WO1997002482A1

WO1997002482A1 - Volumetric cell quantification method and system

Info

Publication number: WO1997002482A1
Application number: PCT/US1996/010977
Authority: WO
Inventors: Sam L. Woo; Thomas M. Baer; Joseph E. Baumgartner; Louis J. Dietz
Original assignee: Biometric Imaging, Inc.
Priority date: 1995-06-30
Filing date: 1996-06-27
Publication date: 1997-01-23
Also published as: AU6542196A

Abstract

Enumeration of white blood cells (WBC) in leukocyte reduced apheresis products using a red helium neon laser to induce fluorescence in intercalating dye-labeled nucleated cells in samples loaded in precision rectangular capillaries. Apheresis samples are incubated with a surfactant which allows the intercalating dye to enter the WBC. A scanning instrument scans, identifies and enumerates the WBC in the apheresis sample. The system uses an adaptive intensity threshold to identify target fluorescent particles. Reagents are provided for lysing platelets or red blood cells in the apheresis product. A quaternary ammonium salt is added to prevent aggregation. The sample is presented to the scanning instrument in rectangular glass capillaries having a depth of 300 microns. The reagents and cartridge are provided in a kit form.

Description

TITLE VOLUMETRIC CELL QUANTIFICATION METHOD AND SYSTEM

CROSS-REFERENCES TO RELATED APPLICATIONS

This application is a continuation in part of patent application Serial No. 08/236,645 filed on May 2, 1994, which is a continuation in part of application Serial No. 08/018,762, filed on 2/17/93, now abandoned. Each of those applications are hereby incorporated by reference.

BACKGROUND OF THE INVENTION

The present invention relates generally to the field of chemical and microbiological assay techniques and more specifi- cally to a new and improved method of preparing an apheresis sample for cell classification and enumeration.

A major contaminant of blood derived transfusion products (packed red blood cells and platelet apheresis) is leukocytes or white blood cells (WBC) from the donor. The donor WBC may be contaminated with virus, such as cytomegalovirus and human immunodeficiency virus (HIV) , or the WBC may stimulate an immune response from the recipient. One method of reducing such effects is to use WBC reduction techniques, for example, cell filtering. Reduction techniques can reduce the concentration of WBC down to less than one cell per microliter. Detecting and monitoring leuko-reduced blood products for WBC contamination at such low levels is a pressing need for blood banks and the like.

Two techniques have been developed to monitor WBC at such low levels. The first technique is a laborious manual method using a Nagoette hemocytometer to count the number of WBC in a known volume. In this method, the WBC are stained using a standard commercial stain called "Turk's solution." The sample is loaded into a Nagoette hemocytometer that can hold about one hundred microliters of sample. An operator scans the hemocytometer using a standard microscope and counts the number of WBC observed in a known volume. This process takes about thirty minutes to an hour and is a very tedious method.

The second method uses a flow cytometer to count the number of WBC in the sample. In the flow cytometer method, a known number of fluorescent beads or nucleated red blood cells from a chicken are added to a measured volume of sample. A lysing agent is added to the sample to make the WBC permeable to a fluorescent intercalating dye, which is also added to the sample. The intercalating dye preferentially binds to the deoxyribonucleic acid (DNA) in the nucleus of a WBC. The dye becomes much more fluorescent after binding to the nucleus. When excited by a laser, the WBC fluoresce significantly brighter than the solution of unbound intercalating dye (background) .

The sample is loaded into a standard flow cytometer and the number of WBC are measured relative to the number of fluo¬ rescing beads (or chicken red blood cells) . Knowing the ori¬ ginal concentration of the beads (or RBC) allows calculation of the original concentration of the WBC in the sample from the formula: Sample WBC concentration = [Measured number of WBC1 * [Initial cone, of beads (or RBC) 1

Measured number of beads (or RBC)

The flow cytometry method is quite sensitive and accurate, but it requires substantial sample preparation, uses a sophisticated cytometry device requiring a highly trained operator and is relatively time consuming. Thus, there is a need for a simple, fast, inexpensive method for enumeration of WBC in blood products.

With the development of light detection devices capable of making highly accurate quantitative measurements of fluores¬ cent intensity, a new potential for automated cell enumeration devices emerged. Flow cytometers were developed with fluores¬ cent sensors to detect fluorescent emission. Scanning devices were developed to scan microscope slides and automatically identify and enumerate target blood cells.

The development of such new instruments for automated fluorescent analysis has resulted in a demand for new and improved sample preparation and presentation techniques that simplify greatly cell enumeration without jeopardizing accur- acy. The general purpose of such techniques is to expand analytical capabilities, improve reliability, simplify prepara¬ tion, minimize handling of samples, and reduce the risk of disease transmission during sample handling. Flow cytometers have the advantage of making rapid and accurate cell enumeration of different cells in a sample, but do not make direct accurate enumeration of cells per volume of whole blood. Flow cytometers present a sample of cells or particles before a light source in a linear flow path to meas- ure the interaction of the laser with each cell or particle. The flow path consists of a downward flowing stream of liquid into which the cells are released one at a time into the center of the flow path.

There is a need for volumetric enumeration of apheresis samples, without the problems associated with flow cytometers. As discussed herein, one known technique is to count all of the cells with a device such as a cell sorter to enumerate the total number of a certain subset of cells in a fixed volume of sample. Another known system for cell enumeration is fluores- cent microscopy, which combines fluorescent labeling with microscopy technology. Since such systems use a microscope slide for sample interrogation, low event targets in large volume apheresis samples cannot be readily determined.

Thus, there is a need for a reliable, cost and time effective method and system for quantification of leukocytes (WBC) in apheresis products, such as platelet apheresis and red blood cell apheresis products.

