INTRODUCTION — During the first half of the 20th century, the complete blood count (CBC), one of the most commonly ordered laboratory tests, was performed using exclusively manual techniques:
●Blood cell counts (red cells, white cells, platelets) were performed using appropriately diluted blood samples and a ruled counting chamber (hemocytometer).
●Hemoglobin concentration was analyzed colorimetrically by the cyanmethemoglobin method.
●The hematocrit (Hct), also called the packed cell volume, was measured by high-speed centrifugation of a column of blood, either in a specially designed tube (the Wintrobe tube) (picture 1), or in sealed microcapillary tubes (ie, the "spun" hematocrit, often obtained by fingerstick blood collection) (picture 2).
●The white blood cell differential was obtained by examining and enumerating by class (eg, granulocytes, lymphocytes, monocytes) 100 to 200 individual white blood cells on a suitably stained blood smear.
In 1932, Wintrobe developed a set of indices that estimated erythrocyte size and hemoglobin content. All were calculated manually from the red blood cell (RBC) count, serum hemoglobin concentration (Hgb), and Hct, as follows:
Mean corpuscular volume (MCV; in femtoliters [fL]) = 10 x HCT (percent) ÷ RBC (millions/microL)
Mean corpuscular hemoglobin (MCH; in picograms [pg]/red cell) = HGB (g/dL) x 10 ÷ RBC (millions/microL)
Mean corpuscular hemoglobin concentration (MCHC), in grams per deciliter (g/dL) = HGB (g/dL) X 100 ÷ HCT (percent)
These early methods were laborious and imprecise. This topic review will discuss the methods and apparatus that have been used to automate the above procedures [1].
DEFINITIONS OF HEMATOLOGY PARAMETERS — The following hematology parameters are reported for a typical complete blood count (CBC) and may be flagged for a value outside the normal range:
●RBC – Red blood cell (RBC) count is the number of RBCs per microL of blood (or number of RBCs x 1012/L). An elevated RBC can reflect polycythemia (reactive or neoplastic) or a disorder of globin synthesis. A decreased RBC typically reflects anemia. (See "Diagnostic approach to the patient with erythrocytosis/polycythemia" and "Diagnostic approach to anemia in adults" and "Approach to the child with anemia".)
●Hemoglobin – Hemoglobin (Hgb) is the concentration of hemoglobin in whole blood, in grams/deciliter (g/dL). An increased Hgb may reflect a polycythemia (reactive or neoplastic) or be due to dehydration. A decreased Hgb typically reflects anemia.
●Hematocrit – The hematocrit (Hct) is the packed spun volume of blood made up of RBC, expressed as a percentage of total blood volume. It can be measured or calculated as Hct = (RBC x MCV)/10. An increased hematocrit can reflect polycythemia (reactive or neoplastic); if calculated, an increased Hct may reflect a normal number of RBC with an elevated MCV.
●MCV – Mean corpuscular volume (MCV) is the average volume (size) of the patient's RBCs. It can be measured or calculated as above (see 'Introduction' above). Anemia can be classified based on whether the MCV is low, normal, or elevated. (See "Diagnostic approach to anemia in adults", section on 'RBC indices'.)
●MCH – Mean corpuscular hemoglobin (MCH) is the average hemoglobin content in a RBC. It is calculated as MCH (pg/red cell) = Hgb (g/dL) x 10 ÷ RBC (millions/microL). A low MCH indicates decreased hemoglobin content per cell, and is typically reflected in hypochromia on the peripheral blood smear. This may be seen in iron deficiency and disorders of globin synthesis.
●MCHC – Mean corpuscular hemoglobin concentration (MCHC) is the average hemoglobin concentration per RBC, in grams/dL. It is calculated as MCHC (g/dL) = Hgb (g/dL) X 100 ÷ Hct (percent). Low and high MCHC values are helpful in classifying anemias. Very low MCHC values are typical of iron deficiency anemia, and very high MCHC values typically reflect spherocytosis or RBC agglutination. Examination of the peripheral blood smear is helpful in distinguishing these findings. (See "Evaluation of the peripheral blood smear", section on 'Red blood cells' and "Diagnostic approach to anemia in adults", section on 'Anemia definitions'.)
●RDW – Red cell distribution width (RDW) is a measure of the variation in RBC size, which is reflected in the degree of anisocytosis on the peripheral blood smear. A high RDW implies a large variation in RBC sizes, and a low RDW implies a more homogeneous population of RBCs. RDW is calculated; the RDW is the coefficient of variation (CV) or the standard deviation (SD) of the red cell volume distribution curve. RDW is helpful in the classification of anemia. A very elevated RDW can be seen in iron deficiency anemia, transfused anemia, myelodysplastic syndrome, and homozygous hemoglobinopathy, whereas normal to slightly elevated RDWs can be seen in thalassemia trait and anemia of chronic disease. (See "Evaluation of the peripheral blood smear" and "Diagnostic approach to anemia in adults", section on 'RBC indices'.)
●WBC – The white blood cell (WBC) count is the number of WBC per microL of blood (or number of WBC x 109/L). An elevated WBC (ie, leukocytosis) may be seen in neoplastic and non-neoplastic conditions. If the WBC is elevated, enumeration of the WBC differential and review of the peripheral blood smear is used along with clinical evaluation to determine the cause. (See "Approach to the patient with neutrophilia".)
If the WBC is decreased (ie, leukopenia), a WBC differential should be obtained and the peripheral blood smear examined to enumerate which cell type is decreased. Since neutrophils make up the greatest percentage of WBCs, leukopenia is usually due to neutropenia at a minimum (and perhaps decreases in other WBCs as well). (See "Overview of neutropenia in children and adolescents" and "Approach to the adult with unexplained neutropenia".)
●PLT – The platelet count (PLT) is the number of platelets per microL of blood (or number of platelets x 109/L). An elevated platelet count (ie, thrombocytosis, also called thrombocythemia) may be seen in reactive and neoplastic conditions. A decreased platelet count (ie, thrombocytopenia) may reflect platelet destruction, sequestration, or ineffective thrombopoiesis. Examination of the peripheral blood smear is helpful in distinguishing among possible causes. (See "Approach to the patient with thrombocytosis" and "Diagnostic approach to the adult with unexplained thrombocytopenia" and "Causes of thrombocytopenia in children" and "Neonatal thrombocytopenia: Etiology".)
●MPV – Mean platelet volume (MPV) is the average volume (size) of the patient's platelets as measured in femtoliters (fL). Evaluation of the MPV should be done in context of the platelet count. (See 'Platelet error flags' below and "Diagnostic approach to the adult with unexplained thrombocytopenia", section on 'Laboratory testing' and 'Approach to an abnormal MPV' below.)
Normal values for RBC parameters are included in the table (table 1). Causes of abnormal "flags" for RBC, WBC, and platelet parameters are discussed in detail below.
SAMPLE COLLECTION — Samples to be used for a complete blood count should be anticoagulated with a suitable agent, such as EDTA or citrate. The use of EDTA may cause platelet clumping in some patients (ie, pseudothrombocytopenia).
It is recommended that blood samples be kept at room temperature if complete blood count (CBC) analysis is to occur within 24 hours of collection. Samples should be refrigerated if the analysis is to occur up to 72 hours after collection [2]. We do not recommend that samples more than 72 hours old be used for CBC testing. Samples from which blood films will be prepared should be prepared within eight hours of collection.
THE COULTER IMPEDANCE APERTURE — In 1953, Wallace Coulter patented a device developed in their basement that used an electric impedance aperture to count red cells and white cells [3]. In the impedance method, blood is diluted in a current-conducting solution, such as isotonic saline. The suspension of cells is drawn by a vacuum through a small aperture positioned between two sensing electrodes connected across a direct current potential. As each cell passes through the aperture, the resistance between the electrodes increases. The non-conductive cell being counted produces a momentary increase in impedance, resulting in an electrical pulse, with the number of pulses indicating the cell count, and the amplitude of the pulse being proportional to the cell's volume. Upper and lower pulse height thresholds could be selected such that only particles within a specific volume range would be counted.
