The Blood Film and Count (Part 2)

Haemopoietic cells

Peripheral blood cells are produced in the bone marrow. Their precursors are referred to as haemopoietic cells (Fig. 1.12). The only significant function of haemopoietic cells is the production of mature end cells. Recognizable haemopoietic precursors are present in the circulating blood of healthy subjects but, except in the neonatal period and during pregnancy, they are quite uncom- mon and are not often noted in a blood film. They are much commoner in patients with leukaemia or other haematological disorders and in patients with severe infection or other serious systemic diseases.

Myeloblasts

Myeloblasts (Fig. 1.13) are very rare in the blood of healthy subjects. They are larger than lymphocytes but often smaller than monocytes. They have a high nucleocytoplasmic ratio and scanty to moderate amounts of cytoplasm, which varies from weakly to moderately basophilic. (Basophilic in this context in- dicates a blue colour consequent on the uptake of basic dyes.) The nucleus is approximately round, nuclear chromatin is dif- fuse and nucleoli may be apparent. In patients with leukaemia and related disorders, the cytoplasm may contain small numbers of azurophilic granules or other inclusions or vacuoles (see page 91). Myeloblasts are precursors of neutrophils, eosinophils and basophils.

Promyelocytes

Promyelocytes (Fig. 1.14) are rare in the blood of healthy people. They are larger than myeloblasts with more plentiful cytoplasm and consequently a lower nucleocytoplasmic ratio. The cyto- plasm is more basophilic than that of a myeloblast and contains azurophilic (pinkish-purple) primary granules. Sometimes there is a more lightly staining zone in the cytoplasm adjacent to the nucleus, which represents the Golgi apparatus, where granules are produced. The nucleus is round or oval, is usually eccentric, shows some chromatin condensation and has a

Fig. 1.12  A diagram showing the  relationship of haemopoietic precursors to each  other and  to the  end  cells  into which they differentiate. Proliferation of cells  occurs simultaneously with maturation or differentiation so that one  myeloblast is likely to give rise  to 16 mature granulocytes and  one  proerythroblast to 16 red cells.  Myeloblasts, promyelocytes and  myelocytes are all cells capable of cell  division or mitosis. Metamyelocytes and  all later cells  are non-dividing cells.  All red cell  precursors with the exception of late  erythroblasts are dividing cells.  Myeloblasts differentiate not  only  into neutrophils, as shown in the  diagram, but also  into eosinophils and  basophils.

Fig. 1.13  A blood  film of a patient showing a myeloblast and  a neutrophil. The  myeloblast has  a high  nucleocytoplasmic ratio, a diffuse chromatin pattern and  a single nucleolus. The  neutrophil is hypogranular

Fig. 1.14  A promyelocyte showing a lower nucleocytoplasmic ratio than that of a myeloblast, an eccentric nucleus, azurophilic granules and  a Golgi  zone  to the  left  of the  nucleus

visible nucleolus. Because they have no specific (lineage- associated) granules, promyelocytes, which are precursors of neutrophils, eosinophils or basophils, cannot generally be distin- guished from each other.

Fig. 1.15  A neutrophil myelocyte showing a smaller cell  than a promyelocyte with some condensation of nuclear chromatin and  no visible nucleolus. On  microscopic examination it is apparent that such cells  have primary and  secondary granules with different staining characteristics.

Myelocytes

Myelocytes (Fig. 1.15) are uncommon in the blood of healthy subjects except in the neonatal period and during pregnancy. They are smaller than promyelocytes. They have not only azuro- philic or primary granules but also secondary granules character- istic of specific lineages, i.e. neutrophilic, eosinophilic or basophilic granules. The myelocyte nucleus is round or oval and shows chromatin condensation; no nucleolus is apparent.

Metamyelocytes

Small numbers of neutrophil metamyelocytes (Fig. 1.16) are present in the blood of healthy subjects. Basophil and eosinophil metamyelocytes are not seen in the blood of healthy subjects. Metamyelocytes have similar characteristics to myelocytes but differ in that the nucleus is indented, U-shaped or C-shaped and the primary granules are usually no longer apparent.

Band cells

Neutrophil band forms (Fig. 1.17) are present as a minor popula- tion in the blood of healthy people. They are intermediate in characteristics between metamyelocytes and mature

Fig. 1.16  A neutrophil metamyelocyte between two  segmented neutrophils. The  nucleus is indented.

Fig. 1.17  A neutrophil band  form  (left) compared with a segmented neutrophil (right).

neutrophils. The nucleus has an irregular shape with some paral- lel edges so that it resembles a band or ribbon. It differs from a mature or segmented neutrophil in that the nucleus is not divided into distinct lobes or segments. Eosinophil and basophil band forms are quite uncommon.