SUMMARY OF INVENTION The present invention provides a method and system for performing a sensitive and accurate white blood cell (WBC) enumeration in apheresis blood products. The present invention specifically discloses the assay chemistry, including reagents in a kit form, and procedure for low event leukocyte enumera- tion in platelet apheresis. The basic data processing steps, including sample interrogation, fluorescence emission data sampling, particle identification, cell classification and WBC enumeration are disclosed. This invention may also be useful for WBC enumeration in packed red blood cell products.

The present invention is unique in that it provides a method and system for enumerating WBC using intercalating dyes in a volumetric format. In addition, a novel combination of assay reagents are used to lyse and permeabilize the cells, prevent clotting, and assisting the staining of the WBC DNA. In addition, the cell recognition processor automatically sets certain gates for detecting the nucleated cells in the sample. Similarly, an adaptive threshold is used to follow changes in dye concentration and baseline (background) fluorescence. Moreover, to increase sample size, a larger capillary is used, wherein the scanning instrument is focused on the lower portion of the capillary.

In the method of the present invention, the WBC are stained with an intercalating dye. After incubation, the absolute number of cells are counted in a known volume. The leukocyte depletion assay labels nucleated cells which are placed within a scan capillary, and allowed to settle. The capillary is scanned with a helium-neon (He-Ne) laser and the fluorescent response of the WBC in the capillary is measured.

The cells are recognized from the fluorescent response using a cell identification algorithm.

The present invention includes an apparatus having a laser based optics module for scanning a capillary of an apher- esis sample containing an unknown amount of WBC. Data sampling circuitry is provided to record fluorescent responses of the sample in the capillary. According to one aspect, the scanner and data sampling circuitry produce a plurality of channels of data, and a corresponding plurality of scanned images. A processing system is coupled to the sampling circuitry, and includes resources to identify and enumerate the WBC which fluoresce in response to excitation by the laser. A cartridge is provided for presenting a plurality of scan capillaries to the scanning instrument. The cartridges are configured with application wells and pedestals for moving the apheresis sample into the scan capillaries for processing a known volume of fluorescently stained sample. The intensity of the fluorescence of a target cell is determined by processing certain identified segments of data from the plurality of fluorescent responses based upon the expected characteristics of target WBC. For example, the segments are processed by defining a neighborhood of pixels for a candidate particle, wherein the neighborhood is larger than the expected size of the target cell. The pixels within the neighborhood are processed to compensate for background fluo¬ rescence and noise. An adaptive threshold on the fluorescence from the candidate particle is used to compensate for changes in dye concentration and baseline fluorescence.

Generally, an assay is disclosed for enumerating target components in an apheresis sample in a fixed volume. The apheresis sample is stained with a fluorescent compound configured to selectively bind to a target component of the apheresis sample. A laser in the sample instrument is used to excite the fluorescent compound, while the data processor identifies, classifies and enumerates target leukocytes.

In particular, the present invention provides for an assay for enumerating leukocytes in a platelet apheresis sample in a fixed volume capillary. A first reagent is added to the apheresis sample which is configured to stain DNA in leukocytes in the apheresis sample. The first reagent includes a fluores¬ cent intercalating dye, such as a cyanine nucleic acid dye. A second reagent is added to the apheresis sample to lyse the platelets, for example, polyoxyethylene ether. In addition, a quaternary ammonium salt, such as dodecyltrimethylammonium bromide is added to the apheresis sample to prevent the lysed platelet particles from aggregating. The present invention includes a system for analysis of an apheresis sample, which uses a scanner configured to interrogate a sample with a light source to generate fluores¬ cent emission data. A data processor is used to sample the fluorescent emission data to generate a plurality of pixel values. The data processor also analyzes the pixel values to identify a candidate particle and a pixel neighborhood. The fluorescence values for the pixel neighborhood are processed to classify the candidate particle as a leukocyte. WBC enumera- tion includes comparing the fluorescent emission data to an adaptive intensity threshold and a slope gate.

A kit is disclosed herein for use in an assay for enumerating leukocytes in a platelet apheresis sample. The kit includes TO-PRO-3 intercalating dye configured to stain DNA in WBC. A known amount of TRITON X-100 is included to lyse the platelets, while a quaternary ammonium salt is added to prevent the lysed platelets from aggregating. The reagents are provided in a container, such as a vial, for incubating an apheresis sample containing leukocytes and platelets with the first, second and third reagents. Further, a cartridge having at least one scan capillary may be included in the kit.

These and other features and advantages of the present invention will become apparent from the following more detailed description, when taken in conjunction with the accompanying drawings which illustrate, by way of example, the principles of the present invention.

BRIEF DESCRIPTION OF THE DRAWING FIGURE 1 is a system diagram of an optical scanner for use with the invention.

FIG. 2 is a side perspective view of a sample-filled capillary tube according to the invention, illustrating an illuminated columnar region from the excitation and emission beams, showing the scanner in schematic representation.

FIG. 3 is a side perspective view of a sample-filled capillary tube showing overlapping beam spots and an illumi¬ nated columnar region according to the invention.

FIG. 4 is a top perspective view of a sample-filled capillary tube, showing overlapping beam spots according to the invention.

FIG. 5 is a schematic representation of the optical scanning path according to the invention.

FIG. 6 is a schematic representation of a labeled cell suspension and the corresponding detector signal.

FIG. 7 is a perspective view in partial cross-section of a sample-filled capillary showing suspended dye particles within the apheresis sample, wherein the white blood cells have been allowed to settle to the bottom of the capillary.

FIG. 8 is an enlarged view of the portion in circle "8" of FIG. 7.

FIG. 9 schematically illustrates an organization of fluo- rescent data pixels into a neighborhood generated by sampling and storing the output of the scanner of FIG. 2.

FIG. 10 is a graph illustrating the emission spectra of

TO-PRO-3 intercalating dye and the transmission spectrum of a dichroic mirror which divides the emission spectra of the dye. FIG. 11 illustrates a cell classification graph used for classifying cells based on two channels of data.

FIGS. 12B and 12B together make up a flow chart for the basic data processing method.

FIG. 13 is a flow chart illustrating the generation of background indices.