Since typical red blood cell (RBC) counts are three orders of magnitude greater than white blood cell (WBC) counts, and since appropriate settings of the lower pulse height threshold eliminate counting of the smaller platelets (platelet volume is approximately 9 fL, or approximately one-tenth the volume of mature red cells), a cell count on appropriately diluted whole blood samples served to yield a RBC count. RBC lysing reagents then need to be added to the sample before the white blood cells can be counted with accuracy.
In the 1960s, the first multichannel automated hematology instruments appeared. Anticoagulated whole blood samples were aspirated into the apparatus and automatically aliquoted and diluted into RBC and WBC counting chambers (baths) for cell counting and sizing with impedance apertures, and into a spectrophotometric cuvette for Hgb determination using the cyanmethemoglobin method. Coincidence correction (to correct for when two or more cells passing through the aperture are counted as one cell) and calculations of RBC indices were automatically performed.
Since these machines measure the RBC count directly, as well as the mean corpuscular volume (MCV), the hematocrit (Hct) could be electronically calculated by rearranging one of the original Wintrobe formulae shown above:
Hct (percent) = RBC (in millions/microL) x MCV (fL) ÷ 10
These first multichannel instruments revolutionized the complete blood count (CBC) as a laboratory test, enabling high throughput, low cost, and fast turnaround time with the ability to produce immediate (STAT) results. However, the platelet count was not a part of the first automated CBC, since the aperture resolution on the early instruments could not adequately separate platelets from RBCs.
The red blood cell volume histogram — Modern cell counting and sizing apertures have sufficient precision to function as "channelizers." As an example, Coulter instruments channel cells within the RBC/platelet aperture into 256 bins in the size range from zero to 360 fL. RBCs are not only enumerated as those particles with a cell volume within the range of 36 to 360 fL, but they are also classified into appropriate "channels" by size (related to the pulse height). For instruments with pulse editing circuitry, aberrant pulses are excluded from these channels. A plot of frequency versus channel size allows the development of a RBC volume histogram, which can be smoothed into a RBC size distribution curve.
The RBC size frequency distribution curve typically has a symmetrical or Gaussian shape. The MCV and red cell distribution width (RDW, the coefficient of variation of the RBC volume distribution) are directly derived from this curve. A number of abnormalities of this curve can be seen in disease settings:
●A left "shoulder" extension to the curve, or failure of the curve to reach baseline on the left side (ie, RBCs with smaller volumes), can indicate the presence of a population of very small RBCs (eg, microspherocytes or schistocytes), or might be associated with the presence of unusually large platelets (macrothrombocytes) or platelet clumps.
●A separate RBC population to the left (ie, microcytic RBCs) can indicate the presence of two populations of red cells, as seen in X-linked sideroblastic anemia (figure 1). (See "Sideroblastic anemias: Diagnosis and management", section on 'X-linked sideroblastic anemias'.)
●A right-sided shoulder usually corresponds to a population of extremely large RBCs (macrocytes) or reticulocytes.
●A trailing erythrocyte population to the extreme right (eg, cells with a MCV >200 fL) can indicate the presence of RBC agglutinins (picture 3), which cause red cells to pass through the aperture in small clumps.
All pulses in the RBC aperture are enumerated for the RBC count, even if they have an aberrant shape. However, RBCs with an aberrant shape generate a pulse height that is also aberrant and this is edited out of the MCV computation. Thus, the MCV excludes aberrant pulses if derived from an impedance analyzer; in some samples, up to 20 percent of counted RBCs may be excluded from the MCV calculation. Thus, impedance apertures have a limited linear range for particle sizing that can result in an MCHC with a limited range.
Red cell distribution width — The red cell distribution width (RDW), is an indicator of the degree of variation in RBC size (ie, anisocytosis). In Beckman Coulter instruments, the RDW is made equal to the coefficient of variation (CV) of the red cell volume distribution curve, while in some Sysmex instruments, the RDW is equal to the standard deviation of this curve.
RDW has been proposed as a tool to distinguish iron deficiency (elevated RDW) from thalassemia trait (normal RDW) in samples with low MCV. However, many cases of thalassemia trait were found to have an elevated RDW. (See "Microcytosis/Microcytic anemia", section on 'RDW (size variability)'.)
Platelet count and size
Platelet count — Platelets may be counted in multiple ways: manual microscopy, impedance, light scatter, immunologic methods, and digital image analysis [4]. Automated platelet counting primarily uses impedance and optical scatter, although immunologic methods are available on some instruments and image analysis is a developing technology. This section will focus on impedance platelet counting with optical platelet counting methods discussed in a subsequent section. (See 'The optical platelet count' below.)
Evaluation of patients with abnormally low or high platelet counts (ie, thrombocytopenia or thrombocytosis, respectively) is discussed in separate topic reviews. (See "Diagnostic approach to the adult with unexplained thrombocytopenia" and "Causes of thrombocytopenia in children" and "Approach to the patient with thrombocytosis".)
On instrumentation that uses impedance-based platelet counting, platelets are enumerated as particles in a given range, (eg, from 2 fL up to 30 fL in volume). Similar to RBCs, a frequency curve is generated from a channelized histogram of platelet volumes (eg, raw data curve), and is extrapolated to 70 fL. In contrast to the Gaussian distribution typical of RBCs, platelets show a lognormal or skewed distribution. A mean platelet volume (MPV) is calculated from a log transformation of the platelet volume distribution curve (eg, fitted curve), yielding a geometric mean for this parameter. Fitting a lognormal curve to the raw data can help exclude interferences (eg, microcytic red cells, cytoplasmic fragments, debris, electronic noise) and allows the inclusion of giant platelets with a size greater than 20 fL (picture 4).
Inspection of the platelet histogram can aid in determining inaccuracies in the platelet count; these are typically "flagged" by the instrument to indicate their presence. A peak or spike at the low end (eg, at volumes <2 fL) suggests the presence of cytoplasmic fragments or sufficient interference by electronic noise, while a failure to return to baseline at the high end (eg, volumes >20 to 30 fL) indicates interference by microcytic RBCs or failure to include giant platelets in the count.
When instrument platelet flags occur, it is important to verify the platelet count by estimation from a peripheral blood smear. In an area of the peripheral blood smear where RBCs barely touch, the number of platelets per 100 power field when multiplied by 20 x 109/L gives an estimate of the platelet count, which can then be compared with the machine-derived count. (See "Evaluation of the peripheral blood smear", section on 'Platelets'.)
Different instruments employ varying techniques to detect interferences that can result in a spurious platelet count. Some instruments enumerate particles between 2 to 3 fL (on the low end) and 20 to 30 fL (on the high end) as platelets and monitor the percent of total particles near the lower counting threshold (ie, 1 to 2 fL) and the upper counting threshold (ie, 20 to 30 fL). When the number of particles in these regions exceeds certain limits, "flags" are generated.
The inability to distinguish platelets from other particles, such as microcytic red blood cells and electronic noise, has been an often cited limitation of impedance methodology. However, some automated hematology analyzers have shown better correlation of their impedance platelet counts with the reference method compared with optical methods [5].
Platelet counting using light scattering and other techniques, which obviates many of the problems enumerated above is discussed in a subsequent section. (See 'The optical platelet count' below.)
Mean platelet volume (platelet size) — The mean platelet volume (MPV) is the volume of the average circulating platelet in femtoliters (fL), similar to the MCV for red blood cells.
