embodiments of the disclosure will be further described by way of example only with reference to the accompanying drawings.

[0015] Figure 1 illustrates a red blood cell (RBC) being illuminated by a laser beam and the laser light scattered by the RBC.

[0016] Figures 2A to 2D schematically depict example devices for generating a metric representing a property of RBCs in a sample.

[0017] Figures 3 A and B show a relationship between measured oxygen unloading time constant and side-scatter (A) and forward-scatter (B) for a sample of RBCs.

[0018] Figure 4 shows performance of a metric representing time constant of oxygen unloading as a surrogate of time constant of oxygen unloading with coefficients obtained by fitting a non-linear model of side-scatter and forward scatter to the measured data of Figures 3A and B. Performance was assessed in terms of specificity (spec), sensitivity (sens), and accuracy (accu).

[0019] Figures 5 A and B show results for benchmarking of a metric representing time constant of oxygen unloading using two WADA calibration standards (numbers 3281 and 3337) representing three distinct levels of RBC shape, issued to four blood banking agencies (National Health Service Blood and Transplant (NHSBT) in England, Canadian Blood Services (CBS) in Canada, Lifeblood in Australia, and Banc de Sang i Teixits (BST) in Spain). Figure 6 shows temporal evolution of markers of storage lesion in blood units from six anonymous donors (labelled A-F) including a metric representing a property of RBCs.

[0020] Figures 7A to 7D show results for oxygen unloading time constant (single-cell oxygen saturation imaging), side scatter, forward scatter and a metric representing a property of RBCs for freshly drawn blood samples and samples stored for 2 days at room temperature.

[0021] Figures 8A to 8D show results for oxygen unloading time constant (single-cell oxygen saturation imaging), side scatter, forward scatter and a metric representing a property of RBCs for freshly drawn blood samples and samples stored for 2 days under refrigerated conditions.

[0022] Figures 9A to 9D show results for side-scatter, forward scatter, a metric representing a property of RBCs, and the time constant of oxygen unloading measured for samples of stored blood after rejuvenation treatment using a mixture of phosphate (Ph), inosine (I), pyruvate (P) and adenine (A), or without individual components as shown.

[0023] Figure 10A shows results for the distribution of a metric representing a property of RBCs over time in a large cohort of donors, stratified by the delay between sample collection and measurement.

[0024] Figure 10B shows correlation between a metric representing a property of RBCs and RBC parameters normally detected on haematology analysers.

[0025] Figure IOC shows a relationship between a metric representing a property of RBCs and blood group.

[0026] Figure 11 A shows a relationship between a metric representing a property of RBCs and age, stratified by sex assigned at birth for a first data set.

[0027] Figure 1 IB shows a relationship between a metric representing a property of RBCs and age, stratified by sex assigned at birth and smoking for the first data set.

[0028] Figure 11C shows a relationship between a metric representing a property of RBCs and age, stratified by sex assigned at birth for a second data set.

[0029] Figures 12A and 12B show a relationship between the measured 2,3- diphosphoglycerate (2,3 -DPG) normalized to haemoglobin (Hb) concentration and a metric representing a property of RBCs with correlation measured by Pearson’s correlation coefficient, and a Bland- Altman plot for the same data. Figures 13 A and 13B show a relationship between the measured adenosine triphosphate (ATP) normalized to haemoglobin (Hb) concentration and a metric representing a property of RBCs with correlation measured by Pearson’s correlation coefficient, and a Bland- Altman plot for the same data.

[0030] Embodiments of the present disclosure relate to a method for generating a metric representing a property of RBCs in a sample, and a device for the same. The device may be configured to perform the method.

[0031] Figure 1 illustrates a RBC 6 being illuminated by a laser beam 1. At least a portion of the laser beam 1 may be scattered in all directions as scattered light 11. The scattered light 11 may result from reflection and / or refraction of the laser beam 1 by the RBC 6.

[0032] Figures 2A-2D depict example devices 100 for generating a metric representing a property of RBCs 6 in a blood sample. The devices 100 may be configured to perform any of the methods for generating a metric representing a property of RBCs in a sample described herein.

[0033] As exemplified in Figures 2A-2D, the device 100 may comprise a laser beam arrangement 10. The laser beam arrangement 10 is configured to illuminate the RBCs 6 of the sample with a laser beam 1. In one implementation, the device 100 is configured to illuminate the RBCs 6 with the laser beam 1 one RBC 6 at a time. Alternatively or additionally, the device 100 may be configured to illuminate multiple RBCs 6 simultaneously. Any laser beam arrangement suitable for providing the data required may be used. Details of the data requirements are provided below.

[0034] In one example implementation, a laser beam arrangement emitting a beam of 633nm is used, but lasers having other operating parameters may be used.