FIG. 14 is a flow chart of the process used for detecting particles from the fluorescent data.

FIG. 14A is a pixel diagram showing the perimeter and adjacent pixel comparison of FIG. 14. FIG. 15 is a flow chart for the process used for classifying cells.

FIG. 16 is a graph showing the relationship of interca¬ lating dye concentration to the fluorescent units exhibited by a target cell and by the background fluorescence from a platelet apheresis sample.

FIG. 17 is a graphical representation fluorescent units exhibited for the background and a target cell for a plurality of assay configurations.

FIG. 18 is a perspective view of an assembled assay cartridge constructed with dual capillaries.

FIG. 19 is an exploded perspective view of the assay cartridge of FIG. 18.

FIG. 20 is a perspective view of an assembled assay cartridge constructed with four capillaries. FIG. 21 is an exploded perspective view of the assay cartridge of FIG. 20. DESCRIPTION OF THE PREFERRED EMBODIMENTS

As shown in the exemplary drawings, the present invention is embodied in a method and system for performing sensitive and accurate white blood cell (WBC) enumeration in apheresis blood products. The present invention includes the assay chemistry, including reagents in a kit form, and procedure for low event leukocyte enumeration in platelet apheresis. The basic data processing steps, including sample interrogation, fluorescence emission data sampling, particle identification, cell classification and WBC enumeration are disclosed. This invention may also be useful for WBC enumeration in packed red blood cell products.

With reference to FIG. 1, the present invention includes an optics module or scanner 25 including a laser 27, such as a helium neon (He-Ne) laser, which generates an excitation beam 28 that impinges on a glass plate 29. The glass plate reflects a significant portion of the laser beam to a power monitor 30 which measures the power output of the laser. The portion of the excitation beam 32 that is passed through, rather than reflected by, the glass plate is directed through a laser line filter 33 and then to a first spectral dispersion device 34, which may be, for example, a dichroic mirror, a prism, or a grating. The spectral dispersion device reflects the excitation beam which, in the case of a He-Ne laser, has a wavelength of approximately 633 nanometers. The excitation beam is then directed to a first mirror 35 and through a right angle prism 36 to a scan assembly 40.

The scan assembly 40 comprises a galvanometer 41 attached to a galvo mirror 42, a first lens 43, a second lens 44, and an objective lens 45. Within the scan assembly, the passed portion of the excitation beam 32 travels from the galvo mirror through the first lens and then through the second lens. From the second lens, the excitation beam is directed through the objective lens so that a focal spot of the beam may impinge upon a transparent capillary tube 50, having an upper surface 51. The capillary contains the sample to be analyzed, which includes target WBC in platelet apheresis mixed with an excess of an intercalating dye used to stain the WBC. The excitation beam 32 impinging upon the upper surface 51 of the capillary 50 traverses the wall and illuminates a region of the sample causing fluorescent emission from the sample. Light collection occurs in an epi-illumination manner. The emitted fluorescence is collected by objective lens 45 and directed back, as retrobeam 53, through scan assembly 40. As shown in FIG. 2, the objective lens has a central portion for passage of the excitation beam and uniform depth of focus of the excitation beam through bottom one-third of the capillary tube. Because fluorescent emission is over a very wide angle, represented by retrobeam rays 54 and 54, fluorescent collection occurs over a wider portion of the objective lens.

Referring now more particularly to FIG. 1, the retrobeam 53 travels from scan assembly 40 to the right angle prism 36 to the first mirror 35 and the first spectral dispersion device 34. The retrobeam has a spectral range which encompasses different wavelength than the excitation beam 28 and may there¬ fore be directed differently than the excitation beam by the same optical components that direct the excitation beam to the scan assembly. Thus, the retrobeam is transmitted through the first spectral dispersion device and through a bandpass filter 61 to a second mirror 62 where it is directed through a colli¬ mating lens 63. The retrobeam is then selectively passed through a spatial filter 64 and into a detection assembly 70. The spatial filter has a predetermined pinhole aperture of a diameter that permits passage of only that fluorescence emission from a columnar region 60 defined by the illuminated segment within the capillary tube 50.

Any light having the wavelength of the excitation beam 32 will be removed from the retrobeam 53, for example,- light from the excitation beam reflected by the upper surface 51 of the capillary 50 along the path of the retrobeam. Such light will either be directed by the first spectral dispersion device 34 away from the path of the retrobeam and back toward the laser 27 or will be removed by the band bass filter 61. The retrobeam is thereby purged of reflected excitation light. The retrobeam is focused by the collimating lens 53 through the pinhole aperture of the spatial filter 54, which acts to eliminate from the retrobeam any light other than fluorescent emissions from a defined region 60 of the sample within the capillary.

In the preferred embodiment, the detection assembly 70 comprises two detection channels, each having a detector 71, 72 which amplifies the fluorescent signal of the retrobeam 53. The detector is a light measuring device, such as a photomulti- plier tube (PMT) or photodiode. A second spectral dispersion device 73 is positioned between the spatial filter 64 and the detectors to split the retrobeam as a function of its wave¬ length. A first beam 77 of one range of wavelength is impinged upon the first detector 71/ and a second beam 78, having another wavelength range, is impinged upon the second detector 72. Each detector is in electronic communication with a data processor 75, which converts analog signals from the detectors to digital data representative of the retrobeam' s fluorescence intensity.

A critical feature of the present invention is illus¬ trated in FIG. 2. The spatial filter 64 is configured with a pinhole aperture that collects light over a large numerical aperture, but confines the depth of detection to a defined portion 60 of the interior depth dimension of the capillary tube 50. The objective lens 45 has been configured to focus the excitation laser beam 32 to provide a uniform illumination waist or columnar region 81 along the lower one-third (one hundred microns) of the capillary tube. Similarly, only the fluorescence from the sample in the bottom one-third of the capillary is focused from the retrobeam 53 and impinged upon the detectors 71, 72. The capillary tube 50 is a transparent sample holder of known dimensions, preferably made of optics quality glass or acrylic. The capillary tube preferably has a rectangular cross-section with a shorter dimension defining an interior depth of two hundred to four hundred microns and a longer dimension defining a width of one to three millimeters. The length of the capillary tube is chosen to provide a desired volume of the sample to be processed. The beginning and ending longitudinal points of the scan in a direction along the length of the capillary tube define the precise volume of the sample scanned.