The MPV value is determined from the geometric mean of the transformed lognormal platelet volume data in impedance technology systems; or it can be measured using optical technology, where in some systems the MPV is the mode of the measured platelet volume. Thus, the MPV will vary with the method by which it is measured (ie, it is instrument-specific). Further, a patient's true platelet distribution may not fit a log-normal distribution, and manufacturers each specify a defined range of MPV within their algorithms. Thus, one must refer to the laboratory's reference range, which will have been established for that patient population using a specific hematology instrument [6]. As an example of the instrument-to-instrument variation, a study examining MPV ranges in adults with normal platelet counts reported that impedance methods had normal values ranging from 6.0 to 13.2 fL, whereas optical methods had normal values ranging from 5.6 to 12.1 fL [6].
Under normal circumstances, there is an inverse relationship between platelet size and number. Therefore, the total platelet mass (ie, the product of the MPV and platelet count or "plateletcrit") is closely regulated. When platelets decrease in number, bone marrow megakaryocytes are stimulated by thrombopoietin; these stimulated megakaryocytes produce larger platelets. Thus, platelets with a higher MPV are expected to be seen in destructive thrombocytopenia when megakaryocytic stimulation is present. Conversely, platelets with a lower MPV are expected in thrombocytopenic states associated with marrow hypoplasia or aplasia. (See "Megakaryocyte biology and the production of platelets", section on 'Regulation of platelet production'.)
An exception to this relationship occurs with splenic sequestration in which a low MPV is seen because the spleen sequesters large platelets. Similarly, a higher MPV is seen in hyposplenic states since there is no spleen to sequester the larger platelets.
Many clinicians do not use the MPV because of difficulties associated with this measurement, although most modern hematology analyzers report the MPV, and the literature indicates that MPV measurement may have diagnostic utility (eg, as a marker for increased cardiovascular risk, venous thromboembolism, or diabetes) [7]. Many laboratories do not report the MPV as a routine part of the CBC, because of technical problems associated with the accurate measurement of MPV (eg, swelling of platelets after prolonged exposure to EDTA, lack of an international standard, reference range of MPV varies with the platelet count).
Approach to an abnormal MPV — In an otherwise healthy patient without a history of bleeding and with an unremarkable CBC, the MPV is unlikely to be useful in diagnosing an underlying disorder or in predicting an increased bleeding risk.
For individuals with other abnormalities in the CBC, an abnormal MPV may be helpful in the evaluation, although it cannot be used to confirm or exclude any of the possible diagnoses. In such cases, we first confirm that the abnormal MPV is accurate by ensuring that the patient's blood is drawn and tested within six hours. This may require that the patient have blood drawn in or near the hematology laboratory where testing occurs. Should a clinician order the MPV specifically, the sample should be analyzed after two hours to allow stabilization of platelet shape changes, but without significant further delays.
We then base our further assessment on whether the MPV is high or low, as follows:
●A high MPV in a thrombocytopenic patient indicates active bone marrow production of platelets (eg, as in immune thrombocytopenia [ITP]). A high MPV is also seen in some congenital thrombocytopenias (eg, gray platelet syndrome, May-Hegglin anomaly, Bernard-Soulier syndrome) and in some patients with myelodysplastic syndromes (table 2). (See "Congenital and acquired disorders of platelet function".)
A sequential increase in MPV over time can indicate megakaryocytic regeneration, such as recovery from a hypoplastic or aplastic state.
●A low MPV in a thrombocytopenic patient is indicative of bone marrow suppression (eg, as in aplastic anemia) [8]. A low MPV may also be seen with some congenital thrombocytopenias (eg, Wiskott-Aldrich syndrome).
Examination of the peripheral blood smear is a logical next step with a "true" flag for an abnormal MPV when one of these conditions is suspected based on other abnormalities in the CBC. Morphologic features of myelodysplasia or consumptive processes, such as microangiopathic hemolytic anemia, can be seen and help lead to a correct diagnosis. Several of the inherited thrombocytopenic disorders have large or giant platelets with leukocyte inclusions. In the patient with ITP, the peripheral blood smear will show normal to large sized platelets, although these features are nonspecific. (See "Evaluation of the peripheral blood smear".)
Correlation of the MPV with reticulated platelet parameters such as the immature platelet fraction (IPF) can also be useful in distinguishing hypoplastic platelet production from platelet destruction syndromes. (See 'Reticulated platelets' below.)
CELL COUNTING BY LIGHT SCATTERING — The 1970s saw the introduction of cell counting by light scattering technology. In this technique, cells are hydrodynamically focused in a flow cell and illuminated by a narrow beam of laser light. The light scattered by each cell is captured by a photodetector and converted to an electrical pulse; the total number of pulses is proportional to the cell count, while pulse amplitude is proportional to cell volume. The number of pulses within a predetermined size range (eg, 30 to 180 fL for red cells and 0 to 20 [or 30] fL for platelets) yields the corresponding red blood cell (RBC) and platelet counts. White cells can be counted with light scattering technology after lysis of the RBCs.
By analyzing light scatter at multiple angles, better separation of white blood cells (WBCs) from interferences, such as unlysed RBCs (a problem with neonatal samples, which have difficult-to-lyse RBCs), and nucleated RBCs can be obtained. Some instruments determine both optical and impedance WBC and compare them to look for interferences.
For erythrocyte evaluation with light scattering technology, the erythroid cells first must be isovolumetrically sphered and fixed. Scattered light is measured at low forward angle (0 to 3 degrees) and high forward angle (5 to 15 degrees), yielding pulses proportional to RBC size and refractive index, respectively. The latter is proportional to the hemoglobin content of the cell, allowing hemoglobin content to be directly determined for individual RBCs. This generates a RBC "cytogram" that allows calculation of RBC volume and Hgb concentration for each individual red blood cell.
RBC volume and RBC hemoglobin concentration histograms and smoothed curves can be constructed, and the mean, standard deviation (SD), and coefficients of variation (CV) electronically calculated, yielding the following measurements:
●Mean corpuscular volume (MCV) and red cell distribution width (RDW) (as with impedance methods)
●CHM (mean of directly measured HGB content per red cell) and the CHDW (cellular hemoglobin content distribution width)
●CHCM (mean of directly measured HGB concentration) and the CHCDW (cellular hemoglobin concentration distribution width)
The mean value of the directly measured RBC hemoglobin concentration (CHCM) may be decreased and the distribution width of the cellular hemoglobin concentration curve (CHCDW) may be increased in moderate to severe iron deficiency, due to a shift in the percent of hypochromic RBCs. However, these parameters have not proven to be a sensitive indicator of early iron deficiency. (See "Causes and diagnosis of iron deficiency and iron deficiency anemia in adults", section on 'Findings on CBC'.)
The optical platelet count — In a manner similar to optical counting of red cells, an optical platelet count and volume can be determined from two-dimensional analysis of low and high angle light scatter measurements. These two light scatter measurements are related to platelet volume (size) and refractive index (density) [9]. Platelet volume histograms and smoothed curves can be constructed from this plot, and the platelet count, mean platelet volume (MPV) and SD or CV of the volume (ie, platelet distribution width, PDW) derived. Interference by microcytic RBC, RBC fragments, and WBC cytoplasmic fragments is eliminated since these particles have a different refractive index than platelets and are not counted.
Evaluation of patients with abnormally low or high platelet counts (ie, thrombocytopenia or thrombocytosis, respectively) is discussed in separate topic reviews. (See "Diagnostic approach to the adult with unexplained thrombocytopenia" and "Causes of thrombocytopenia in children" and "Approach to the patient with thrombocytosis".)
