Beyond the numbers: Things your CBC machine won't tell you, but you need to know (Proceedings)

In the diagnostic work up of any case, one of the most frequently used tests is the complete blood count (CBC). The CBC generally includes the gross examination of the sample, the data generated by an automated analyzer and the microscopic evaluation of a blood smear. From the data collected, additional tests may be indicated and/or pursued. Increasingly powerful and easily (for a price of course) available table top hematology analyzers provide an avenue to produce seemingly accurate and precise data rapidly and efficiently. It is very easy to accept the data produced by the analyzer and insert it into an animal's workup. In many instances, one can elect to perform an additional test, choose not to perform a different test, make a diagnosis and begin treatment or change the differential diagnoses list all together. The lure of ease and speed often prove too great for busy clinicians and they use the information from the analyzer exclusively; however, there is a tremendous amount of data that can only be obtained by making a diagnostic blood smear and critically evaluating it using a high quality, diagnostic microscope.

The first step to generating accurate, usable data is to insure the sample was collected and handled in an appropriate manner. A fasting period of 12 hours will generally eliminate the risk of lipemia. Lipemia not only causes artifacts on the serum chemistry panel, but it can also affect the plasma protein, fibrinogen, hemoglobin, mean corpuscular hemoglobin concentration (MCHC) and erythrocyte morphology. A seemingly simple concept to adhere to is sample labeling. In larger practices, with multiple samples being processed by several people, the risk for a handling error increases. Using the appropriate tube is also important. Almost without exception, a vacuum tube with an anticoagulant is used for sample collection. The proper amount of blood must be added to the tube for optimal conditions to be achieved. Glass tubes are less likely to cause clot formation than are plastic tubes. If possible, allow the vacuum of the tube to draw the blood into the chamber, as they are often calibrated to obtain only as much blood as needed, thereby avoiding positive pressure and trauma to the cells. Cell trauma can also be reduced by using a larger bore needle, such as an 18 or 20 gauge to draw blood from a large, free-flowing vein.

For small animals, the anticoagulant of choice is ethylenediaminetetraacetic acid (EDTA), with the hematology tube being the priority tube for filling. It provides optimal retention of in vivo morphology as well as staining characteristics. If there are no contraindications, a liquid anticoagulant is preferred to lyophilized. An incomplete sample draw will decrease the hematocrit (HCT) via cell shrinkage and cause dilution of all cell lines, if liquid anticoagulant is used. Heparin may also be used if WBC clumping occurs. Heparin is not ideal as an anticoagulant for hematologic examinations for multiple reasons. Poor leukocyte staining occurs, likely because heparin binds to them. It is also more prone to platelet clumps forming, and it is not a permanent anticoagulant. A third choice for hematologic sample anticoagulation is citrate.

Once a sample is obtained, it is grossly evaluated within the tube. The tube should be rocked and rolled to check for macroscopically evident clumps or clots. If present, a new sample should be automatically be obtained and no further testing performed. If clinical suspicions suggest the presence of methemoglobinemia, a drop of blood can be placed on white filter paper to see if it is bright red. If the methemoglobin content is >10%, the blood with have a noticeable brown color when placed on white filter paper. If no abnormalities exist, the blood can then be placed into the hematology analyzer and two to four fresh smears can be made as soon as reasonably possible. When the smear is made, additional evaluation for macroagglutination can be performed. A properly made smear should have a head, body, feathered edge and an adequately large monolayer (Fig. 1).

Figure 1

At the same time a blood smear is made, two or more microhematocrit tubes should be prepared and spun in a centrifuge. A packed cell volume (PCV) will be generated to compare to the machine's calculated hematocrit (Hct). Likewise, a crude correlation of the buffy coat can be made to the machine derived white blood cell count (WBC) and platelet count. In addition, the column of plasma can be evaluated for evidence of hemolysis, icterus or lipemia that may affect how all of the data, including the CBC, are interpreted.

The microscopic examination begins with a low power (I prefer 20x) scan of the smear. A suggested pattern is shown in Figure 2.

Figure 2

This scan insures that no platelet (PLT) or leukocyte clumps are present. Clumping will falsely decrease the PLT count and the white blood cell count (WBC), respectively. Platelet clumping can also decrease the hematocrit, increase the mean platelet volume (MPV), and decrease the mean corpuscular volume (MVC) and red cell distribution width (RDW) on some machines. It is impossible to provide a firm, accurate platelet count on a clumped sample. A minimum number can be offered; however, some laboratories will not report any number on clumped samples. Leukocyte clumping affects the WBC in a similar fashion; however, it can also affect the differential counts, as one cell line may be preferentially clumped. Having freshly made smears available allows the clinician to evaluate morphology as close to in vivo as possible. This can be particularly useful for things such as neutrophil degeneration or vacuolation as well as hemoparasites such as Mycoplasma spp. that have a tremendous propensity to fall off of erythrocytes, making them impossible to identify in blood smears made from old blood. The low magnification scan also makes the identification of abnormal RBC arrangements easier. Rouleaux and microagglutination are both easy to appreciate at a lower magnification. Agglutination can falsely elevate the MCV if small clumps of cells are perceived as a single cell. When either RBC arrangement is noted, a saline agglutination test should be performed to differentiate the two.