[0035] Referring to Figure 2A, in one implementation the device 100 comprises a first optical detector 21. The first optical detector 21 is configured to measure a first scatter of the laser beam 1 by the RBCs 6 to obtain first scatter data. The device 100 further comprises a second optical detector 22. The second optical detector 22 is configured to measure a second scatter of the laser beam 1 by the RBCs 6 to obtain second scatter data.

[0036] The first optical detector 21 and the second optical detector 22 may be configured as necessary to perform method steps described herein that comprise measuring scatter of a laser beam. Scatter data may be obtained by measuring at least a proportion of the laser light scattered by the RBCs 6 of the sample within an angular range. The first scatter corresponds to scattering in a first angular range. The second scatter corresponds to scattering in a second angular range. In the example of Figure 2 A, the first optical detector 21 is configured to measure the first scatter in the angular range 51. The second optical detector is configured to measure the second scatter in the angular range 52. The first angular range 51 and second angular range 52 are different, e.g. the first angular range 51 and second angular range 52 do not encompass exactly the same portion of three- dimensional space. The angular ranges may be similar in other aspects, such as having the same absolute angular size. The angular ranges may be otherwise identical other than a displacement in space or around one or more axes. In some implementations, the first angular range 51 and second angular range 52 are non-overlapping, e.g. none of the second range is contained within the first range.

[0037] The first scatter and the second scatter of the laser light by the RBCs 6 may be measured simultaneously. However, this is not essential; for example, the first scatter and second scatter may be measured one after the other at closely spaced intervals and / or in an interlaced measurement mode.

[0038] The device further comprises a data processing system (not illustrated) configured to generate a metric representing a property of RBCs in the sample using the obtained first scatter data and the obtained second scatter data. The data processing system may comprise any suitable combination of data processing hardware (e.g., CPUs, memory, data transmission lines, etc.), firmware, and / or software for performing the functionality required to perform the method.

[0039] The generation of the metric may include normalising one of the first scatter data and second scatter data using the other of the first scatter data and second scatter data. Normalising may involve a division of the first scatter data with the second scatter data, for example.

[0040] In an implementation, the generation of the metric representing a property of RBCs comprises evaluating a term of the form <z(SSC + ?) / (FSC + y) + <5, where SSC represents a measured amount of second scatter derived from the second scatter data, FSC represents a measured amount of first scatter derived from the first scatter data, and a, (3 , y and 5 are constants. The constants ?, y and / or 5 may be zero, and in this case it is not necessary to include the constants ft, y and / or in the term above. The constants may be determined experimentally, for example by using multivariate linear regression to compare the obtained scatter data obtained from a control sample to a standard laboratory method of measuring the property of RBCs of the same control sample. Once the constants have been determined for a particular apparatus or method of gathering the scatter data, the metric representing the property can be generated for further samples. This prevents the need for full laboratory analysis for each sample.

[0041] In the example of Figure 2A, the first optical detector 21 is substantially aligned with the axis of the incident laser beam 1 (e.g., to primarily receive light propagating substantially parallel to the axis of the laser beam 1 after interaction with the RBCs 6) and the second optical detector 22 is substantially perpendicular to the axis of the incident laser beam 1 (e.g., to primarily receive light propagating substantially perpendicularly to the axis of the incident laser beam 1).

[0042] In this example the first optical detector 21 is configured to measure a first scatter that corresponds to measuring a forward scatter of the laser beam by the RBCs. The second optical detector is configured to measure a second scatter that corresponds to measuring a side scatter of the laser beam by the RBCs 6. In an implementation, the average angle between the measuring of the forward scatter and the measuring of the side scatter is greater than 20 degrees.

[0043] The angular ranges in which the scatter is measured may be defined as a cone with an apex on the incidence point A of the laser beam on the sample. In an implementation, the first angular range 51 is a first cone 51. The first cone 51 has an axis aligned with the axis of the laser beam 1 and a half angle of 20 degrees. The second angular range 52 is a second cone 52. The second cone 52 has an axis aligned perpendicular to the axis of the laser beam 1 and a half angle of 70 degrees. The forward scatter may be obtained by a measurement of at least a proportion of the laser light that is scattered by the red blood cells RBC within the first cone 51. Side scatter may be obtained by a measurement of at least a proportion of the laser light that is scattered by the RBCs 6 within the second cone 52. It is not necessary to measure the entirety of the laser light scattered by the RBCs 6 within an angular range to obtain a measurement of scatter.

[0044] When the first scatter data corresponds to forward scatter and the second scatter data corresponds to side scatter, the normalisation of the side scatter data using the forward scatter data, as discussed above, may control for a measure of size of the RBCs when generating the metric representing a property of the RBCs. This can improve the accuracy of the metric generated from the scatter data.