In the present invention, it is important to use a sufficiently large sample volume to ensure detection of WBC in sufficient number to make the results statistically signifi¬ cant. For a platelet apheresis sample, a capillary being three hundred microns deep, three millimeters wide and forty milli¬ meters long provides optimal scanning characteristics. The diameter of the columnar region 81 defined by the excitation beam 32 is preferably ten to fifteen microns, or about the same size as a target cell. The cross section of the columnar region or waist is referred to herein as the "beam spot."

As shown in FIG. 3, after a columnar region 81 of the capillary tube is illuminated and the fluorescence emitted from its contents is detected and recorded, the excitation beam spot 80 (shown for illustration purposes along the upper surface 51 of the capillary 50) is moved to a new position to illuminate an adjacent columnar region 82 at the lower one-third of the capillary. The excitation columnar region is moved a distance that is only a fraction of the beam spot diameter, so that each illuminated columnar region 81 partially overlaps adjacent illuminated columnar region 82. The optical scanning of the capillary continues in this manner of illuminating and fluores- cently exciting a region from which fluorescent emission is detected and recorded, then is moved slightly to illuminate a new columnar region and to repeat the data collection process.

The scanning assembly 40 moves the excitation beam spot

80 along a scan path, first in a direction transverse to the longitudinal axis (arrow 85) of the capillary tube 50, i.e. along its width, and then in a direction along the length or longitudinal axis of the capillary (arrow 86) . Such a sideways and lengthwise scanning protocol provides a two-dimensional array of beam spots and corresponding data readings from the detectors 71, 72. As shown in FIG. 4, the transverse scan begins at point 88 inside a first edge 91 of the upper surface 51 of the capillary and ends at a point 89 inside the second edge 92 of the capillary upper surface.

FIG. 5 shows a schematic representation of scan path 95 from above capillary tube 50, according to the preferred embod¬ iment of the present invention. The excitation beam spots are moved along the scan path in a transverse direction from inside the first edge 91 to inside the second edge 92, then returned to the first edge to follow a closely-spaced and overlapping parallel path in the transverse direction. In the preferred embodiment, the scanning starts and ends one millimeter inside the capillary edges. The scanning process is repeated contiguously so that the scan covers a desired distance along the longitudinal axis, of the capillary tube. In this manner, fluorescence emission occurs and is detected from any chosen length of the capillary tube. The volume of the sample scanned is determined by multiplying the depth of the capillary times the distance of the transverse scans (point 88 to 89) times the longitudinal distance scanned.

As shown in Fig. 6, the fluorescent signal from the detectors 71, 72 and stored by the data processor 75 may be represented graphically in a plot 96 of fluorescent units (Y- axis) versus time (X-axis) . The background level 97 of fluorescence is caused by unbound dye 101 in the sample. This background level will not be constant and is disturbed by artifacts, such as clumped platelets and uneven dispersion of the dye particles. Each WBC 100 is represented by a fluores¬ cent event 98. The method of the present invention does not require removal of unbound intercalating dye. As explained in more detail herein, the assay chemistry of the present inven¬ tion optimizes the ratio of the peak fluorescence from a cellular event versus the background level fluorescence.

Referring now to FIG. 7 and in more detail to FIG. 8, the capillary 50 is shown with a partial cutaway of the capillary wall 52, demonstrating the location of the target WBC 100 in the sample of platelet apheresis 102. The capillary is shown in a stasis preferred just prior to interrogation by the scanner 25. The sample has been permitted to stand after vortexing, allowing the heavier WBC to settle near the bottom 103 of the capillary. The unbound dye particles 101 remain suspended in the platelet apheresis within the volume of the capillary. Thus, the optics module 25 allows transverse and longitudinal scanning of the laser beam 32 across the sample capillary 50. The scan assembly 40 includes a galvanometer 41 driven mirror 42 which rotates a few degrees back and forth in a rapid fashion at about 20 - 200 Hz (peak-to-peak 6-12°) . The excitation beam is deflected by the galvanometer mounted mirror to the first lens 43, through the second lens 44 and through the objective lens 45 to the upper surface 51 of the capillary.

The two scan assembly lenses 43, 44 are configured with equal focal lengths and are separated along the optical path by twice their focal length. In addition, the lenses have overlapping focal planes. Similarly, the distance between the second lens 44 and the objective lens 45 must be precisely controlled so that the excitation beam 32, as it is displaced by the galvanometer driven mirror 42, appears to be rotating about a virtual point directly in front of the scanning assembly objective.

For a capillary 50 of about three millimeters in width, and a beam spot 80 having a diameter of about ten microns, the transverse scan path 95 is about one millimeter long. This provides two hundred-fifty samples along each scan path. For a forty millimeter long capillary, with scan lines separated by four microns, about ten thousand scan lines are collected for each apheresis sample. The data processor 75 is configured to sample and store in data memory each fluorescent data value (pixel 105) from the detectors 71, 72 at a rate which creates a data sampling array representing a four micron transverse or longitudinal displace¬ ment of the excitation beam spot 80. FIG. '9 illustrates a seven-by-seven array of pixels, referred to herein as a

"neighborhood" 110. Thus, in the upper left hand corner, pixel of row 1, column 1 is found. In the upper right hand corner, pixel of row 1, column 7 is found. In the center of the neighborhood of pixels, pixel of row 4, column 4 is found. Similarly, in the lower right hand corner, pixel of row 7, column 7 is found.