In some Abbott instruments, platelets are first plotted using scattering angles of 90 degrees versus 7 degrees. The ascending angle of the platelet population is determined and a histogram is formed perpendicular to the axis of this angle. Thresholds are set to isolate the platelets from large and small cells and fragments on either side of the histogram. A second histogram is constructed along the main body of this refined platelet population in order to determine the upper threshold separating platelets from microcytic RBCs.
Some Sysmex instruments provides both a fluorescence optical platelet count and an impedance count, utilizing a computer algorithm to give the best count based on whether, as an example, interference or an abnormal platelet distribution is detected. The fluorescence-based platelet count uses a patented fluorescent dye containing oxazine, which stains mitochondria and rough surface endoplasmic reticulum, and is platelet-specific when used with a dedicated fluorescence channel, thus minimizing interference with red cell fragments, small red blood cells, and white blood cell cytoplasmic fragments [10].
Immunologic counting of platelets — The immunologic method of platelet counting by flow cytometry is now the proposed International Reference Method for enumeration of platelets [11]. Some of Abbott's instruments use an immunologic platelet counting method in which platelets are identified by cluster analysis on a two-dimensional plot of fluorescent monoclonal antibody CD61 reactivity versus high forward angle light scatter. This optional platelet count analysis may be especially useful for very low platelet counts, or when flagged interferences with the platelet count are noted. In one study, that compared the accuracy of platelet counting among automated hematology analyzers in severe thrombocytopenia, this method was found to be the most accurate of all analyzers [12].
In another comparative study, the optical and flow cytometric (immunoplatelet) methods were found to be in good agreement for platelet counts in the range of 25 to 547 x 109/L (25,000 to 547,000/microL) [13]. However, for platelets counts <25 x 109/L, the optical method tended to overestimate the platelet count (figure 2). The lowest absolute platelet count that these methods could reliably distinguish from true zero was 1.730 x 109/L for the optical method and 0.020 and 0.009 x 109/L for the anti-CD41a and the anti-CD61 immunoplatelet methods, respectively.
In a follow-up study, the authors compared three different hematology analyzers with the immunoplatelet method and found a coefficient of determination (R2) >0.98 for the overall platelet range. However, examination of each individual platelet range (hematology analyzer versus immunoplatelet method) showed R2 values below 0.98 [14]. The XE-2100 (optical mode) and LH-750 showed values R2 >0.9 for all platelet ranges, while for platelet counts less than 50 x 109/L, the Advia 120 showed the best correlation with the immunoplatelet method.
Reticulated platelets — Reticulated platelets are the youngest circulating platelets, analogous to the relationship between reticulocytes and mature RBC. Like reticulocytes, reticulated platelets have an increased RNA content. Measurement of the RNA content of platelets with thiazole orange dye can be useful in the diagnostic classification of thrombocytopenia and in monitoring recovering thrombopoiesis (figure 3) [15-17]. However, a lack of standardization of reticulated platelets has hindered clinical use of this parameter [18].
Another method to quantify reticulated platelets, expressed as the immature platelet fraction (IPF), has been developed on some Sysmex instruments using its proprietary fluorescent dye containing oxazine and polymethine. This method has demonstrated increased IPF values (ie, an increased population of young platelets) in patients with thrombocytopenia due to increased peripheral platelet destruction [10], consistent with clinical observations in such patients.
Comparison of the IPF with reticulated platelets as measured by flow cytometry has shown good correlation [19], and studies support that the IPF has value in the role of platelet recovery after chemotherapy or hematopoietic cell transplantation, risk index of thrombosis in patients with thrombocytosis, timing for platelet transfusion, evaluation of platelet turnover, myelodysplastic syndrome, disseminated intravascular coagulation, and as a marker of megakaryopoiesis in liver cirrhosis [20-25].
ALTERNATIVE MEASUREMENTS OF HEMOGLOBIN CONCENTRATION — Concerns about the disposal of the potassium cyanide reagent have led to measurement of hemoglobin by methods other than the conversion of hemoglobin to hemiglobin cyanide (HiCN). These alternatives include the use of sodium lauryl sulphate (SLS), which interacts with hemoglobin to form an SLS-Hgb derivative, methemoglobin conversion with imidazole as the heme-ligand, and azide methemoglobin conversion. These methods may show less interference from substances that spuriously elevate HiCN measurement via increase in turbidity, such as high WBC, high triglycerides, or the presence of poorly soluble (at room temperature) IgM paraproteins.
AUTOMATED COUNTING OF RETICULOCYTES — Automated counting of reticulocytes became available at the start of the 1990s, significantly improving the precision of reticulocyte counts as compared with manual counting. A reticulocyte is defined in this methodology as the stage of erythrocyte development following loss of its nucleus in which the cell still contains measurable amounts of RNA. Automation provides an absolute reticulocyte count (ie, the number of reticulocytes per microL of blood), as well as the reticulocyte percentage, as a percent of all red cells (see "Diagnostic approach to anemia in adults", section on 'Reticulocyte count'). Automation often involves use of a fluorescent dye, such as nucleic acid binding dyes, including thiazole orange, first used in flow cytometry. As well as yielding a reticulocyte count, an immature reticulocyte fraction (IRF) can be measured as the sum of the reticulocytes with medium and high fluorescence [26,27]. Features of the combined absolute reticulocyte count and IRF are characteristic of different disease states and may be clinically useful in classifying anemia [28]. Reference intervals for the IRF are dependent on each method used.
Automated reticulocyte counts are included on most automated hematology analyzers:
●Siemens Advia instruments stain reticulocytes with Oxazine 750 and measure light adsorption [28]. Other reticulocyte parameters are reported in sphered reticulocytes (eg, MCVr, CHCMr, HDWr, CHr, CHDWr) similar to those for the sphered mature red blood cells (RBCs) [29]. High, medium, and low fluorescent fractions are also calculated.
The hemoglobin content of reticulocytes (CHr), measured on the Siemens Advia instruments, appears to be a sensitive and specific indicator in detecting iron deficiency, monitoring iron therapy in patients with renal failure on chronic dialysis, and monitoring for the advent of iron deficiency following treatment with erythropoietin in such patients [30,31].
Another reticulocyte index, the mean reticulocyte volume, has been proposed to be helpful in diagnosing iron deficiency, monitoring response to treatment of nutritional anemias, an early indicator of erythropoiesis after bone marrow transplantation, and as a tool in the evaluation of erythropoietin abuse in athletes [28,32-34].
●Some Abbott instruments use a proprietary fluorescent RNA dye called CD4K540 [35]. This dye is 10 times as bright as Auramine O and fluoresces at 530 nm (green). A plot of seven degree light scatter versus fluorescence separates platelets from RBC and very highly fluorescent non-viable WBC and nucleated RBC from the reticulocytes. The combined RBC and reticulocytes are gated; a histogram of the mature RBC plus reticulocyte fluorescence is analyzed with a valley-finding algorithm to separate and enumerate the reticulocytes. The IRF is identified arbitrarily by setting a threshold 30 channels above the RBC/reticulocyte threshold.
●The Beckman-Coulter instruments stain blood with new methylene blue, followed by clearing with a hypotonic solution [36]. Red cells are measured for volume, conductivity, and laser light scatter; the latter signal is proportional to the residual RNA within the red cell. Reticulocyte volume is measured, and those reticulocytes with the most RNA (corresponding to those reticulocytes in the highest light scatter regions) are measured. The DxH800 captures five laser light scatter measurements for each cell, in contrast to one light scatter measurement for the LH750/780. The IRF is then the ratio of those reticulocytes with the most RNA over the total number of reticulocytes.