[0045] In other implementations, the detectors may be arranged at other angles and / or configured to measure other angular ranges, provided that the first optical detector and second optical detector are configured to measure a first angular range and a second angular range that are different.

[0046] In the example of Figure 2B the first optical detector 21b is arranged at an angle of 45 degrees to the axis of the laser and is configured to measure scattered light within the first angular range 51b. In this case, the first scatter detected by the first optical detector 21b may no longer fall within the definition of forward scatter. The first angular range 51b and second angular range 52b depicted in Figure 2B are substantially the same size, i.e. have the same absolute angular range.

[0047] Measuring angular ranges that have a larger average angular distance between the measurements of first scatter and measurements of second scatter may be preferable. A large angular distance between the first and second measurements of scatter may improve the accuracy of the metric generated from the obtained scatter data. If further scatters of the laser light by the RBCs are measured, as described later, the same advantage may apply. The example of Figure 2A is configured to have a large average angular distance between the measurements of first scatter and second scatter. However, a large angular distance is not essential. In an example illustrated in Figure 2C, an average angle between the measuring of the forward scatter by the first optical detector 21c and the measuring of the side scatter by the second optical detector 22c is less than the example of Figure 2 A. In the example of Figure 2C, the average angle between the measuring of forward scatter and the measuring of side scatter is greater than 20 degrees; however the angle may be less than 20 degrees in some implementations.

[0048] In the examples of Figures 2A to 2D, the optical detectors are arranged to lie within their corresponding angular ranges. Other arrangements are possible, provided the optical detectors can measure at least a proportion of the laser light scattered within the corresponding angular range. For example, an arrangement of mirrors and / or lenses could be configured to appropriately transmit the scattered light from an angular range to an optical detector such that the scattered light can be detected. The RBCs in the sample may be aligned when they pass through the laser beam. Aligning the RBCs in a consistent manner as they pass through the laser beam increases the quality of the scatter data collected, and consequently increase the accuracy of the metric generated from the scatter data representing a property of the RBCs. For example, the RBCs may be aligned such that a characteristic axis of the RBCs is perpendicular to a common direction. The characteristic axis may be an axis of rotational symmetry. Other characteristic axes, such as other axes of symmetry, may be suitable.

[0049] In an implementation, RBCs are entrained in a flow of liquid during the illumination of the RBCs. This is illustrated in the examples of Figures 2A to 2D. The flow of liquid 40 flows in a conduit 30 in the direction indicated by the arrow F. The conduit 30 may be arranged in any way that is suitable to enable a flow of liquid 40 to flow. The way in which the flow of liquid is created is not particularly limited. A device for generating the metric may include a flow controller (not illustrated) configured to provide the flow of liquid 40 for entraining the RBCs 6 in the flow of liquid during the illumination of the red blood cells.

[0050] The RBCs 6 may be injected into the flow of liquid 40 in the conduit 30. The injection may take place substantially in the centre of the flow, such that the RBCs 6 are entrained into the flow of liquid 40 approximately along the centreline of the flow. In implementations of this type, the laser beam 1 may be configured to intersect the conduit 30 such that the laser beam illuminates the RBCs 6 of the sample by intersecting the flow of liquid in the conduit. The RBCs 6 may be injected into the flow of liquid one at a time, so that each RBC 6 passes through the laser beam individually and measurements of scatter can be made for each RBC 6. In the examples of Figures 2A-2D, the laser beam 1 intersects the conduit 30 perpendicular to the flow direction of the flow of liquid 40 and in such a way that the RBCs 6 entrained in the flow of liquid 40 pass through the laser beam 1. The exact angle of incidence of the laser beam 1 relative to the flow of liquid 40 is not critical. Any angle can be used as long as the laser beam 1 can be arranged to illuminate the RBCs 6 and the light scattered by the RBCs can be collected. At least a portion of a wall of the conduit 30 may be at least partially transparent to the wavelength of the laser beam 1 to allow the laser beam 1 to illuminate the RBCs 6 through the wall of the conduit 30 and to allow scattered light to pass though the wall and be measured. The common direction referred to above (for defining alignment of the RBCs 6 in the flow) may be parallel to the direction F of the flow of liquid 40. The RBCs 6 may be aligned such that the characteristic axis (e.g., an axis of rotational symmetry) is perpendicular to this common direction. In this example, the RBCs are aligned such that, if the axis of rotational symmetry is considered a Z-axis in a Cartesian coordinate system, a vector representing the direction F of flow of the liquid 40 lies in the X-Y plane of the cell. The RBCs 6 will retain a degree of rotational freedom around the characteristic axis (e.g., axis of rotational symmetry).