FIG. 9 also illustrates the size of the excitation laser beam spot 80 relative to the pixel 105 dimensions, wherein oversampling occurs. In the preferred embodiment, the beam spot has a diameter of about ten microns. For example, the first excitation beam spot 111 excites a region of ten microns in diameter for the pixel at row 7, column 1. At row 6, column 1, a second beam spot 112 illuminates a region ten microns in diameter which substantially overlaps with the first beam spot for row 7, column 1. Similarly, the adjacent third beam spot 113 for row 7, column 2 substantially overlaps with the first beam spot and the second beam spot in column 1. Referring to FIG. 1 and more particularly to FIG. 10, retrobeam 83 is comprised of fluorescence from the bound and unbound intercalating dye in the illuminated sample. The retrobeam is directed to a spectral dispersion device 73 which separates the fluorescence into two beams. The first beam 77 contains fluorescent emissions with wavelengths less than the cutoff point of the spectral dispersion device. The second beam 78 contains fluorescent emissions with wavelengths greater than the cutoff point of the spectral dispersion device. The first resultant beam is directed to the first detector 71 (channel 0) , while the second resultant beam is directed to the second detector 72 (channel 1) . A portion of the fluorescence from the dye will be directed into each of the channels. FIG. 10 depicts the cutoff curve 115 for a dichroic mirror which splits the fluorescent wavelengths at approximately 684 nano- meters.

Suitable intercalating dyes include monomeric cell- impermeant nucleic acid stains. Especially useful are membrane-impermeant cyanine nucleic acid dyes which are excitable by a He-Ne laser at about 633 nanometers. One such intercalating dye is a benzothiazolium-4-quinolinium dye available under the trade name TO-PRO-3 from Molecular Probes, Eugene, OR. TO-PRO-3 has an absorption maximum at 642 nano¬ meters and an emission maximum of 661 nanometers. TO-PRO-3 has an emission spectrum from about 600 nanometers to about 800 nanometers, as shown by line 117 in FIG. 10.

FIG. 11 represents a graph of fluorescent responses for WBC 100 and other sample artifacts 120. Because of the significantly different fluorescent responses between the nucleated WBC and other sample artifacts, such as clumped dye and platelet particles, the WBC may be differentiated by comparing the absolute and relative responses from the two detectors (PMT) 71, 72. One differentiation is based comparing the fluorescent intensity from the first detector I₀ (x-axis) to a minimum adaptive intensity threshold 122. The adaptive threshold is derived from the fluorescent intensity value of the baseline or background value, which is a function of dye concentration. See FIGS. 16 and 17 and associated text herein. Any fluorescent particle or artifact which does not generate the minimum fluorescent response relative to the background intensity is disregarded, and is not counted as a WBC.

Another differentiation of the raw fluorescent data is based on comparing the fluorescent intensity from the first detector I₀ relative to the fluorescent intensity from the second detector Ii- This comparison is accomplished by comparing the ratio of the intensities to a empirically determined calibration slope 124. In addition, a high tolerance slope 125 and a low tolerance slope 126 around the calibration slope brackets the acceptable range of fluorescent values from the two detectors for discriminating a WBC from other fluorescent artifacts in the sample. Any fluorescent particle or artifact which does not generate relative values of the channel 0 intensity I₀ and channel 1 intensity I within the tolerance around the calibration slope is disregarded, and is not counted as a WBC.

Once the scan capillary 50 has been interrogated by the scanner 25 and the fluorescent response of each excitation beam spot 80 stored by the data processor 75, the data processor initiates WBC identification and enumeration, FIGS. 12A and 12B. Cell identification is performed by software programming stored in the memory of the data processor. WBC discrimination includes analysis of the fluorescent data derived from the detection assembly 70. The linear regression technique used to discriminate the fluorescent data stored for the two channels is described in copending U.S. application serial number 08/236,645 (hereinafter "Sitzo et al. ") and is incorporated herein by reference. In particular, cell classification analysis is performed on a seven-by-seven neighborhood of pixels centered on each target cell which is detected using the method of FIGS. 12A and 12B. FIG. 13 illustrates a method for processing background noise indices to define the tolerances 125, 126 around the calibration slope 124 in FIG. 11 within which the fluorescent response measured for a neighborhood will be characterized as a WBC. This background index is recalculated for each neighborhood (block 209) . The process is carried out for the data from both detectors 71, 72 (channels) to achieve two separate background noise indices.

FIG. 14 illustrates the algorithm for detecting a particle (block 211) , which may be a candidate cell for enumeration. The particle detection algorithm uses the average or sum of the fluorescent data stored for each pixel (block 400) . A five-by-five pixel map (block 401) surrounding each pixel is tested (block 402) . The test ensures that the fluo¬ rescent values for the particle match the expected intensity profile for a WBC, and are not indicative of noise or other non-WBC artifacts.

Referring to FIG. 14A, the cell detection test first determines whether the center pixel 410 fluorescent intensity value minus an adaptive intensity threshold value is greater than each of the perimeter pixel 412 fluorescent intensity values. The test also determines whether the absolute value of the center pixel fluorescent intensity value is greater than each of the adjacent pixel 414 values. If either of these tests are negative, i.e., center pixel not greater, then the particle is disregarded and not counted as a WBC. If both tests are satisfied (block 403) , then the location of the center pixel in the data array (capillary position) is recorded. In addition, a seven-by-seven pixel map, or neighborhood, surrounding the center pixel is saved

(block 213) . These particle values are be used for further processing of the data and ultimate WBC classification and enumeration.

FIG. 15 illustrates the enhanced method for processing WBC fluorescent data in an apheresis sample. Cell classifica- tion according to a preferred embodiment of this invention is similar to that found in Sizto et al. , though it differs in several material ways. The method includes retrieving (block 500) the seven-by-seven neighborhood of fluorescent data values stored for each particle, as described herein (block 213) . Additional parameters are extracted from the particle neighbor¬ hood which describe the particle in geometric or morphologic terms, such as, but not limited to, area, radius, circularity, etc. With reference to FIG. 11 and specifically to the steps outlined in FIG. 15, particles are classified as to cell type. First, the coefficient of regression (r) is compared to a threshold (K₀) disclosed in Sizto et al. (block 501) . If the coefficient is not greater than the threshold, then the parti- cle is disregarded and not enumerated as a WBC (block 507) .