Important information gathered by a low power scan that a machine will not find are microfilaria. Dirofilaria immitis, Dipetalanema sp. and Setaria sp. can also easily be noted on a low power scan. Likewise, potentially diagnostically significant cells present in low numbers can also be noted on a low power scan. Good examples include mast cells and megakaryocytes.

Higher (50x and 100x) magnification reveals a wealth of information that automated analyzers simple can not identify. Probably the most important things not identified are neutrophil maturity and toxic changes. Segmented neutrophils normally have five lobes. Any more and it is considered hypersegmented, indicating senescence. They should have smooth, pink to blue cytoplasm. Band neutrophils have no constrictions in the nucleus greater than 50% of the widest part. These are considered an immature form; however, may be seen in very low numbers normally. Metamyelocytes have a kidney shaped nucleus with a mild indentation and are slightly larger than band neutrophils. They are even more immature than band neutrophils and should not normally be seen in circulation. Myelocytes are generally the most immature neutrophil seen in circulation and have a round nucleus, with a chromatin pattern similar to the others in the neutrophil line, only less intensely staining. A left shifts is when there is an increase in immature neutrophils and indicates non-specific inflammation / demand for neutrophils. A regenerative left shift is when the total number of immature forms is increased, but still less than the total number of segmented neutrophils, in the face of a neutrophilia. A degenerative left shift is when the total number of immature forms outnumbers the segmented neutrophils, or an increase in immature neutrophils with a normal or decreased neutrophil count. This indicates that the bone marrow cannot keep up with demand and thus has resorted to releasing more and more immature forms. Toxicity is another key feature that must be assessed microscopically. Toxic changes are the morphologic changes seen that occur when production of neutrophils is markedly increased or when endotoxin is present, both of which result in abnormal maturation within the bone marrow. Döhle bodies represent the first sign of toxicity. These are blue staining inclusions in the cytoplasm, often at the periphery or near the nucleus, as seen in figure 3.

Figure 3

The next stage of toxicity is blue and foamy cytoplasm. Both must be present to be considered the second level of toxicity. Toxic granulation consists of fine pink granules in the cytoplasm and represent the most severe toxic change. This toxicity is rarely seen in small animals. This change should be accompanied by prominent examples of the first two stages of toxicity.

Similar to neutrophil morphology identification, lymphocyte sub-classification can be a problem. Atypical, reactive, large granular and neoplastic lymphocytes are not identified. Mature lymphocytes are roughly 1.5 to 2 times the size of a red cell, have an evenly stippled, finely stippled chromatin pattern and a uniformly high N:C ratio. They will get larger as they are less mature. Reactive lymphocytes / plasmacytoid cells / immunocytes are slightly larger than a normal mature lymphocyte and have more cytoplasm, which is often royal blue, save the clearing around the nucleus. They indicate a non-specific antigenic stimulus. Atypical lymphocytes include a feature that is not completely normal such as a pleomorphic (notched, oblong, cleaved, multiple, etc.) nucleus or vacuolated cytoplasm. Their significance is speculative. Lymphoblasts are generally 20μsm or greater in size, have a clumped, poorly staining chromatin and one or more prominent nucleoli. Similar shortfalls come with the identification of most leukemic cells, particularly when poorly differentiated. Large Granular Lymphocytes (LGL) are larger than normal mature lymphocytes, have more cytoplasm, which tends to be pale blue to almost clear and may or may not contain a few pink to purple granules in the cytoplasm. Approximately 10-15% of normally circulating lymphocytes are LGLs.

Another short coming of automated analyzers is cell identification. WBC produced are generally accurate; however, with very low or very high numbers, accuracy can be erroneous. The leukocyte differential can be inaccurate and outright misleading. Depending on the methodology employed, a differential may include three cells or five cells; however, this differential absolutely must be verified with a human eye. Hematology analyzers are notoriously poor at identifying basophils. Basophils are about the same size as a neutrophil, have a narrow, segmented nucleus that is often described as "ribbon" due to its length and winding course through the cell. Dog basophils contain a few to several dark purple granules, whereas cat basophils are generally full of pale lavender-colored, rod shaped granules. Fortunately, marked basophilias are rare.

Another, and potentially profoundly important feature of blood that analyzers generally don't do, or do incorrectly is to evaluate RBC morphologic changes. Polychromasia is the bluish cast immature RBCs will have and is due to retained to ribosomes and polysomes. A few are expected in regenerative anemias; however, their numbers should rise in cases of erythropoiesis (e.g. regenerative anemia). Anisocytosis is the variability in cell size, indicating young and more mature cells present. If an analyzer does not provide a red cell distribution width (RDW), anisocytosis can be used. Heinz bodies are formed secondary to oxidative injury to the erythrocytes, causing hemoglobin to precipitate, aggregate and adhere to the internal surface of RBC membranes. Felines are particularly prone due to extra sulfhydryl groups and flimsy spleen. Allium sp. intoxication, acetaminophen intoxication in dogs, Zn intoxication, repeated propofol administration in cats, hyperthyroidism, diabetes mellitus (especially if ketoacidotic) are all causes of increased Heinz bodies. On Romanowsky stained smears, they appear slightly lighter colored than the remainder of the cell and slightly refractile. New Methylene Blue staining improves visualization. Generally they are on one side and protrude, like a nose. These can be masked by poikilocytosis. Occasionally seen along with the Heinz bodies are eccentrocytes, who have darker cytoplasm that is moved over to one side of the cell, leaving a thin membrane on the opposite side. They too are due to oxidative damage