Reticulocytes and the immature reticulocyte fraction (IRF) are determined on Sysmex instruments using a polymethine dye that stains reticulocyte RNA. Graphing fluorescence versus forward light scatter separates mature RBCs from reticulocytes. The IRF is the sum of the moderately and highly fluorescent reticulocytes (MFR+HFR). Using a combination of RBC and reticulocyte parameters on the Sysmex XN, some authors have reported a screening method for iron deficiency anemia and hereditary RBC diseases with a sensitivity and specificity of 95.2 and 99.9 percent, respectively, using a classification and regression tree analysis [37]. (See "Causes and diagnosis of iron deficiency and iron deficiency anemia in adults", section on 'Iron studies (list of available tests)'.)
AUTOMATION OF THE WHITE CELL DIFFERENTIAL — In addition to cell counts, the classical complete blood count (CBC) includes examination of a Wright-Giemsa stained blood film to subclassify the leukocytes (leukocyte or white blood cell [WBC] differential) and assess the morphology of the leukocytes, erythrocytes, and platelets. Attempts to automate this process started with image analysis instruments in the 1970s and progressed to a flow-through automated differential [38].
The three-part differential — Differentiating leukocyte subsets in liquid suspension, rather than on a stained smear, began in the 1980s. Flow-through automated differentials were first performed using standard electrical resistance or impedance technology. A weak lysing reagent resulted in lysing of the RBCs and shrinking of the leukocyte membranes about the nucleus, resulting in three peaks of leukocyte subsets in the normal individual:
●Lymphocytes and basophils in the smallest size group (35 to 90 fL)
●Segmented and band neutrophils and eosinophils ("granulocytes") in the largest size group (>160 fL)
●Monocytes and other mononuclear cells, including immature granulocytes and a portion of the eosinophils, were found in a smaller intermediate size peak between 90 and 160 fL
The presence of abnormal cells obscured the valleys between these three peaks, resulting in error "flags" and inaccurate counts. The "granulocyte" fraction did not always give a true absolute neutrophil count, particularly if significant numbers of bands, immature granulocytes, or eosinophils were present. In addition, increased numbers of bands and eosinophils, as well as more extreme abnormalities (eg, blasts, reactive lymphocytes, nucleated RBC) all give flagged samples, resulting in a slide review rate >30 percent for a hospital inpatient population. Thus, the three-part differential functioned best in the situation of screening a largely normal population for potential abnormalities.
Many of the smaller automated hematology instruments still use this technology.
The five-part differential — Five-part differentials were first introduced in the early 1990s [38]. Different instrument manufacturers all report at a minimum the basic five leukocyte subsets (neutrophils, eosinophils, basophils, lymphocytes, and monocytes) and have flags to detect the presence of abnormal cells (eg, blasts, nucleated RBC, reactive lymphocytes, lymphoma cells, mononuclear granulocytes). In fact, many of the newer analyzers report a seven-part differential, including the quantification of immature granulocytes and nucleated red blood cells. Performance in the detection of these abnormalities varies; most instruments are >90 percent sensitive in detecting large numbers of abnormal cells (ie, when the population exceeds 5 percent), but lack sensitivity to low frequency abnormalities (ie, <5 percent). Although these instruments claim to flag increases in band neutrophils, sensitivity and specificity for detection of elevations in this cell type are poor.
The current generation of instruments use a combination of technologies for the leukocyte differential, including:
●Impedance volume with direct current (DC)
●Radiofrequency (RF) conductivity (with impedance aperture)
●Laser light scattering
●Peroxidase staining
●Propidium iodide fluorescence (for nucleated RBC and non-viable cells)
●Cell-specific lysing reagents
●Polymethine RNA/DNA histone dye
●Digital imaging
Both DC and RF methods use impedance technology in which a cell breaks electrical current generated across an aperture opening, producing a voltage change. When DC current is applied, pulse height is related to overall cell volume. In contrast, when high radiofrequency (RF) current is applied across the counting aperture, the pulse height or voltage change is proportional to nuclear size and density.
Suspect flags — Instruments performing automated WBC differentials have a variety of "suspect" flags used to indicate that abnormal cell populations may be present and that the automated differential may be inaccurate. In general, these "suspect" flags can be divided into four types:
●Interfering particles are present at the lower WBC counting threshold, or lowest forward/side light scatter region for lymphocytes. Typical resulting flags include: nRBC (nucleated red blood cell), CLUMP (platelet), GIANT (platelet).
●Large mononuclear cells are present at the monocyte/neutrophil interface or with high values for high angle (90 degree) light scattering. Typical resulting instrument flags: BLAST.
●Large cells are present in the lymphoid region or at the interface between lymphoid and monocyte regions. Typical instrument flags include: ATYPICAL LYMPH, BLAST.
●There is a shifted position in the neutrophil cluster, with a large amount of forward or side light scatter. Typical instrument flags include: IMMATURE NEUTROPHILS, BANDS.
When one encounters flagged samples in the laboratory that indicate a possible abnormal cell population or interfering population, a peripheral blood smear should be examined to confirm whether an abnormality is truly present and to further define this. Typically, this occurs within the clinical laboratory by medical technologists, and a further subset of these samples is sent for pathologist review based on individual laboratory criteria. (See "Evaluation of the peripheral blood smear".)
Cell population data (CPD) — Hematology analyzers also provide additional data in the form of population data for an individual type of leukocyte, which is generated from the differential analysis of WBCs [39]. CPD on neutrophils has been used to screen for bacterial infection and sepsis, and CPD on lymphocytes has been used for discriminating viral infections from lymphoid malignancies [40,41]. Mathematical models of the population dynamics of subtypes of leukocytes has allowed for risk stratification of patients with acute coronary syndrome [42].
AUTOMATED SAMPLING MODES AND COMPUTER ADVANCES — Advances in automated hematology instruments were made possible by new, improved, and cheaper computer processing technology and memory. Instruments contained on-board computers to perform curve fitting, multi-dimensional population cluster discrimination, quality control data, and moving averages for quality control. In addition, there is sufficient memory to store and sort samples and to store a finite number of sample scatterplots in "list" mode for reanalysis.
Analyzers now include an automated sampling mode. Blood is aspirated from closed tubes, with automated bar code readability, minimizing operator contact with the blood, and allowing for positive sample identification. A built-in mixing system ensured sampling of a well-mixed specimen, and instruments usually included a sample aspiration monitor to aid in detecting the presence of clots and determining that the correct sample volume has been aspirated. Sporadic tests, such as reticulocytes or an automated differential, can be selected by random access and performed only when ordered, saving reagent costs. Automated slide makers also are available that can produce and stain blood films as directed by laboratory-programmed algorithms, which consider both the results of specific parameters as well as the presence of instrument flags. The combination of automated hematology analyzers coupled with automated slide makers and stainers have resulted in automated hematology workcells which offer high throughput and increased efficiency [43].
The current generation of automated hematology instruments uses laser light scatter, fluorescence, DNA/RNA dyes, and monoclonal antibodies to expand the scope of the automated differential. Some Abbott instruments use fluorescein isothiocyanate (FITC)- and R-phycoerythrin (PE)-conjugated monoclonal antibodies to enumerate CD3, CD4, and CD8 counts [44]. The Sysmex XE and XN analyzers use a polymethine dye that binds RNA, DNA, and histones and light scatter to enumerate hematopoietic progenitor cells, which some authors have suggested be used as a surrogate for CD34 stem cell quantification for peripheral blood stem cell harvesting [45,46].
Some analyzers also enumerate immature granulocytes (IGs). The automated IG is being studied as a predictor of sepsis, and studies show that IG counts, while highly specific for infections, show low sensitivity [47]. Hematology analyzers are also being used to generate patterns of data to identify patients with myelodysplastic syndromes, infections, myocardial infarction, and lymphoproliferative disorders [48-52].
Enumeration of abnormal cell types, such as blasts, immature granulocytes, and reactive lymphocytes is progressing with laser light scatter, fluorescence, and monoclonal antibodies.
CAUSES OF SPURIOUS RESULTS — Although modern hematology instruments are accurate and precise, interfering substances occasionally produce erroneous results [53,54].