If the first criteria is met, then an adaptive intensity threshold 122 based on the intensity of the channel 0 baseline is used to qualify cell intensity (block 502) . This second test reduces or eliminates any sensitivity the method may have to variations in baseline or median particle intensity. Thus, the particle intensity (I₀) from channel 0 (first detector 71) which is above the baseline (background) is compared to an adaptive threshold (K_λ) proportional to the intensity of the channel 0 baseline. If the particle intensity is not greater than the intensity threshold, then the particle is disregarded and not enumerated as a WBC (block 507) .

Continuing with the classification, the method of the present invention uses a symmetric region 125, 126 around the calibration slope 124 to include candidate WBC particles and exclude noise particles (block 503) . In particular, the absolute difference between the channel 1 (second detector 72) particle intensity above baseline (I_x) and the channel 0 particle intensity above baseline (I₀) projected onto channel 1 through the calibration slope (J-) is compared to the adaptive threshold proportional to the mode of the channel 0 baseline variation (K₂) . If the channel 1 brightness is not within the slope gate, then the particle is disregarded and not enumerated as a WBC (block 507) . The last criterion for classification is based on the geometric properties of the candidate particle (block 504) . Specifically, the channel 0 particle area (A₀) must be greater than a first empirical threshold (K₃) and less than a second empirical threshold (K₄) . The channel 0 particle area (A₀) is defined as the number of pixels in the seven-by-seven neighbor¬ hood with channel 0 intensity greater than or equal to a minimum percentage, for example, one-third, of the intensity of the pixel with the greatest intensity in the neighborhood. If the particle area is not within the empirical thresholds, then the particle is disregarded and not enumerated as a WBC (block 507) .

Thus, the identified particle must satisfy four criteria to be classified as a WBC: (block 501) r > K₀ (block 502) I₀ > K_x (block 503) I Ii - j I < K₂ (block 504) K₃ < A₀ < K₄.

As shown in Fig. 16, the fluorescent emission of a WBC stained with a fluorescent intercalating dye may be plotted against the concentration of the dye in the initial incubating solution. Most cyanine nucleic acid stains will exhibit a fluorescent curve 130 which has a high-noise-ratio region 132 followed by a linear region 134. The linear region ends at a non-linear region 136 finalizing in a flattened saturation portion 138 of the curve.

An aspect of the present invention is to use the interca¬ lating dye in a concentration in the linear region of the curve. Using a dye concentration in the linear range provides a significant difference between the fluorescent units exhib¬ ited by the white blood cell stained with the intercalating dye and the corresponding background fluorescence 140 exhibited by the unbound intercalating dye in the solution. The background fluorescence exhibits a linear relationship to the initial dye concentration in the incubating solution. It has been found for certain cyanine nucleic acid stains, such as TO-PRO-3 (benzothiazolium-4-quinolinium) , that a concentration of 0.2 micromolar to 1.0 provides a preferred ratio of WBC fluorescence to background fluorescence.

As shown in Fig. 17, it is critical that the target cell fluorescence significantly stands out over (greater magnitude) the background fluorescence. Four curves are shown to demonstrate the relative difference between the background fluorescence and the peak fluorescence from a WBC stained with an intercalating dye. For any particular scanning instrument, the preferred difference is a function of the scanning optics and cell identification software. For the present invention, a ratio of about 5:1 is preferable. In addition to manipulation of the percentage dye concentration in the incubating solution, the depth of the capillary 50 may be configured to adjust the ratio of target cell fluorescence to background fluorescence.

Curve A exemplifies the background fluorescence 141 at a very low level, while the target cell fluorescence 142 is also at a low level. Such a combination is unacceptable, because there is an insufficient difference between the target cell fluorescence and the background fluorescence. Moreover, there is a high noise-to-signal ratio, wherein the noise (shot, thermal, etc.) dominates the fluorescent signal from the background, blurring the distinction between the target cell fluorescence and the background fluorescence. Curve A is indicative of too low a dye concentration in the incubating solution.

Curve B demonstrates the condition where the dye concen¬ tration reaches the saturation point 138 for maximum WBC peak fluorescence 144. The corresponding background fluorescence level 145, however, is too high to readily distinguish the WBC peak fluorescence. The ratio of the target cell fluorescence to background fluorescence is too low. Curve B is indicative of excessively high dye concentration in the incubating solution. Curve C demonstrates the condition where the dye concentration approaches the saturation point 138 for maximum WBC peak fluorescence 144, but remains in the linear range 134. Here, the peak fluorescence 146 of a target cell is near its optimum point, but the background fluorescence 145 is unaccept- ably high with respect to the peak fluorescence. Since the ratio of the target cell fluorescence to background fluores¬ cence is too low, the background fluorescence level is too high to readily distinguish the WBC peak fluorescence. Curve C is indicative of configuring the capillary 50 too deep for the concentration of intercalating dye used in the incubating solution.

Curve D demonstrates the condition where the incubating solution dye concentration is properly chosen in the linear range 134, and the depth 58 of the capillary 50 is chosen to optimize the background fluorescence 147. Preferably, the peak fluorescence 148 of a target cell is near its optimum point, while the background fluorescence is sufficiently low. Conse- quently, the incubating solution dye concentration and capil¬ lary depth are configured to provide a WBC peak fluorescence near the saturation level, while exhibiting a background fluo¬ rescence level about one fifth the peak fluorescence. In the present invention, using a three hundred micron deep capillary and a TO-PRO-3 dye concentration of about 0.9 micromolar provides an approximate ratio of WBC fluorescence to background fluorescence of 5:1.