[0051] In an implementation, the alignment of the RBCs 6 is achieved by the flow of liquid 40 being such as to hydrodynamically align the RBCs 6 in the flow. The RBCs 6 may be aligned by injection of the RBCs 6 into a flow of liquid 40 where an average linear velocity of the flow is greater than the velocity of the RBCs immediately post-injection into the flow of liquid 40. This effect may be further intensified where the average linear velocity of the flow of liquid 40 increases for at least a portion of the conduit downstream of the point at which the RBCs 6 are injected, and / or if the flow of liquid 40 in the conduit 30 is laminar. The increase in velocity may be achieved by the reduction in cross section of the conduit 30 in the direction F of the flow.

[0052] The alignment of the RBCs 6 may be achieved in other ways than by hydrodynamic forces acting on the RBCs 6 in a flow of liquid. For example, acoustic focusing, optical tweezers, optical lattices, or dielectrophoresis may be used, instead of or in conjunction with the hydrodynamic focusing in any combination, in order to align the RBCs as they are illuminated by the laser beam. Additionally, factors such as viscosity, ionic strength, pH, and the presence of specific additives or surfactants in the liquid can influence the behaviour of cells in the flow stream and their interaction with the focusing forces.

[0053] Further scatter data may be collected other than the first scatter data and second scatter data. Each further scatter may correspond to a further angular range, and each further angular range may be different to the first angular range, second angular range and any other further angular range. For example, one or more further scatters of the laser beam by the RBCs may be measured and the metric may be generated using the obtained first scatter data, the obtained second scatter data and the obtained further scatter data. In one implementation, a further measurement of scatter may be taken in an angular range between the first angular range and second angular range in one plane. The example illustrated in Figure 2D shows a further optical detector 23 configured to measure further scatter data within the further angular range 53. The further angular range 53 lies between ranges 5 Id and 52d where the measurements of first scatter and second scatter are made. The one or more further angular ranges that may be measured are not limited to the example of Figure 2D. Any number of further scatters of the laser beam by the RBCs in further angular ranges may be measured.

[0054] In an implementation a flow cytometer is used for illuminating the RBCs of the sample with a laser beam, measuring a first scatter of the laser beam by the RBCs to obtain first scatter data, and measuring a second scatter of the laser beam by the RBCs to obtain second scatter data. The flow cytometer may also perform the functions of aligning the RBCs and illuminating the RBCs of the sample with a laser beam.

[0055] The metric generated from the scatter data may represent a time constant of oxygen unloading from the RBCs. The obtained scatter data may represent information about the shape of the RBCs. Healthy human RBCs have a thin, approximately biconcave shape. RBCs stored in a blood bank may undergo storage lesion, which can involve a remodelling the shape of the RBCs towards a more spherical shape. RBC diseases affecting cytoskeletal proteins, such as certain inherited anaemias, can affect shape. RBC diseases affecting metabolism can affect shape. RBC geometry is rate limiting for oxygen diffusion in and out of the RBCs. In general, the closer the shape of the RBCs is to a sphere, the slower the oxygen transport out of the RBCs occurs. As a functional parameter, the time constant is important to measure because it affects tissue oxygenation. The scatter data can be used to generate a metric which represents the time constant of oxygen unloading in the blood with a high degree of accuracy. This implementation of the metric provides a way to assess the metabolic health of RBCs using a single number which is easy and cost-effective to generate, and the single numerical reference is a characteristic of the cells which is intuitive to understand and allows samples of blood to be readily compared. Furthermore, large datasets of scatter data already exist. If the methods of collecting the scatter data can be calibrated against a property of RBCs, the metric can also be retroactively generated for previous datasets.

[0056] Oxygen unloading kinetics are described as a diffusion-reaction process that is critically dependent on metabolic state and diffusion pathlength, implicating both haemoglobin-oxygen stability and cell shape. In some implementations, the metric corresponds to one or more biochemical readouts as well as, or in place of, the time constant of oxygen unloading. The biochemical readouts may include intracellular 2,3- DPG and ATP concentrations. A metric corresponding to intracellular 2,3-DPG and intracellular ATP concentrations may be expressed as a concentration relative to a mass of haemoglobin in the blood (Hb).

[0057] The scatter data may take various forms. The scatter data may, for example, comprise an average of the individual measurements of scatter made for each RBC of a plurality of RBCs, where the light scattered is measured for each cell individually.

[0058] If the angular range encompasses the axis of the laser beam, such as when measuring forward scatter as in the arrangement in Figures 2A, 2C and 2D, the angular range 51 may be adjusted such that a portion of the range corresponding to the axis of the laser beam is excluded, to prevent the optical detector being exposed to the laser beam. The direct exposure of an optical detector to the laser beam may prevent an accurate measurement of scatter. The exclusion may be achieved by appropriate configuration of the detector, by a physical block or filter placed between the laser beam and the detector, or any other suitable means.