Spurious increase in the MCHC — Monitoring the mean corpuscular hemoglobin concentration (MCHC) for each sample is one of the best ways to assess for both instrument and sample problems. Although a true increase in the MCHC can be seen when spherocytes are present, a spuriously elevated HGB or spuriously low RBC can result in elevation of the MCHC. Common sample problems leading to these results include:
●A very high WBC, lipemia, or a precipitating monoclonal protein, all of which produce turbidity in the HGB colorimetric method, will result in a spuriously high HGB.
●The presence of a cold agglutinin spuriously decreases the RBC count since RBCs go through the counter in small groups, rather than one-by-one.
●Artificial hemoglobin-containing solutions increase HGB without an increase in RBC count. (See "Oxygen carriers as alternatives to red blood cell transfusion".)
Spurious decrease in the MCHC — A true decrease in the MCHC can be seen in iron deficiency anemia. However, a high glucose concentration (due to hyperglycemia in the patient or contamination of the blood sample with an intravenous solution containing glucose) causes temporary osmotic imbalance in the diluted RBC in the analyzer chamber, causing them to swell with water for a short period, spuriously increasing the mean corpuscular volume (MCV) and decreasing the MCHC.
Platelet error flags — Platelet error flags by automated instruments are important because they may indicate an erroneous platelet count. Examples include electronic noise, cell cytoplasmic fragments, and microcytic RBCs, all of which can cause a spurious increase in the platelet count. The platelet count can be spuriously decreased when giant platelets are present, since such large platelets may be omitted from the platelet count because of their extreme size. As an example, platelets can be larger than red cells in some patients with the May-Hegglin anomaly. (See 'Platelet count and size' above and "Diagnostic approach to the adult with unexplained thrombocytopenia", section on 'Pseudothrombocytopenia'.)
Verification of the platelet count by slide estimate is usually the first step if flags occur, and, if necessary, a manual phase contrast microscopy method is used. (See "Evaluation of the peripheral blood smear".)
Leukocyte counting errors — Interferences at the lower size/volume threshold of WBC counting can be an important clue to the presence of artifacts. These include sample clotting, platelet clumping or giant platelets that can lead to spuriously low platelet counts, high WBC counts, and high percent lymphocytes. Most instruments flag significant particle interference in this region.
Most instruments also flag the automated leukocyte differentials when significant interferences are identified. A slide review is usually necessary to validate the instrument results and detect the presence of abnormal cells (eg, atypical lymphocytes, blasts).
Platelet clumping with pseudothrombocytopenia — Platelet clumping can occur, either due to in vitro blood clotting or in association with an EDTA-dependent platelet agglutinin (picture 5). Often these platelet clumps are so large that they are beyond the threshold for particle size included in the platelet histogram, leading to spuriously low platelet counts. These platelet clumps are the same size as small WBC and can spuriously increase the WBC count. Nucleated RBCs and non-lysing RBCs show a similar interference pattern.
If a slide examination confirms the presence of platelet clumps, it is likely that the platelet count is spuriously low and that the WBC count is spuriously high. If clotting of the blood sample is present, the sample should be redrawn. A correct platelet count can often be obtained for patients with an EDTA-dependent platelet agglutinin by collecting the blood in citrate anticoagulant (and multiplying the results by the dilution effect from the liquid anticoagulant), or by using a heparinized blood sample, in which case there is no need to correct for dilution.
SELECTED AUTOMATED HEMATOLOGY INSTRUMENTS — There are a number of commercially available automated hematology instruments, which differ in the methodologies used for determination of the individual elements of the complete blood count (CBC). Some of these are listed below.
Beckman Coulter — A number of Beckman Coulter models are available [55-57]. The Beckman Coulter DxH900/DxH800/LH780/LH750 report the white blood cell (WBC) count, red blood cell (RBC) count, hemoglobin (Hgb), hematocrit (Hct, calculated), mean corpuscular volume (MCV), mean corpuscular hemoglobin (MCH, calculated), mean corpuscular hemoglobin concentration (MCHC, calculated), red cell distribution width (RDW), platelet count, and mean platelet volume (MPV). A five-part differential includes percent and absolute number of total neutrophils, total lymphoid cells, monocytes, eosinophils, and basophils. When specifically ordered, reticulocytes (percent and absolute number) are analyzed and reported, as well as an immature reticulocyte fraction (IRF). The DxH900, DxH800, and LH750/LH780 also enumerate nucleated red blood cells (NRBCs), a feature not present on earlier instruments, using the combination of impedance, conductivity, and laser light scatter. The DxH900 provides a monocyte distribution width (MDW), which has been used as an early sepsis indicator [58].
The LH750/LH780 uses electric impedance technology to enumerate WBC, RBC, and platelet counts, and to determine RBC and platelet volume. The RBC/platelet aperture centers cells by virtue of its architecture (very long and narrow). In addition, sweep flow and pulse editing technology is employed. Hemoglobin is determined by CN-methemoglobin colorimetry. The five-part differential (lymphocytes, neutrophils, monocytes, eosinophils, basophils) is generated by analyzing events in a flow cell with three different technologies: volumetric impedance using direct current (related to cell size or volume), conductivity using high frequency electromagnetic energy (signal related to internal cell complexity), and laser light scatter (related to both cell size and structure). Approximately 8000 events are classified using these methods. Data are plotted in a three-dimensional matrix (VCS), and clusters are identified as specific cell populations. If the populations are not in expected locations, or a significant number of events are detected outside of the population clusters, flags are generated, indicating that abnormal cells may be present, as well as a need for manual slide review [59].
The DxH900 and DxH800 enumerate NRBCs using a whole blood sample that is diluted and lysed to remove non-nucleated cells and is then directed through the flow cell. Seven parameter measurements are captured characterizing size, shape, and morphology, and these data are analyzed using proprietary algorithms. This analyzer features high-definition digital signal processing and flow cytometric digital analysis using volume, conductivity and multi-angle light scatter, and is notable for a trackless magnetic transport system that contains no exposed moving parts. Reports on the evaluation of NRBCs in neonates using the DxH800 have shown better correlation with the reference manual technique compared to other analyzers [56,60]. The LH500 reports the presence of NRBCs as a flag but does not specifically enumerate them.
Siemens Advia — The Advia 120/2120 reports the WBC, RBC, Hgb, Hct (calculated), MCV, MCH (calculated), MCHC (calculated), RDW, platelet count, and MPV [61]. Other parameters include directly measured RBC hemoglobin content (mean = CH, SD = HDW) and directly measured RBC hemoglobin concentration (mean = CHCM, SD = CHDW) [28,62]. Reticulocyte parameters include the reticulocyte count and percentage, the reticulocyte hemoglobin content (CHr), the reticulocyte hemoglobin concentration (CHCMr), and the mean corpuscular volume of reticulocytes (MCVr). Reticulocyte parameters for laboratory use-only include RDWr, HDWr, and CHDWr.
A six-part differential (figure 4) includes percent and absolute number of neutrophils, lymphoid cells, monocytes, eosinophils, basophils, and large unstained cells (LUC, which includes reactive lymphocytes, blasts, and other abnormal cells). When specifically ordered, reticulocyte percent and absolute number are reported, as well as an immature reticulocyte fraction (IRF). Reticulocyte MCV and RDW, as well as directly measured reticulocyte hemoglobin content and concentration parameters, similar to those for mature erythrocytes, are also available. Instrument flags indicate the presence of potential interferences with individual parameters.
The Advia instruments count RBCs by electronic impedance, but measure their volume by two-dimensional laser light scatter, after sphering and fixing the erythrocytes. A RBC cytogram computes the volume and Hgb concentration for each individual erythrocyte. From this information, RBC volume and Hgb concentration curves can be constructed. Hgb concentration is also determined by colorimetry with a sodium lauryl sulfate-hemoglobin derivative.