An unexpected problem was solved by the present inven¬ tion. It was known that unbound (free) cyanine nucleic acid dye in a whole blood or lysed blood solution will fluoresce at a relatively known value. It was also known that when the dye is intercalated into the DNA of a white blood cell the result¬ ing fluorescence value is about two hundred times that of the unbound dye. It was an unexpected result that, when the cyan- ine nucleic acid intercalating dye was mixed in a solution of lysed platelets, the background level of fluorescence was five times higher than the background fluorescence in a solution containing lysed red blood cells and having the same dye concentration. Because of the unexpectedly high background level in the platelet apheresis solution, it was critically important to determine and use a concentration of the TO-PRO-3 dye in the linear range 134 of fluorescence. Likewise, it was critically important to determine the appropriate depth 58 of the scan capillary 50 which would provide a about 5:1 ratio of peak target WBC fluorescence to background fluorescence. Appar¬ ently, the problem of the unexpectedly high background fluores- cence in an apheresis platelet solution was not known to those of ordinary skill in the art of WBC enumeration. Moreover, the present invention of mixing the intercalating dye concentration to provide a fluorescence of a WBC in the linear range, while configuring a optimum depth capillary to provide the desired ratio of peak to background fluorescence was certainly not known to those of ordinary skill in the art .

Another aspect of the present invention is to enhance the activity of the intercalating dye by permeabilizing the target cell membrane with a surfactant, detergent or other mild lysing agent. One such compound is polyoxyethylene ether, available under the trade name TRITON X-100 from Bio-Rad, Richmond, CA. The lysing agent is used in sufficient concentration in the sample to adequately lyse the platelets. This allows the target WBC to settle to the bottom of the capillary, within the focal range of the scanning instrument (one hundred microns) . In addition, platelet lysing reduces fluorescent scatter and other optical interference from the platelets. The lysing agent, however, should be selected and provided in a concentration that will not lyse the target cells, here the WBC.

One of the aspects of the present invention is to con¬ figure the concentration of the sample chemistry, including the concentration of TRITON X-100, such that after the period of incubation the cell membranes of the WBC are sufficiently permeabilized or weakened to allow the intercalating dye TO- PRO-3 to enter the WBC and stain the cell DNA. The optimum concentration range of TRITON X-100 is 50.0 millimolar (mM) to 200.0 mM, and preferably 90.3 mM. At such a concentration, the WBC are sufficiently permeabilized, but will not lyse for at least four hours. If the lysing agent is too strong, activity or concentration, or the WBC are left in a weak solution too long, then the WBC will lyse, defeating the enumeration process. Another aspect of the assay chemistry of the present invention is to add a quaternary ammonium salt to the sample in sufficient concentration to prevent agglutination or similar aggregation of platelet particles after lysing. One such quaternary ammonium salt having suitable surface active properties is dodecyltrimethylammonium bromide, available from Aldrich, Milwaukee, WI . In addition, the quaternary ammonium salt may be dissolved in an isotonic diluent, for example, dimethyl sulfoxide (DMSO) also available from Aldrich, Milwaukee, WI .

Another aspect of the present invention is to provide the reagent chemistry in a kit, including a dispensing vial. For example, the intercalating dye (TO-PRO-3) , detergent (TRITON X- 100) , quaternary ammonium salt (dodecyltrimethylammonium bromide) and diluent (DMSO) may all be provided in sufficient concentration such that when a known amount of platelet apher¬ esis sample is added to a vial containing the chemistry, the reagents are all in their optimum concentrations. Further, a kit may comprise the reagent chemistry and a sample cartridge having at least one scan capillary and application well. Similarly, a pipette may be supplied with the kit to remove a desired amount of the platelet apheresis sample from its container and to dispense the reagent-apheresis mixture to the application well of the cartridge. In accordance with the present invention and as shown in FIGS. 18-21, a cartridge 150 is provided to present a bio¬ logical sample for analysis by an imaging or scanning instru¬ ment. The cartridge uses one or more glass or acrylic capil¬ laries 160 and adjacent application wells 165 to present an aliquot of incubated sample for interrogation by the instrument optics. Since a predetermined aliquot of sample is drawn from the application well by each capillary, there is no need to meter the sample into the application wells. Thus, a practical and cost effective cartridge and assay process is provided which eliminates the need for complicated and costly fluidics frequently found in prior art cartridges. Such a cartridge is especially useful with fixed volume assays.

Referring now more particularly to FIGS. 19 and 21, the cartridge 150 comprises three molded plates 152, 154 and 156, preferably made of a plastic or the like, such as acrylonitrile butadiene styrene (ABS) , polystyrene or polymethyl methacry¬ late. ABS suitable for manufacture of a cartridge incorpor- ating the present invention may be purchased from BASF Corp. of Wyandotte, MI under the trademark "TERLUX 2802 TR. " The ABS plates are fused together, preferably by ultrasonic welding, as shown in FIGS. 18 and 20. The cartridge may be configured with a plurality of scan capillaries, for example, two capillaries (FIGS. 18, 19) or four capillaries (FIGS. 20, 21) .

The top plate 152 is configured with a groove or slot 162 to hold each capillary 160. The capillary may be press fit or glued into the slot. An application well 165 is configured adjacent one end of the capillary slot so as to provide a res- ervoir of sample to the capillary. The application well is further configured with a pedestal 158 for directly applying the sample droplet. The sharp edges of the pedestal act as a stop junction, preventing the sample from being drawn under the capillary into the slot. The remainder of the application well forms a moat around the pedestal.

The top plate and middle plate 154 are further configured with registration markers 167 to ensure proper alignment of the cartridge within the scanning instrument. The middle plate is also configured with an arcuate handle portion 170 for ease of insertion of the cartridge into the scanning instrument. All three cartridge plates are preferably configured in a triangular ^• shape so as to fit symmetrically within a round rotatable platter in the scanning instrument . A cartridge with a rectangular or similar shape may be suitable for other instrument configurations.