[0059] The high turnover of circulating human RBCs (lifespan of -100 days) reflects the vulnerability of their physiological function to metabolic and mechanical stressors in vivo. This vulnerability also manifests under ex vivo storage, and may affect transfusion outcomes. A full appraisal of RBC physiological quality necessitates measurements of their O2 handling ability but these are difficult to implement in clinical or blood banking settings, which is why correlations are routinely sought with metrics that are readily obtainable, e.g. from standard haematology analysers, such as flow cytometers, that are in use globally. Some surrogates, such as the colour (e.g., light absorbance) of blood used as a metric of oxygen-carrying capacity on haemoglobin, are calibratable and accurate. In contrast, appraisal of O2 unloading rate, a kinetic quantity, is conceptually more difficult to attribute to any currently measurable parameter.

[0060] In an implementation, the metric according to this disclosure relates to side-scatter, normalized to forward-scatter, which has been found to be correlated to the time constant of oxygen unloading. The basis for this correlation is that the time constant of oxygen unloading describes a diffusion-reaction process that is highly sensitive to diffusion pathlength, i.e. cell thickness, as shown by the effect of storage under standard NHS Blood and Transplant (NHSBT) protocols and on an accelerated scale without additional storage lesion-mitigating measures. As cells become metabolically depleted and more spherical, diffusion pathlength increases. This has an amplified effect on the time constant because of Fick’s law, i.e. a doubling of distance increases diffusion time by four-fold. Side-scatter is highest with biconcave cells because such dumbbell-shaped cells may produce two, broader peaks of voltage response on the optical detector when passing through the flow cytometry cell. Spherically remodelled RBCs, in contrast, are likely to produce one shorter side-scatter peak. Normalization to forward- scatter improves signal-to-noise by controlling for a measure of cell size, and generates a metric according to this disclosure that is stable over time (correlation has been measured as stable over a period of 3 years) and between at least two centres that use multiple analysers each. Since metabolism, shape, and O2 handling are inter-related, variants of the metric according to this disclosure with different coefficients can predict [2,3 -DPG] and [ATP], The former is particularly timely given the cost-ineffectiveness and withdrawal of commercial biochemical kits for 2,3-DPG, and the continued statutory need to test proposed changes in materials or methods in blood manufacture and storage.

[0061] An immediate use of the metric according to this disclosure is as a cost-effective quality-control marker of transfusion products. Since haematology analysers are widely available, the metric according to this disclosure can be implemented immediately in screening blood units for quality and possible stratification to match product with the most appropriate recipient, e.g. faster O2 unloading for those needing massive transfusion in traumatic brain injury or paediatric intensive care. Additionally, the metric according to this disclosure can be used to ensure compliance between different blood banking systems, especially in an effort to bring facilities in lower / middle income countries in line with best practice and identifying vulnerable handling steps that critically determine quality of the end-product. Any changes that inevitably arise to challenge blood banking manufacture or process, such as the upcoming ban on DHEP plasticiser, will mandate quality-control to ensure that the proposed substitutes produce a product that is no worse from the original. The metric of the disclosure may be used in a large-scale meta-analysis of the metric against factors providing information about oxygen transport, for example in relation to diseases and / or athletic performance. Figures 3 to 13 provide data illustrating the uses and performance of a metric representing a property of RBCs of the disclosure.

[0062] Figure 3 A and B show a relationship between measured oxygen unloading time constant and side-scatter (A) and forward-scatter (B) for a sample of red blood cells.

[0063] Figure 4 shows performance of a metric representing a quantity of RBCs as a surrogate of time constant of oxygen unloading with coefficients obtained by fitting a nonlinear model of side-scatter and forward scatter to the measured data of Figures 3 A and B. In this example, the generation of the metric involved evaluating a term of the form <z(SSC + ?) / (FSC + y) + 5, (in this case the term was of the equivalent form a + b*(SSC + c) / FSC). FN represents false negative, FP represents false positive, TN represents true negative, and TP represents true positive. A Chi-square test (y2statistic 115.0012) indicates highly significant correlation (P<0.00001), with good specificity (spec), sensitivity (sens), and accuracy (accu), greater than 80%. 95% confidence interval for specificity: 81 %-93%; sensitivity: 72%-86%. These data show that a metric representing a quantity of RBCs of the disclosure provides a good surrogate for Ch-handling kinetics in RBCs.

[0064] Figure 5 A and B show results for quality control of a metric representing time constant of oxygen unloading. The quality control was performed using two so-called WADA calibration standards (numbers 3281 and 3337) representing three distinct levels of RBC shape, issued to four blood banking agencies (National Health Service Blood and Transplant (NHSBT) in England, Canadian Blood Services (CBS) in Canada, Lifeblood in Australia, and Banc de Sang i Teixits (BST) in Spain). A metric representing a property of RBCs of the disclosure was calculated by each blood banking agency without specialist calibration of the haematology analysers used. The error bars are 95% confidence interval (CI) around mean. The metrics representing a property of RBCs of the disclosure generated by all of the blood banks were within 5% error of the global mean. The results show that a metric representing a property of RBCs of the disclosure is comparable between locations, even without specialist calibrations.