Nucleated red blood cells

Nucleated red blood cells (NRBC) or erythroblasts (Fig. 1.18) are present in very small numbers in healthy people, except during the neonatal period. Those which are most likely to be released

Fig. 1.18  Three nucleated red blood  cells  (NRBC)  showing a small densely staining nucleus and  cytoplasm which is pink because of the  presence of haemoglobin

into the blood stream are late erythroblasts. They can be readily recognized because the cytoplasm is at least partly haemoglobin- ized giving them a pinkish or lilac tinge. NRBC have a superficial resemblance to lymphocytes but can be distinguished from them not only by the colour of the cytoplasm but also by the lower nucleocytoplasmic ratio and the denser, more homogeneously staining nucleus.

The blood count

Haematology laboratories not only examine blood films. They also perform various measurements relating to the haemoglobin content of the blood, the characteristics of red cells and the number of red cells, white cells and platelets. These measure- ments are collectively referred to as a blood count or full blood count (FBC). During illness, abnormalities can develop in any of the cells in the blood. The purpose of performing a blood count and examining a blood film is to detect quantitative and qualitative abnormalities in blood cells. Their detection often helps in diagnosis and in the treatment of the patient.

Haemoglobin concentration

If red cells are lysed, the haemoglobin is released from the red cells and forms a solution in the plasma. The haemoglobin con- centration (Hb) can be measured biochemically by light absorp- tion at a specified wave length after a chemical reaction which converts haemoglobin to cyanmethaemoglobin or to lauryl sul- phate haemoglobin. Hb is measured in either grams per decilitre (g/dl) or grams per litre (g/l). A fall in the Hb is referred to as anaemia.

Haematocrit or packed cell volume

An alternative way of detecting anaemia is to centrifuge a tube containing an aliquot of blood and measure the proportion of the column of blood which is occupied by the red cells. Nowadays an equivalent measurement is made by various automated instru- ments using a quite different principle to get the same infor- mation. This test is called a packed cell volume (PCV) or a haematocrit (Hct). Some haematologists use these two terms interchangeably while others used PCV to refer to a measure- ment made after centrifugation and Hct for an estimate made by an automated instrument. This measurement is expressed as a decimal percentage, i.e. as litres/litre (e.g. 0.45).

Cell counts

Traditionally blood cells were counted by diluting a small quan- tity of blood in a diluent which could also stain the cells or, if white cells or platelets were to be counted, could lyse the more numerous red cells. The diluted blood was placed in a counting chamber of known volume and the number of cells present was counted microscopically. Such a method of counting blood cells is very labour-intensive and not suited to the large number of blood counts needed in modern medical practice. Nowadays blood cells are counted by large automated instruments.

A stream of cells in a diluent passes through a sensing zone. They are sensed either because they pass through an electric field or because they pass through a beam of light. Each cell passing through the sensing zone generates an electrical impulse, which can then be counted. Red cells are both relatively large and relatively numerous and so can be readily counted. White cells can be counted by lysing the more numerous red cells or by altering the red cells in some way so that they are ‘invisible’ to the instrument. Platelets are distinguished from other cells by their smaller size. Cell counts are expressed as the number of cells in a litre of blood. The red blood cell count (RBC) is ex- pressed as a number x 1012 per litre (e.g. 5 x 1012/l). The white blood cell count (WBC) and platelet count are expressed as a number x 109 per litre (e.g. 7.5 x 109/l and 140 x 109/l). A white cell count of 7.5 x 109/l means that there are 7 500 000 000 cells in a litre of blood.

Red cell indices

Red cells can vary in their size and in the amount of haemoglo- bin contained in an individual cell. Abnormalities in both these cell characteristics are common in certain inherited abnormal- ities and when people are sick. Diagnostically useful information can be obtained by measuring them. Traditionally the size of red cells was estimated by dividing the PCV by the number of cells in the blood to give a mean cell volume (MCV). The haemoglobin content of individual cells was estimated by dividing the Hb by the RBC to give a mean cell haemoglobin (MCH). The Hb of individual cells was estimated by dividing the Hb by the PCV to give a mean cell haemoglobin concentration (MCHC). Nowadays, not only is the PCV estimated electronically but the size of a red cell can be calculated from the height of the elec- trical impulse which is generated when the cell passes through a light beam or through an electrical field. As the automated instruments also measure the total Hb of the blood, it is a simple matter for the red cell indices to be produced automatically as part of the blood count. Instruments can be designed to measure the MCV and calculate the PCV/Hct from the MCV and the RBC or, alternatively, to measure the PCV/Hct and calculate the MCV from the PCV/Hct and the RBC. The formulae which relate the various red cell indices to each other are as follows:

Normal ranges

In order to interpret blood counts it is necessary to know what is normal. This is usually done by reference to either a normal range or a reference range. A reference range is more strictly defined than a normal range but both represent the range of test results which would be expected in healthy people of the same age and sex (and, if relevant, of the same ethnic origin) as the person being investigated. Conventionally, both types of range are expressed as the central 95% of test results that would be expected in healthy people. The reason for excluding the top 2.5% and the bottom 2.5% is that there is usually an overlap between test results of healthy people and of those who are sick. A 95% range has been chosen to avoid either classifying too many healthy subjects as abnormal or missing relevant abnor- malities in patients who are sick. It is clear that for any one test 5% of healthy subjects will have results falling outside the ‘nor- mal’ range. Conversely, a patient who is sick may have a test result which is abnormal for him or her but which is still within the normal range. For example, a man may have a large gastroin- testinal haemorrhage, causing his Hb to fall from its normal level of around 16 g/dl to 14 g/dl. The latter 14 g/dl is within the range expected for a healthy adult man but for this particular patient it is abnormal. This is because the range of test results expected in a group of healthy people is much wider than the range expected if the same test is repeated day after day in the same person. Usually we have no way of knowing what is ‘nor- mal’ for a particular individual and so we have to resort to comparing his or her test results with a normal range.

The statistical distribution of test results differs for different tests. Many tests, e.g. the Hb, show a normal or Gaussian distri- bution. This means that if the distribution of the test results is plotted on graph paper a bell-shaped curve is obtained (Fig. 1.19a). If this is so, the 95% range can be calculated by estimat- ing the mean ± 2 standard deviations. Other test results, e.g. the WBC (Fig. 1.19b), have a skewed distribution which only be- comes bell-shaped if the test results are plotted on logarithmic graph paper. Test results with this type of distribution require special statistical treatment to derive the normal range.

Fig. 1.19  Smoothed histograms showing (a) the  normal distribution of Hb and  (b) the  log normal distribution of the  white cell  count

Some normal ranges applicable to healthy people are shown in Tables 1.2–1.4. However, it should be noted that the test results for some haematological variables, e.g. the MCV, vary according to the method of measurement and it is desirable for laboratories to derive their own normal ranges for their own automated instruments by obtaining blood samples from a large number of healthy people. In the case of children, it is always difficult to obtain blood samples from large numbers of healthy individuals of various ages. As a consequence, published normal ranges for children are not as reliable as those for adults.

How to examine a blood film

Blood films should be examined in a systematic way. First the film should be examined without using the microscope, to make sure it is well spread (not too thick, too long or too short) and that its staining characteristics are normal. A film that is a deeper blue than other films stained in the same batch is usually indicative of an increase in the concentration of plasma proteins. This can be diagnostically important since it is often caused by multiple myeloma (a plasma cell malignancy) (see page 87) or by chronic inflammatory disease.

Table 1.2  Normal ranges for healthy Caucasian adults.

RDW, red cell distribution width; HDW, haemoglobin distribution width. The differential white cell counts and the platelet counts are for Technicon H.1 series automated instruments. The ranges are wider for manual differential counts, particularly for monocytes, eosinophils and basophils. Platelet counts are very dependent on the method used for 

counting and should be assessed only in relation to a normal range derived for the instrument or method in use. 

* Coulter S Plus IV. 

† Technicon H.1 series.

Next the film is examined microscopically at low power (e.g. with a x25 objective) so that a large part of the film can be scanned rapidly to detect any abnormal cells present in small numbers. Finally the film is examined at a higher power (e.g. with a x40 or x50 objective) so that the detailed structure of cells can be assessed. The great majority of films can be evaluated perfectly adequately without using high power (i.e. a x100 oil immersion objective). High power can be reserved for making a detailed assessment of films that show significant abnormalities requiring further assessment. In examining a film be sure to look

Table 1.3  Normal ranges for Afro-Caribbean and  Africans for those haematological variables where the  ranges differ  from  those of Caucasians.

It should be noted that the lower RBC, Hb, PCV and MCV observed in Afro-Caribbean and Africans are likely to be consequent on a high prevalence of thalassaemia trait and haemoglobinopathies rather than on other ethnic differences. It is therefore appropriate to use Caucasian reference ranges for red cell variables for Afro-Caribbean and Africans.