There are too many morphologic changes to include in this session; however, a few key ones will be noted. Acanthocytes have irregular, variably sized, often wild projections of cytoplasm and can be associated with liver failure, RBC fragmentation such as seen with hemangiosarcoma or disseminated intravascular coagulation (DIC). Spherocytes are formed when a portion of the membrane is removed by macrophages of the MPS. They are most commonly, and significantly, associated with immune-mediated hemolytic anemia. They can only reliably and consistently be detected in canines due to their consistently present central pallor. Spherocytes are smaller, darker and lack central pallor. Although they appear smaller on microscopic examination, their volume remains unchanged; therefore, MCV is not altered. Hypochromasia is noted when RBCs that have an exaggerated central pallor with a thin rim of pale staining cytoplasm. This change is often seen in iron deficiency anemia, as with chronic blood loss. It is important to consider that the MCV and MCHC both can and will change with iron deficiency; however, they represent the mean of the entire population. Therefore, the RBC indices may not reveal the influence of a significant, but smaller population of cells. Keratocytes indicate fragmentation either due to fragile RBCs or trauma in circulation. Common differential diagnoses include iron deficiency, vasculitis, DIC, hemangiosarcoma, caval syndrome, etc. Keratocytes may present as blister, strap or helmet cells. Blister cells have a single colorless vesicle on the edge of the cell. Strap cells occur when the vesicle (i.e. blister) ruptures, leaving a single strand of cytoplasm that will eventually break off. Helmet cells are also formed when the vesicle ruptures, leaving two strands of cytoplasm. Schistocytes are fragments of erythrocytes that have been sheared apart. They are small and irregularly shaped. It should be remembered that since they are fragments of whole RBCs, their volume should likewise be a fragment of normal RBC volume. They can be an extremely important indicator of DIC and should be assessed along with other indicators such as platelet counts, coagulation testing (PT, aPTT), antithrombin levels and fibrin split products (e.g. FDPs, FSPs, D-dimers). Other differential diagnoses include severe iron-deficiency anemia, myelofibrosis and liver disease.

The final and possibly most important features that automated hematology analyzers will not identify are hemoparasites. As previously mentioned, diagnostic smears should be made as soon as possible once the blood is properly anticoagulated. Parasites that are one the surface of RBCs will fall off and organisms that live within cells may leave.

References

Bain BJ. Diagnosis from the Blood Smear. New England Journal of Medicine. 353(5):498-507, 2005.

Cowell RL, Tyler RD, Meinkoth JH, DeNicola DB. Diagnostic Cytology and Hematology of the Dog and Cat. 3nd Edition. Mosby Elsevier. St. Louis. 2008.

Feldman BF, Zinkl JG, Jain NC. Schalm's Veterinary Hematology. 5th Edition. Lippincott Williams & Wilkens. Philadelphia. 2000.

Harvey JW. Atlas of Veterinary Hematology: Blood and Bone Marrow of Domestic Animals. W.B. Saunders. Philadelphia. 2001.

Jain, NC. Essentials of Veterinary Hematology. Lea & Febiger. Philadelphia. 1993.

Latimer KS, Mahaffey EA, Prasse KW. Duncan & Prasse's Veterinary Laboratory Medicine. 4th Edition. Iowa State Press. Ames. 2003.

Metzger FL and Rebar AH. Three-Minute Peripheral Blood Film Evaluation: Preparing the Film. Veterinary Medicine. 99(12):1020-1024, 2004.

Metzger FL and Rebar AH. Three-Minute Peripheral Blood Film Evaluation: The Erythron and Thrombon. Veterinary Medicine 99(12):1026-1034, 2004.

Metzger FL and Rebar AH. Three-Minute Peripheral Blood Film Evaluation: The Leukon. Veterinary Medicine. 99(12):1036-1040, 2004.

Meyer DJ, Harvey JW. Veterinary Laboratory Medicine. 3rd Edition. W.B. Saunders. Philadelphia. 2004.

Reagan WJ, Sanders TG, DeNicola DB. Veterinary Hematology: Atlas of Common Domestic Species. Iowa State University Press. Ames. 1998.

Rebar AH, Feldman BF, MacWilliams PS. A Guide to Hematology in Dogs and Cats. Teton New Media. Jackson. 2001.

Thrall MA, Baker DC, Campbell TW, et al. Veterinary Hematology and Clinical Chemistry. Lippincott Williams & Wilkins. Philadelphia. 2004.

Willard MD, Tvedten H. Small Animal Clinical Diagnosis by Laboratory Methods. 4th Edition. W.B. Saunders. Philadelphia. 2004.