RBC volume and RBC Hgb concentration histograms and smoothed curves, as well as the MCV and RDW, are reported. In addition, a number of erythrocyte research parameters can be obtained on alternate displays. These include the CHM (mean of directly measured Hgb content), the HDW (SD of the Hgb content, or RBC hemoglobin distribution width), the CHCM (mean of directly measured Hgb concentration) and the CHDW (SD of the Hgb concentration, or RBC hemoglobin concentration distribution width). The Advia 120 has been used to quantify schistocytes and has shown a high negative predictive value [63]. The percentage of hypochromic red cells (%HYPO) and reticulocyte hemoglobin content (CHr) have been shown to be a useful tool in diagnosing iron deficiency [64].
For platelet enumeration and sizing, the platelets are sphered, fixed, and counted by laser light scatter in two dimensions, similar to RBC analysis. This technique allows discrimination between platelets on the one hand and RBCs (microcytic RBCs, RBC fragments, and ghosts) and WBCs on the other, because the latter sets of particles have a different refractive index than platelets. Platelet volume histograms and smoothed curves can be constructed from this plot, and the platelet count, MPV and SD or CV of the volume (PDW) derived.
The WBC is determined by duplicate methodologies. The WBCs are fixed and counted in the optical peroxidase channel as well as enumerated in the dual laser light scatter basophil flow cell with stronger lysing reagent. If interference at the lower counting threshold is noted in the peroxidase channel due to unlysed RBCs, the WBC count from the basophil channel is used.
A six-part WBC differential is performed by cytochemical and laser light scattering technology. After lysing the RBCs, WBCs are fixed and stained for myeloperoxidase using 4-chloro-1-naphthol substrate at 50 to 70°C. Stained cells are hydrodynamically focused through a flow cell illuminated by tungsten-halogen light, where light scatter as well as light absorption of individual cells is measured with a pair of photodetectors. The data are displayed in a two-dimensional scattergram, plotting cell size (light scatter) versus peroxidase activity. In addition to discriminating between neutrophils, lymphocytes, monocytes, and eosinophils, a sixth population, large unstained cells, is enumerated (figure 4).
Because myeloperoxidase activity is employed in this method for obtaining the WBC differential, neutrophils in patients with myeloperoxidase deficiency will be enumerated under the population of "large unstained cells." (See "Myeloperoxidase deficiency and other enzymatic WBC defects causing immunodeficiency", section on 'Myeloperoxidase deficiency'.)
Basophils are classified in a separate "basophil/lobularity" channel. Using a strong lysing agent, both RBCs and WBCs are lysed, leaving only bare nuclei, except for basophils, which are resistant to this agent. The remaining nuclei/cells are analyzed with two-angle laser light scatter. Basophils appear the largest, while the smaller bare nuclei can be analyzed as well. The mononuclear round cells scatter the least light, while lobulated cells scatter the most, generating a lobularity index. This information is used to detect the presence of blasts and mononuclear left-shifted neutrophils.
Abbott — The newest Abbott hematology analyzer, the Alinity hq, employs optical scatter and fluorescence methods to analyze blood cells and reports a six-part WBC differential including immature granulocytes, nucleated RBCs, percent reticulated platelets, and reticulocyte hemoglobin (ret-He) content [65]. The previously released Cell-Dyn Sapphire analyzer reports the WBC (optical), RBC (impedance), RBC (optical), Hgb, Hct (calculated), MCV, MCH (calculated), MCHC (calculated), RDW, platelet count (optical and impedance), MPV, and PDW [66]. The machine also calculates a plateletcrit, the percent of the sample's volume (ie, cells plus plasma) occupied by platelets. The differential includes percent and absolute number of neutrophils, lymphoid cells, monocytes, eosinophils, and basophils. Nucleated red blood cells are enumerated [67], and a nucleated RBC/WBC ratio is given. A leukocyte viability fraction is also given. When specifically ordered, reticulocyte percent and absolute number are analyzed and reported, as well as an immature reticulocyte fraction (IRF). In addition, research screens on the instrument enumerate blasts, atypical lymphocytes, and immature granulocytes. Instrument flags indicate the presence of potential interferences for individual parameters.
RBCs on the Cell-Dyn Sapphire are counted by both impedance and laser light scatter, using hydrodynamically focused and sphered RBCs for both measurements. The impedance measurement is used for RBC count and size parameters. A two-dimensional optical RBC count serves to validate the impedance count. Using multi-angle light scatter and fluorescence, a three-dimensional RBC differential analysis has been generated allowing automated analysis of normal and abnormal red blood cells [68].
Hemoglobin is measured colorimetrically on both the Alinity hq and the Cell-Dyn Sapphire using a cyanide-free reagent (methemoglobin method with imidazole as the heme-ligand). Sodium hydroxide and lauramine oxide dissolve all cell particles and lipids and destroy bilirubin, minimizing interferences.
Platelets on the Cell-Dyn Sapphire are enumerated by both optical and focused impedance methods. The optical platelet count and volume is determined from two-dimensional analysis of low and high angle light scatter measurements. Platelets are first plotted using 90 degree versus 7 degree scattering. The ascending angle of the platelet population is determined and the histogram is formed perpendicular to the axis of the angle. High and low thresholds are set to isolate the platelets from large and small cells and fragments. A second histogram is constructed along the main body of this refined platelet population to determine the upper threshold separating platelets from microcytic RBCs. Interference by microcytic RBCs or RBC fragments and WBC cytoplasmic fragments is eliminated, since these particles have a different refractive index than platelets. Once the platelet particles are identified, the platelet count, MPV, PDW, and PCT are determined.
The impedance platelet count appears as a lognormal distribution and serves to validate the optical count. Flagging of the impedance platelet count occurs when a high percentage of particles are noted near either the lower or upper counting threshold.
Optional platelet enumeration by immunologic analysis is available on the Sapphire. Platelets are identified by cluster analysis on a two-dimensional plot of fluorescent monoclonal antibody CD61 (directed against glycoprotein IIIa) reactivity versus high forward angle light scatter. The CD61 immunoplatelet method correlates the closest with the flow cytometry reference method for platelets, compared with impedance and optical platelets on the Sapphire [69]. Using the CD61 method as a gold standard compared with impedance and optical methods results in platelet undertransfusion and overtransfusion [70].
WBCs are counted by laser light scatter. Events are considered "WBC" using analysis in three dimensions: size (0 degree scatter), complexity (7 degree scatter), and DNA fluorescence. Only events that exceed lower thresholds for these three measurements qualify for being counted. Thresholds are set low enough to capture all true WBCs, but exclude significant noise. Further analysis eliminates nucleated RBCs, RBCs resistant to lysis, platelet clumps, and other interfering materials from the final white blood cell count.
For the leukocyte differential, blood is diluted with a leukoprotective diluent that maintains the leukocytes close to their native state and includes propidium iodide (PI) to enumerate nucleated RBCs and assess leukocyte viability. Mature erythrocytes are rendered transparent to laser light and do not produce light scatter under these conditions. The WBC dilution is hydrodynamically focused through the sensing region of a laser beam. Laser light scatter is measured at multiple angles, including (figure 5):
●0 degree measurement is related to cell size (axial light loss from incident ray)
●7 degree measurement is related to internal complexity
●90 degree measurement is related to lobularity
●90 degree depolarization measurement is specific for eosinophilic granules
In addition, red fluorescence is measured, which is proportional to the amount of DNA in each cell that has become membrane compromised, such that the PI reagent has access to the cell nucleus. Up to 10,000 events are classified, and the measurements are stored in list mode, allowing the algorithm to work through the data incrementally.