Table 1.4  Approximate 95%  ranges for red cell  variables and  for automated* total and  differential white cell  counts for Caucasian infants and  children.

* Ranges will be wider for manual differential counts than for automated counts. † The lymphocyte count is up to 8 x 109/l in 2-year-olds, up to 5.5 x 109/l in 3- and 4-year-olds and up to 4.5 x 109/l in 5-year-olds.

specifically at red cells, white cells and platelets so that no abnormality is inadvertently overlooked. Be sure to look at the edges and tail of the film where abnormal cells may be found. 

Finally, decide if a differential count is needed. Nowadays this will often have been performed by an automated instrument but you may need to verify its accuracy and in leukaemia you may need to carry out a manual differential count, i.e. one performed with the aid of a microscope.

Learning to look at blood films 

When learning to recognize cells for the first time it is useful to compare cells seen down the microscope with photographs. Examining films on a double-headed microscope with an ex- perienced laboratory worker is also very valuable. To learn to recognize high and low WBC and platelet counts, start by com- paring the film appearance with the count on an automated instrument. After you have had some experience try to estimate what the count will be before you look at the test results. Later you will need to be able to do this fairly accurately so that you can recognize erroneous instrument counts. Similarly, start by looking at films with high and low MCVs and compare the size of the red cells with neutrophils and lymphocytes until you can recognize large and small red cells. When you have had some experience try to estimate the approximate MCV before you look at the test results. Eventually you will be able to judge the MCV, at least to within 5–10 fl. 

Recognizing problems with the blood sample 

Before carrying out a detailed assessment of a blood film it is important to detect any abnormal characteristics of the speci- men which might interfere with your assessment of the film or with the accuracy of the automated count. The most common problem is storage artefact (Fig. 1.20). This occurs when blood has been at room temperature for a day or more before reaching the laboratory. The red cells turn into echinocytes, i.e. their shape alters so that the surface is covered with numerous short, regular projections. This process is also known as crenation.

Fig. 1.20  Storage artefact. The  red cells  are crenated, a lymphocyte (right) has  a fuzzy  outline and  one  of the  two  neutrophils (left) has  a nucleus which has  become round, dense and  homogeneous. (Compare the degenerating neutrophil with the  nucleated red cells  shown in Fig. 1.18.)

Some of the white cells develop fuzzy outlines or disintegrate entirely when the blood film is spread. The nuclei of neutrophils become dense, homogeneous and round and may break up into two or more round masses. It is important not to confuse these degenerating neutrophils with NRBC. They have a lower nucleo- cytoplasmic ratio and the cytoplasm is pink and slightly granular rather than reddish-brown. It is impossible to give any reliable opinion of films showing storage artefact. If the blood count is normal they can usually be ignored but if there is any reason to suspect a haematological abnormality a fresh blood sample must be obtained. A common cause of inaccurate blood counts is partial clotting of the specimen or aggregation of the platelets. Platelets may aggregate because they have been activated (i.e. the process of blood clotting has started) or because there is an antibody present in the plasma which leads to platelet aggregation in blood that is anticoagulated with ethylenediaminetetra-acetic acid (EDTA). Aggregated platelets form masses between the red cells, that may contain intact platelets (Fig. 1.21) or may be composed of totally degranulated platelets, which stain pale blue. Less often, partial- ly clotted samples contain fibrin strands, which are seen as pale blue or almost non-staining linear structures running between and deforming red cells (Fig. 1.22). Another in vitro artefact, less common than platelet aggregation but which can also lead to falsely low platelet counts, is platelet satellitism (Fig. 1.23).

Fig. 1.21  A platelet aggregate containing a mixture of intact and degranulated platelets.

Fig. 1.22  Fibrin strands passing between and  over  red cells.

Fig. 1.23  Platelet satellitism.

Less common artefacts which should be recognized are those due to accidental freezing or overheating of the blood specimen before it reaches the laboratory and the presence of lipid (fat) in the plasma. All these abnormalities cause anomalous blood counts. 

Interpreting blood films 

When assessing blood films, always note the age, sex and ethnic origin of the patient and keep in mind what would be normal for that individual. Also consider the clinical details so that you can look carefully for any specific abnormalities which might be relevant, keeping in mind that the clinical details may provide you with an obvious explanation for an abnormality you have noted. For example, if the clinical details were ‘alcohol excess’ you would not be surprised to find that the patient had macro- cytosis and you would go on to see if there were stomatocytes or any of the other abnormalities which could be caused by alcohol. Your report of these abnormalities would give the clinician very specific information which would help to confirm his/her clini- cal suspicion.