EXAMPLE

A 25.0 microliter (μL) solution of 0.5 micromolar (μM) TO-PRO-3 (Molecular Probes, Eugene, OR) , 90.3 millimolar (mM) Triton X-100 (Bio-Rad, Richmond, CA) and 63.0 mM dodecyltri¬ methylammonium bromide (Aldrich, Milwaukee, WI) in DMSO (Aldrich, Milwaukee, WI) was mixed thoroughly with 500.0 μL of platelet concentrate obtained from either a Cobe or Fenwal apheresis instrument. A volume of 50.0 μL of the mix was then loaded into a rectangular glass capillary with the internal dimensions of 0.3 millimeters (mm) thick, 3.0 mm wide and 50.0 mm in length that was mounted on an scanning cart-ridge. The cartridge was loaded into a scanning instrument of the present invention and the sample was then analyzed using the method of the present invention. The method included delaying the interrogation of the sample in the scanning capillary for fifteen minutes. The delay permitted the stained cells to settle to the bottom of the capillary. The depth of field of the scanning instrument was configured to be less than the thickness of the capillary, i.e., 0.1 mm. The instrument was focused at a region to best detect the stained cells, i.e., at the bottom one third of the capillary. The capillary was then scanned over three minutes and the WBC counts were obtained.

While several particular forms of invention have been illustrated and described, it will be apparent that various modifications can be made without departing from the spirit and scope of the invention. Accordingly, it is not intended that the invention be limited, except as by the appended claims.

Claims

We claim :

1. An assay for enumerating target components in an apheresis sample, comprising the steps of: providing an apheresis sample in a fixed volume; staining the apheresis sample with a fluorescent compound configured to selectively bind to a target component of the apheresis sample; and exciting the fluorescent compound.

2. An assay for enumerating leukocytes in a blood products apheresis sample, comprising the steps of: providing an apheresis sample in a fixed volume capillary, the apheresis sample containing leukocytes and a second blood constituent other than leukocytes; adding a first reagent to the apheresis sample configured to fluorescently stain DNA in the leukocytes; and adding a second reagent to the apheresis sample to lyse the second blood constituent in the apheresis sample.

3. The assay of claim 2, further comprising adding a third reagent to the apheresis sample to prevent the lysed second blood constituent from aggregating.

4. An assay for enumerating leukocytes in a platelet apheresis sample, comprising the steps of: providing an apheresis sample containing platelets; adding a first reagent to the apheresis sample configured to stain DNA in leukocytes in the apheresis sample; adding a second reagent to the apheresis sample to lyse the platelets; adding a third reagent to the apheresis sample to prevent the lysed platelet particles from aggregating; and placing the apheresis sample in a fixed volume capillary.

5. The assay of claim 4, wherein said adding a first reagent step includes staining the leukocyte DNA with a fluo¬ rescent intercalating dye.

6. The assay of claim 4, wherein said adding a first reagent step includes staining the leukocyte DNA with about 0.5 μM of TO-PRO-3.

7. The assay of claim 5, wherein said adding a second reagent step includes lysing the platelets with polyoxyethylene ether.

8. The assay of claim 6, wherein said adding a second reagent step includes lysing the platelets with about 90 mM TRITON X-100.

9. The assay of claim 7, wherein said adding a third reagent step includes mixing the lysed apheresis sample with a quaternary ammonium salt.

10. The assay of claim 8, wherein said adding a third reagent step includes mixing the lysed apheresis sample with about 60 mM dodecyltrimethylammonium bromide.

11. The assay of claim 4, wherein said placing the apheresis sample step includes providing a capillary about three hundred microns deep, about forty millimeters long and at least one millimeter wide.

12. A system for analysis of an apheresis sample, said system comprising: a scanner configured to interrogate a sample with a light source to generate fluorescent emission data; sampling means for processing the fluorescent emission data to generate a plurality of pixel values; particle detection means for processing the pixel values to identify a candidate particle and a pixel neighborhood; and enumeration means for processing the pixel neighborhood to classify the candidate particle as a leukocyte.

13. The system of claim 12, wherein said scanner includes a laser which emits light having a wavelength of about 630 to 650 nanometers.

14. The system of claim 13, wherein said sampling means includes a dichroic mirror configured to transmit a fluorescent emission and impinge the fluorescent emission onto a first a photomultiplier tube and a second photomultiplier tube.

15. The system of claim 12, wherein said particle detection means includes means for comparing the fluorescent emission data to an adaptive intensity threshold.

16. The system of claim 15, wherein said enumeration means includes means for comparing the fluorescent emission data to an adaptive intensity threshold.

17. The system of claim 16, wherein said enumeration means includes means for comparing the fluorescent emission data to a slope gate.

18. A kit for use in an assay for enumerating leukocytes in a platelet apheresis sample, said kit comprising: a first reagent configured to stain DNA in leukocytes; a second reagent configured to lyse the platelets; a third reagent configured to prevent the lysed platelets from aggregating; and a container for incubating an apheresis sample containing leukocytes and platelets with the first, second and third reagents.

19. The kit of claim 18, wherein said first reagent includes a fluorescent intercalating dye.

20. The kit of claim 18, wherein said first reagent includes about 0.5 μM of TO-PRO-3.

21. The kit of claim 19, wherein said second reagent includes polyoxyethylene ether.

22. The kit of claim 20, wherein said second reagent includes about 90 mM TRITON X-100.

23. The kit of claim 21, wherein said third reagent includes a quaternary ammonium salt.

24. The kit of claim 22, wherein said third reagent includes about 60 mM dodecyltrimethylammonium bromide.

25. The kit of claim 18, further comprising a cartridge having at least one scan capillary.

26. The kit of claim 25, wherein said cartridge is triangular shaped.

27. The kit of claim 23, further comprising a triangular shaped cartridge having a glass capillary about three microns deep, about forty millimeters long and at least one millimeter wide.

********************