Hematology analyzers demystified (Proceedings)

Introduction

The manual PCV and plasma protein are quick, reliable methods for determining blood volume and hydration status when performed correctly. However, more than a bare-bones level of hematologic information has been made available with advances in technology. Wallace H. Coulter revolutionized hematology when he developed an automated means of counting particles that resulted in a patent in the early 1950s. His concept ultimately became known as "The Coulter Principle". Around the same time, others were involved with refining flow cytometry for sizing cells. In the 1980s, rudimentary automated differentials were developed for screening purposes. Continued refinement has resulted in the waning of manual differentials in human medicine and advances in the human health care market have played a significant role in the development of automation for veterinary hematology.

The hematology technology today involves three main principles: centrifugal, impedance (the Coulter principle) and light scatter. Spectrophotometry is shared by impedance and light scatter analyzers for the determination of hemoglobin concentration. Impedance is the most commonly used method in the clinic setting although progress has been made towards a cost-effective light scatter analyzer. A simplified review of the methods is discussed here along with a brief overview of instrument selection, common interferences and quality assurance. The Quality Assurance and Lab Errors sessions provide more in depth discussions of interferences and QA.

Centrifugal

The only veterinary centrifugal analyzer available is known as the QBC (Quantitative Buffy Coat) VetAutoread by Idexx. It uses the principle that cells form layers based on cell density when centrifuged in what resembles a giant hematocrit tube. Fluorescent dyes are used to stain lipoproteins, DNA and RNA to help differentiate the cells and the layers are amplified with the use of a float. This system performs reasonably well with normal samples and platelet estimates, and it can identify leukocytosis, the presence of reticulocytes and microfilaria. Since it measures the size of the layers, cell counts are estimated and cell sizes are not directly determined. Eosinophils are reportedly inconsistent in dog samples and there has been poor correlation with absolute reticulocyte counts.

Non-centrifugal sample processing

The sample is handled in a similar fashion by both impedance and at least some flow cytometry-based analyzers prior to the actual counting and sizing of cells: An aliquot of the blood sample is aspirated from the tube and divided into two portions. Analyzers typically use either a shear valve or a syringe displacement mechanism to obtain an aliquot of the sample. The aliquot is diverted into two parts: one to the "white blood cell (WBC) side" with a ~40-fold dilution in a reagent that lyses the erythrocytes for WBC counting and hemoglobin determination, and the second to the "red blood cell (RBC) side" for ~50,000-fold dilution in isotonic solution to allow RBC and platelet (Plt) counting and sizing. After this point, how the cells are counted and sized is unique for each technology.

Impedance technology

Impedance analyzers can be obtained from multiple sources including Coulter (AcT), Abaxis (VetScan HM5), Heska (HemaTrue), scil (Vet abc), Drew Scientific (HemaVet 950) and others. These use electronic methods to count and size cells (the Coulter Principle). Known volume of diluted blood is pulled through a small aperture flanked on each side by electrodes. Each cell displaces an equivalent volume of ionic solution as it passes through the aperture and impedes the electrical current passing between the electrodes proportionately, creating an electronic "blip". Because a specific sample volume is analyzed and a detector tracks the number and size of each blip, cell concentrations and their sizes (volume) can be accurately determined. This data is converted to a plot of particle size vs. numbers in the form of a histogram.

On the "RBC" side, both platelets and RBCs are counted and sized. When histograms representing the volume in femtoliters (fl) vs. number of RBCs from several species are compared, human erythrocytes are relatively large (~ 80-98 fl), those of dogs are smaller (62-73 fl), RBCs of cats are smaller yet (37-50 fl) and goat RBCs are tiny (~12-20 fl). The Mean Platelet Volume (MPV) of most species typically falls in the goat erythrocyte size range. Impedance analyzers distinguish types of particles based on size only. In humans, because of the wide size difference between erythrocytes and platelets, these two cell types are easily distinguished. This is not always true in our animal species.

Leukocytes are present in the solution and may be counted along with the erythrocytes; however they, too, are normally relatively few in number, and in most cases larger with no apparent impact on the RBC count. The average of the erythrocyte and platelet size curves is reported as the mean cell volume (MCV) and mean platelet volume (MPV), respectively. The width of the RBC and Plt histograms provides a visual indication of the anisocytosis (size variation) within each population. The amount of anisocytosis is calculated as the coefficient of variation (CV = SD/mean) of the respective cell volumes and is called the Red cell Distribution Width (RDW), or Platelet Distribution Width (PDW). The wider the curve, the higher the value and the more variation (anisocytosis) there is in cell size.