[0065] Figure 6 shows temporal evolution of markers of storage lesion in blood units from six anonymous donors (labelled A-F) including a metric representing a property of RBCs. The shaded range reference a metric representing a property of RBCs of the disclosure

[0066] (labelled Flowscore), [ATP] or [2,3-DPG] in freshly drawn RBCs, and a statutory range of haemolysis. Performance of the metric representing a property of RBCs was benchmarked against current markers of storage lesion ([2,3-DPG], [ATP], and haemolysis) which guide blood-banking practice and improvements to materials and methods. Red cell concentrates (RCC) from six donors showed a dramatic depletion of 2,3-DPG over the first 3 weeks of storage, beyond which measurements provided no further insight into the progression of the storage lesion. In early storage, ATP levels were supra-physiological (supported by a buffering effect, e.g. consumption of 2,3-DPG), and then decreased linearly. Thus, for much of the storage duration, [ATP] remained within the range for fresh blood and associated with acceptable post-transfusion recovery (expected count increment). Haemolysis accelerated after week 5 of storage. Significantly, the metric representing a property of RBCs met important criteria for a progressive marker of storage lesion that relates mechanistically to 02 transport by RBCs. Unlike [2,3-DPG] or haemolysis but akin to [ATP], the metric representing a property of RBCs changed steadily during storage, reflecting the progressive attrition of 02 unloading kinetics. Unlike [ATP], the metric representing a property of RBCs in the early phase of storage was near its reference range and worsened over time. The slope of the metric representing a property of RBCs timecourse correlated significantly with the initial rate of 2,3-DPG depletion (Pearson’s test rho=-0.875, P=0.022) but not ATP depletion (Pearson’s test rho=-0.586; P=0.22), possibly because the latter is heavily buffered. Overall, a metric representing a property of RBCs of the disclosure meets criteria for a sensitive and accurate measure of storage lesion progression.

[0067] Figures 7A to 7D show results for oxygen unloading time constant (single-cell oxygen saturation imaging), side scatter, forward scatter and a metric representing a property of RBCs of the disclosure (labelled Flowscore) for freshly drawn blood samples and samples stored for 2 days at room temperature.

[0068] Figures 8A to 8D show results for oxygen unloading time constant (single-cell oxygen saturation imaging), side scatter, forward scatter and a metric representing a property of RBCs for freshly drawn blood samples and samples stored for 2 days under refrigerated conditions.

[0069] Figures 7 and 8 demonstrate that interventions that affect oxygen handling kinetics also affect a metric representing a property of RBCs of the disclosure in a proportional manner. Figures 9A to 9D show results for side-scatter, forward scatter, a metric representing a property of RBCs of the disclosure (labelled Flowscore), and the time constant of oxygen unloading measured for samples of blood with different rejuvenation treatments. The test for concordance between the metric representing a property of RBCs and oxygen unloading time constant was the effect of biochemical rejuvenation of red cell concentrations using a mixture containing phosphate, inosine, pyruvate, and adenine (PhlPA). Paired measurements of the metric representing a property of RBCs and time constant were made on expired blood units split five ways: receiving no treatment, PhlPA treatment, or one of three variants of PhlPA with one component among pyruvate, inosine or adenine replaced iso-osmotically with NaCl. Phosphate omission was not tested because its absence would compromise pH control. There was a significant recovery of the metric representing a property of RBCs and time constant following PhlPA treatment, and inosine was deemed the most significant contributor to rejuvenation efficacy. The observation that the metric representing a property of RBCs and time constant responded similarly to rejuvenation confirms the predictive power of the metric representing a property of RBCs (N=4 biological repeats).