The first step involves a two-dimensional plot of 0 degrees versus fluorescence, to separate the cells which pick up the DNA stain, either nucleated RBCs or WBCs that are fragile or leaky. Based on the 0 degree size, these fluorescent cells are separated into nucleated RBC (red) or fragile white cells (white). The nucleated RBC particles are eliminated from further analysis, while the fragile WBCs are identified by subpopulation in further steps.
In the next step, mononuclear (blue) and polynuclear cells (orange) are separated by a plot of 7 degree versus 90 degree light scatter.
The polynuclear (orange) cells are analyzed with polarized laser light, so that eosinophils (green) can be identified within this fraction by the ability of their granules to depolarize light at 90 degrees.
The mononuclear cells (blue) are plotted in two dimensions (7 degree versus 0 degree light scatter) to define the basophil region (white). Once the basophil region is defined, floating thresholds separate the lymphocytes (blue) and monocytes (purple) using a histogram along the 0 degree size axis with the basophils removed. The remaining events below the lymphocytes (gray) are non-WBC (and non-nucleated RBC) events, and if in sufficient number, may indicate significant interference from RBCs resisting the lysing reagent.
Finally, a plot of fluorescence versus 0 degree size is reviewed again, with classified leukocytes appropriately colored. Non-viable leukocytes (highly fluorescent) can be identified, and a viability ratio is calculated.
The plots of the various populations are analyzed for abnormal populations of cells by proprietary algorithms. Flags are reported if abnormal populations are identified. The neutrophil population is analyzed for the presence of bands and immature granulocytes, the mononuclear population is analyzed for the presence of blasts, variant lymphocytes are identified using data on both viable (non-fluorescent) and non-viable (highly fluorescent) lymphocytes.
Sysmex — The Sysmex XN series report the WBC (optical/fluorescent), RBC (impedance), Hgb, Hct (from measurements of RBC pulse heights), MCV, MCH (calculated), MCHC (calculated), RDW-SD, RDW-CV, platelet (impedance), platelet (alternate, optical/fluorescent), MPV, PDW, P-LCR (the percent of larger platelet) and a plateletcrit (PCT) [71,72]. A seven-part differential including percent and absolute number of neutrophils, lymphoid cells, monocytes, eosinophils, basophils, immature granulocytes and NRBCs, and hemoglobin equivalent parameters is also reported [73]. The XN-30 quantifies malaria parasites within erythrocytes using a violet laser and flow cytometry [74].
Hematopoietic progenitor cells (HPC) are reported as a research parameter. Some authors have reported that the XN-HPC is an effective surrogate marker for the CD34+ cell count in peripheral blood apheresis products used for allogeneic hematopoietic stem cell transplants [75]. When specifically ordered, reticulocyte percent and absolute number are analyzed and reported, as well as the reticulocyte hemoglobin (RET-He), and an immature reticulocyte fraction (IRF), which is the sum of the moderately and highly fluorescent reticulocytes (MFR+HFR). An immature platelet fraction (IPF) also is available off of the fluorescent platelet channel.
XE series instruments use impedance apertures for RBC and platelet counting and sizing, with hydrodynamic focusing and back sweep flow technologies. The RDW is expressed in terms of both SD and CV. The impedance platelet count is fitted to a lognormal distribution; flagging of this count occurs when a high percentage of particles is noted near either the lower or upper counting threshold. In addition, an optical/fluorescent platelet count is determined when the impedance platelet count is flagged, or when the platelet count is less than 30 x 109/L (30,000/microL). A computerized switching algorithm will give a best platelet count if both methods are enabled on the analyzer. Hemoglobin is determined colorimetrically using sodium lauryl sulphate.
The leukocyte count and differential are performed using a combination of DC and RF measurements, as well as laser light scatter (forward and side) and fluorescence with a polymethine DNA/RNA histone dye. XE instruments contain a semiconductor diode laser with a short oscillation wavelength (630 nm), measuring fluorescence in the near red region.
Four different lysing reagents are employed. First, RBC are lysed with a strong acid hemolytic reagent, removing even lysis-resistant RBC as well as platelet aggregates. Basophils are resistant to lysis, and can be separated from other WBC in a side versus forward light scattergram. The WBC count is measured as the sum of the basophil and other leukocyte events.
The second step involves a lysing reagent that acts on the leukocyte membrane to allow dye passage. The fluorescent polymethine dye stains DNA, RNA, and histones. A plot of side light scatter versus fluorescence separates lymphocytes, monocytes, neutrophils plus basophils, and eosinophils. Total neutrophils are computed by subtracting the number of basophils measured in step one. In the third step, a lysing reagent is used to completely expose and shrink the nuclei of the nucleated erythrocytes. The WBCs, however, retain their cytoplasm, and are semi-permeabilized. The polyamine dye stains the small nucleated RBC nuclei and the larger WBC nuclei as well as the WBC cytoplasmic organelles. A plot of side fluorescence versus forward light scatter separates nucleated RBC from the WBC and the non-nucleated RBC ghosts and membrane fragments. An "IMI (immature cell information) channel" is used to detect small numbers of immature granulocytes, blasts, and hematopoietic progenitor cells. Quantitation of immature granulocytes has been used by some authors as a predictor of infection comparable to the absolute neutrophil count [76]. The granularity index has been reported as a useful indicator of neutrophils with toxic granulation on the XE-5000 [77].
Cells with a higher content of lipid are lysed with a fourth hemolytic reagent, while immature cells retain their membranes. The sample is analyzed in an impedance aperture with direct (DC) and radiofrequency (RF) current. In normal samples, no intact cells are seen in the IMI area. The information from the IMI channel is used to generate flags for left shift, immature granulocytes, and blasts, and to enumerate hematopoietic progenitor cells. Additional flags for abnormal cell types occur as abnormal cell populations are detected in the plot of side light scatter versus fluorescence in step two.
Reticulocytes and the IRF are determined using the same polymethine dye capable of staining reticulocyte RNA in a special diluent. A plot of fluorescence versus forward light scatter separates mature RBC from reticulocytes, and enables enumeration of highly fluorescent reticulocytes. Small size platelet and highly fluorescent WBC do not interfere. When the impedance platelet count shows interference, the platelet count is determined from a log-scaled version of this plot. The fluorescent platelet RNA enables separation of platelet from RBC fragments. Fragmented red blood cells can also be quantitated on some instruments; this is performed using the reticulocyte channel in which RBC fragments are gated from the RBC area (small size, low RNA content) [78]. RBC extended parameters and reticulocyte parameters show utility in disorders of iron deficiency, anemia due to chronic renal disease, thalassemia, and schistocytes [79,80].
SUMMARY
●Various methods have been employed to automate hematologic testing for the following:
●Complete blood count (CBC), including red cell indices, red cell distribution width, and the reticulocyte count. (See 'The red blood cell volume histogram' above and 'Red cell distribution width' above and 'Alternative measurements of hemoglobin concentration' above and 'Automated counting of reticulocytes' above and 'Spurious increase in the MCHC' above and 'Spurious decrease in the MCHC' above.)
●White blood cell count and differential, including "flags" for abnormal cells (eg, blasts). (See 'Automation of the white cell differential' above and 'Leukocyte counting errors' above.)
●Platelet count, including measurement of early (reticulated) platelets. (See 'Platelet count and size' above and 'The optical platelet count' above and 'Platelet error flags' above and 'Platelet clumping with pseudothrombocytopenia' above.)
ACKNOWLEDGMENT — We are saddened by the death of Stanley L Schrier, MD, who passed away in August 2019. The editors at UpToDate gratefully acknowledge Dr. Schrier's role as Section Editor on this topic, his tenure as the founding Editor-in-Chief for UpToDate in Hematology, and his dedicated and longstanding involvement with the UpToDate program.