The WBC portion is exposed to a reagent that lyses the RBCs leaving free hemoglobin. The same reagent collapses the cytoplasm of the nucleated cells, leaving a nuclear remnant for counting and sizing. Three-part differentials that classify nucleated cells as granulocytes, monocytes and lymphocytes are based on volume differences between nuclear particles. Typically, granulocyte particle remnants (neutrophils, eosinophils and basophils) are the largest and lymphocyte nuclear remnants the smallest. Monocytes fall in between even though they appear to be the largest cells viewed under the microscope. Depending on the proportions of cell types present and the reagent system, computer-generated histograms may contain one to several adjacent curves. Hidden within the curves are overlapping peaks that represent cell categories. The cursors observed indicate the computer-derived cell groupings. Monocytes often comprise too small a portion to be a visibly distinct population in the histogram and their enumeration is generally not very reproducible unless present in large numbers. Eosinophils remain intact in select reagents and large numbers may appear as a shoulder off of the granulocyte curve of some instruments; smaller numbers are also not counted in a reproducible fashion.

The advantages of impedance analyzers relate to the accuracy of the obtaining aliquot and subsequent dilutions along with analysis of thousands of cells in a short period of time resulting in improved efficiency and precision over manual methods. In addition, cells sizes are determined and several calculations (MCHC, RDW, PDW, and Hct) are performed automatically. Disadvantages are related to the fact that cells are distinguished by size alone and there can be several reasons for size overlap, and therefore, error. Manual film evaluation is still necessary. In spite of these potential sources of error, impedance technology has been the gold standard in hematology for decades.

Light scatter technology

Companies using flow cytometers have various proprietary light scatter methods to count and size cells. Idexx is currently the only company to market an analyzer for typical veterinary clinics based on flow cytometry alone (LaserCyte). Cells are transported in a stream, encased in a sheath of fluid and passed single file in front of a beam of light. When the light beam strikes a cell, light is scattered off it at various angles. As with impedance methods, thousands of cells are analyzed. Detectors are placed in several locations to capture the scattered light. Minimally deflected light (forward angle scatter) typically indicates the particle size, while light scattered at higher angles (side scatter) typically indicates complexity of the nuclear remnant (lobularity) and/or the cell contents (cytoplasmic granularity). In addition to two dimensional impedance-like histograms showing cell size vs. number, flow cytometers provide three- dimensional scatter-plots showing number (each dot) as well as complexity vs. size. The resulting cytogram displays cell types in clusters. Since cell features other than size alone are used, differentiation between large platelets and erythrocytes and between leukocyte types is potentially improved over impedance technology. Negatives include higher instrument cost, sensitivity to laser alignment, and cells can still be missed or misclassified when groups overlap. The light scatter is more predictable for erythrocytes when the normally biconcave cells are sphered. Manual film evaluation is still necessary.

Hemoglobin determination

The hemoglobin concentration is determined by spectroscopy on both impedance and flow cytometry analyzers. Historically, after exposure to a lytic agent the liberated hemoglobin reacts with a cyanide-containing solution to form a stable compound, cyanmethemoglobin. Light of a specific wavelength is absorbed in proportion to the hemoglobin concentration and is measured by a spectrophotometer. Since cyanide is a hazardous substance, cyanide-free reagents have recently become routinely available. This measurement is typically very reproducible and accurate, but by itself, adds little to the clinical interpretation. Anything increasing the turbidity of the sample as lipemia (commonly due to insufficient fasting or less likely certain metabolic conditions as diabetes mellitus, hypothyroidism, and Cushings), incomplete cell lysis, many Heinz bodies, or extreme leukocytosis may falsely elevate the hemoglobin concentration. Anything falsely elevating the Hgb will falsely elevate the mean corpuscular hemoglobin concentration (MCHC). The MCHC is calculated by dividing the [Hgb] by the Hct and multiplying the result by 100. The typical reference interval of 33 - ~36 g/dl is a physiologic standard that should not be exceeded if everything is working appropriately and a quality sample is being tested. Note that hemolysis will not interfere with the hemoglobin determination since the sample is hemolyzed as part of the process anyway. Synthetic hemoglobin products will be measured as hemoglobin and can significantly increase the MCHC.

Instrument selection

Primary reasons for having an in-house hematology system include immediate medical management associated with health screens and urgent care and a potential for improvement over manual methods including speed, precision, and accuracy. Automation allows for counting thousands of cells in seconds over the more tedious manual differential counts.

Instruments are now available to fulfill the needs of most any laboratory center and vary as to expense, complexity and throughput. Human medicine is trending towards the use of large throughput instruments exceeding $100,000 placed in referral centers. This may be attributed to increased regulation effectively removing laboratory medicine out of the physician's offices. Veterinary referral laboratories take advantage of the technology offered by these larger analyzers, while considerable growth is visible in the market of smaller point of care-type analyzers for the clinic setting. Buyers often base their decision on convenience, up-front costs, and marketed features. While the purchase price is not an insignificant sum of money, it may not be the largest expense. Using approximate inflation-adjusted dollars, instruments that used to cost ~$200,000 and $0.50 in reagents/sample now cost ~$20,000 and ~$5.00/sample. More important issues include needs, performance and on-going costs, including tech time, reagents including calibrator and quality control material (QCM) and maintenance. Be sure to ask the sales rep provide a cost/billable test estimate rather than a cost/test estimate.