[0070] Figures 10 and 11 show that a metric representing a property of RBCs of the disclosure can be extracted from previously collected haematology analyser datasets where there are side- and forward- scatter measurements to identify novel associations. Figure 10A shows the distribution of a metric representing a property of RBCs of the disclosure (labelled Flowscore) calculated from NH4R COMPARE study data by frequency histograms stratified by time delay between blood sampling and measurement. NH4R COMPARE includes recordings from 29,021 participants, 94% of whom had SSC and FSC measurements. Figure 10B shows correlation between a metric representing a property of RBCs of the disclosure (labelled Flowscore) and RBC parameters annotated with units (Pearson’s test). Data obtained after >1 day delay from sampling were excluded. MCV: mean corpuscular volume (fL), MCH: mean corpuscular haemoglobin (pg), RBC He: haemoglobin content of mature red cells, RET He: haemoglobin content of reticulocytes, MacroR: percentage of macrocytic RBCs, RDW SD: red cell distribution width (standard deviation), HYPER-He: percentage of RBC with cellular haemoglobin content higher than 49 pg, RET: percentage of reticulocytes, LFR: percentage of low fluorescence reticulocytes, MFR: percentage of medium fluorescence reticulocytes, IRF: percentage of immature reticulocytes, NRBC: percentage of nucleated RBCs; FRC: percentage of fragmented RBCs, HYPO-He: percentage of RBC with cellular haemoglobin content lower than 17 pg, MicroR: percentage of microcytic RBCs. Figure IOC shows a relationship between a metric representing a property of RBCs of the disclosure and blood group (mean, 95% confidence interval, CI). Figure 11 A shows a relationship between a metric representing a property of RBCs and age, stratified by sex assigned at birth for the NIHR COMPARE data set. Figure 1 IB shows a relationship between a metric representing a property of RBCs and age, stratified by sex assigned at birth and smoking for the same data set. Figure 11C shows a relationship between a metric representing a property of RBCs and age, stratified by sex assigned at birth for the LifeLines dataset, a multidisciplinary prospective population-based cohort study examining the health and health- related behaviours of 167,729 persons in northern Netherlands using a three-generation design and employs a broad range of investigative procedures in assessing the biomedical, socio-demographic, behavioural, physical, and psychological factors contributing to health and disease.

[0071] Figures 12A and 12B show a relationship between the measured 2,3-DPG normalized to Hb concentration and a metric representing a property of RBCs of the disclosure (labelled MetaFlowScore) for data from six studies analysing 518 samples of stored blood measured on a Sysmex XN-1000 analyser with correlation measured by Pearson’s correlation coefficient, and a Bland- Altman plot for the same data. The Bland- Altman plot shows good agreement between measurement and a metric representing a property of RBCs of the disclosure, without systematic bias. This data shows a metric representing a property of red blood cells in a sample representing an intracellular 2,3-DPG concentration relative to a mass of Hb (haemoglobin). In some embodiments, the intracellular 2,3-DPG concentration may be expressed in other ways, such as a mM concentration.

[0072] Figures 13 A and 13B show a relationship between the measured ATP normalized to Hb concentration and a metric representing a property of RBCs of the disclosure (labelled MetaFlowScore) for data from six studies analysing 518 samples of stored blood measured on a Sysmex XN-1000 analyser with correlation measured by Pearson’s correlation coefficient, and a Bland-Altman plot for the same data. The Bland-Altman plot for these data again shows good agreement between measurement and a metric representing a property of RBCs of the disclosure, without systematic bias. This data shows a metric representing a property of red blood cells in a sample representing an intracellular ATP concentration relative to a mass of Hb (haemoglobin). In some embodiments, the intracellular ATP concentration may be expressed in other ways, such as a mM concentration.

[0073] Cross reference to related applications

[0074] This application claims priority from GB 2408457.6 filed on 13 June 2024, the contents of which are hereby incorporated by reference.

Claims

CLAIMS1. A method for generating a metric representing a property of red blood cells in a sample, the method comprising: illuminating the red blood cells of the sample with a laser beam; measuring a first scatter of the laser beam by the red blood cells to obtain first scatter data; measuring a second scatter of the laser beam by the red blood cells to obtain second scatter data; and generating a metric representing a property of red blood cells in the sample using the obtained first scatter data and the obtained second scatter data, wherein the first scatter corresponds to scattering in a first angular range and the second scatter corresponds to scattering in a second angular range, and the first angular range and second angular range are different.

2. The method of claim 1, wherein the first angular range and the second angular range are non-overlapping.

3. The method of claim 1 or 2, wherein the generation of the metric comprises normalising one of the first scatter data and second scatter data using the other of the first scatter data and second scatter data.

4. The method of any preceding claim, wherein the generation of the metric comprises evaluating a term of the form <z(SSC + ?) / FSC, where SSC represents a measured amount of second scatter derived from the second scatter data, FSC represents a measured amount of first scatter derived from the first scatter data, and a and / 3 are constants.

5. The method of any of claims 1 to 3, wherein the generation of the metric comprises evaluating a term of the form <z(SSC + ?) / (FSC + y) + 8, where SSC represents a measured amount of second scatter derived from the second scatter data, FSC represents ameasured amount of first scatter derived from the first scatter data, a, / 3, y and 5 are constants, and ?, y and 5 may be zero.

6. The method of any preceding claim, wherein the measuring of the first scatter corresponds to measuring a forward scatter of the laser beam by the red blood cells and the measuring of the second scatter corresponds to measuring a side scatter of the laser beam by the red blood cells.

7. The method of claim 6, wherein an average angle between the measuring of the forward scatter and the measuring of the side scatter is greater than 20 degrees.