Unfortunately, instrumentation is not infallible. Concerns about the use of these systems relate to both man and "machine". The problems relating to man exist in part because those using the equipment in clinics are generally people, by nature, training and necessity, who are focused on the hands-on care of the patient. While this is a good thing for the patient, it leaves little time, and possibly desire, to understand the technical features of a relatively complex analyzer. In addition, analyzers approved for human use are strictly regulated regarding performance claims and product manufacture. While veterinary instrumentation is often modeled after the human counterpart, there is a lack of regulation for these products. Ideally, these instruments are finely tuned in order to produce precise and accurate results. Even with strict regulations, bioanalytical instruments require a certain level of expertise obtained through education, experience, and the use of an appropriate quality assurance program for proper utilization. Understanding the principles of operation is the first step towards insuring a successful outcome.

Determining precision (reproducibility) is a crucial piece of the selection process. To do so, obtain a quality sample (no clots, lipemia, or hemolysis) from a normal dog and run the sample multiple times (10 is ideal). Be sure to mix the tube sufficiently in between runs. Evaluate whether the values are within your clinical expectations. Alternatively, the CV may be calculated. Based on a recent publication [Becker 2008] the CVs can be expected run ~<4% for WBC, RBC, Hgb and Hct and slightly higher for Plt. Insure the MCHC remains within the stated reference interval. Other considerations include the sample volume, run time, steps necessary to change species and to begin a new lot of reagents, the daily, weekly and monthly maintenance requirements, and whether dedicated personnel can/will be committed to following the required maintenance as well as quality assurance procedures. These instruments can "lock up" if left unattended or idle for several days at a time. Tubing can become worn and become pinched or develop leaks and samples with microclots may induce plugs.

Quality assurance and calibration

Because the above changes can occur at any time, quality control material (QCM) should be run, values documented on a Levy-Jennings plot, and the results evaluated regularly. AAHA requires daily running of QCM and the tech must be able to follow up appropriately when data are outside expected ranges. This is a necessary cost of doing business and should be included in a cost/reportable evaluation. This follow up may require phone consultation with a technical service representative and possibly a calibration. Biochemistry instruments have essentially eliminated the need for calibration of biochemistry instruments. This is not the case for hematology analyzers. Once precision is verified through a reproducibility study, calibration material is used to adjust the obtained results to the values provided in the package insert. Any adjustments should be verified by running one or preferably two levels of QCM to insure the slope of the calibration line is appropriate. Rerunning the same material that was used as a calibrator only confirms reproducibility. Accuracy is only potentially verified at that single point of concentration for each analyte. Mishandling the calibrator material or misadjusting the analyzer can actually induce error that will not be detected by running the same material. This is a good time to emphasize the importance of following the material handling instructions exactly as described in the package insert.

Microscopes

It is recommended that a good microscope be among the first items purchased for the clinical laboratory. Microscope quality is directly proportional to the expense and one can expect to pay $3-5,000 for a quality scope. Plan lenses provide for viewing flat field viewing and are available as achromat lenses, for the budget minded and apoachromat, for higher resolution. Objectives should include a 10x, for scanning blood films, manual WBC counts on avian/reptile blood or cerebrospinal fluids, and urine wet mounts; a 40x for estimating WBCs on blood films, manual avian/exotic WBC counts and cerebrospinal fluid counts and urine wet mounts and a 100x oil for performing manual differentials, evaluating cell morphology and looking for bacteria and small parasites. The 40x objectives typically require the use of a coverslip for clear viewing; an acceptable alternative can be a thin coat of oil or simply place a glass coverslip on the slide to be viewed. If the 40x objective becomes contaminated with oil, use a cotton swab moistened with lens cleaner to thoroughly remove the oil. Avoid applying lens cleaner directly to the lenses to prolong the lifespan of the seals. As with any piece of equipment regular maintenance is necessary. For the microscope, this includes removing oil from the objectives at least daily and contract for at least annual cleaning by a professional.

References and recommended reading

Becker M, Moritz A, Giger U, Comparative clinical study of canine and feline total blood cell count results with seven in-clinic and two commercial laboratory hematology analyzers. VetClinPath 37:4 Dec 2008

Lassen ED and Weiser MG, General Principles of Laboratory Testing and Diagnosis in Veterinary Hematology and Clinical Chemistry, ed Mary Anna Thrall, 2004, Lippincott, Williams & Wilkins. Pg 14-18 and 21-37.

Weiser, M.G., L.M. Vap, and M.A. Thrall, Perspectives and Advances in In-Clinic Laboratory Diagnostic Capabilities: Hematology & Clinical Chemistry in Clinical Pathology and Diagnostic Techniques, An Issue of Veterinary Clinics: Small Animal Practice ed. D. Robin Allison and James H. Meinkoth, 2007: Saunders. Pg 221-236.