8. The method of claim 6 or 7, wherein: a point at which the laser beam is incident upon the sample defines an incidence point; a first cone is defined by having an apex at the incidence point, an axis parallel with the axis of the laser beam, and an included half angle of 20 degrees; and the forward scatter data is obtained by a measurement of at least a proportion of the laser light scattered by the sample within the first cone.

9. The method of claim 8, wherein: a second cone is defined by having an apex at the incidence point, an axis perpendicular to the axis of the laser beam, and an included half angle of 70 degrees; and the side scatter data is obtained by a measurement of at least a proportion of the laser light scattered by the sample within the second cone.

10. The method of any preceding claim, further comprising aligning a characteristic axis of the red blood cells perpendicular to a common direction.

11. The method of claim 10, wherein the characteristic axis is an axis of rotational symmetry.

12. The method of claim 10 or 11, wherein:the red blood cells are entrained in a flow of liquid during the illumination of the red blood cells; and the common direction is the direction of the flow of liquid.

13. The method of any of claims 1 to 11, wherein the red blood cells are entrained in a flow of liquid during the illumination of the red blood cells.

14. The method of claim 12 or 13, wherein the flow is such as to hydrodynamically align the red blood cells in the flow.

15. The method of claim 13 or 14, wherein: the flow of liquid flows in a conduit; the method comprises injecting the red blood cells of the sample into the flow of liquid in the conduit; and and the laser beam illuminates the red blood cells of the sample by intersecting the flow of liquid in the conduit.

16. The method of claim 15, wherein an average linear velocity of the flow of liquid is greater than the velocity of the red blood cells immediately post-injection into the flow of liquid.

17. The method of claim 16, wherein the average linear velocity of the flow of liquid increases for at least a portion of the conduit downstream of the point at which the red blood cells are injected.

18. The method of any of claims 15 to 17, wherein the flow of liquid in the conduit is laminar.

19. The method of any preceding claim, wherein a flow cytometer is used for: illuminating the red blood cells of the sample with a laser beam; measuring a first scatter of the laser beam by the red blood cells to obtain first scatter data; andmeasuring a second scatter of the laser beam by the red blood cells to obtain second scatter data.

20. The method of any preceding claim, wherein the method further comprises measuring one or more further scatters of the laser beam by the red blood cells, wherein each further scatter corresponds to scattering in a further angular range, and each further angular range is different to the first angular range, second angular range, and any other further angular range.

21. The method of any preceding claim, wherein the metric represents at least one of the following: a time constant of oxygen unloading from the red blood cells; an intracellular 2,3 -DPG concentration, optionally relative to a mass of Hb; and an intracellular ATP concentration, optionally relative to a mass of Hb.

22. The method of any preceding claim, wherein the first scatter and the second scatter are measured simultaneously.

23. A device for generating a metric representing a property of red blood cells in a sample, the device comprising: a laser beam arrangement configured to illuminate the red blood cells of the sample with a laser beam; a first optical detector configured to measure a first scatter of the laser beam by the red blood cells to obtain first scatter data; a second optical detector configured to measure a second scatter of the laser beam by the red blood cells to obtain second scatter data; and the data processing system configured to generate a metric representing a property of red blood cells in the sample using the obtained first scatter data and the obtained second scatter data, wherein the first scatter corresponds to scattering in a first angular range and the second scatter corresponds to scattering in a second angular range, and the first angular range and second angular range are different.

24. The device of claim 23, wherein the device is configured to align a characteristic axis of the red blood cells perpendicular to a common direction.

25. The device of claim 23 or 24, wherein the device comprises a flow controller configured to provide a flow of liquid for entraining the red blood cells in the flow of liquid during the illumination of the red blood cells.

26. The device of claim 25, wherein the flow is such as to hydrodynamically align the red blood cells in the flow.

27. The device of any preceding claim, wherein the measuring of the first scatter corresponds to measuring a forward scatter of the laser beam by the red blood cells and the measuring of the second scatter corresponds to measuring a side scatter of the laser beam by the red blood cells.

28. The device of claim 27, wherein an average angle between the measuring of the forward scatter and the measuring of the side scatter is greater than 70 degrees.

29. The device of any of claims 23 to 28, wherein: a point at which the laser beam is incident upon the red blood cells in the fluid conduit defines an incidence point; a first cone is defined by having an apex at the incidence point, an axis parallel with the axis of the laser beam, and an included half angle of 20 degrees; and the first optical detector is configured to measure the forward scatter by measuring at least a proportion of light scattered within the first cone.

30. The device of claim 29, wherein: a second cone is defined by having an apex at the incidence point, an axis perpendicular to the axis of the laser beam, and an included half angle of 70 degrees; and the second optical detector is configured to measure the side scatter by measuring at least a proportion of light scattered within the second cone.

Images