Leukocyte Production

Granulocytes (i.e., neutrophils, eosinophils, basophils) and monocytes are produced in the bone marrow. Although the bone marrow produces some lymphocytes, most lymphocytes are produced by the peripheral lymphoid tissues (i.e., thymus, lymph nodes, spleen, tonsils, bronchial-associated lymphoid tissue, gut-associated lymphoid tissue). Leukocytes develop in the bone marrow from pluripotent and committed stem cells influenced by interleukins and colony-stimulating factors (Figure 4-2).²⁶

FIGURE 4-2 Differentiation of the bone marrow pluripotent stem cells. Baso, Basophil; BFU, burst-forming unit; CFU, colony-forming unit; DC, dendritic cell; E, erythroid; Eo, eosinophil; GEMM, granulocyte-erythroid-monocyte-megakaryocyte; GM, granulocyte-macrophage; Mega, megakaryocyte; NK, natural killer.

Neutrophils compose the majority of leukocytes in blood, and usually leukocytosis is caused by neutrophilia. Kinetics of neutrophils has been well studied and is best understood. Thus the initial discussion here focuses on neutrophils. Cellular “pools” are used to conceptualize and describe the “location” of neutrophils within the bone marrow and blood and to simplify interpretation of bone marrow and CBC data (Figure 4-3). Bone marrow is divided into two pools. The first, the mitotic pool of myeloblasts, promyelocytes, and myelocytes, provides a steady supply of neutrophils to meet tissue demand for these cells. The second pool is maturation and storage, which consists of metamyelocytes, bands, and segs that lack mitotic ability. Precursor cells undergo progressive maturation and provide a reserve pool of segs to meet sudden increased tissue demands for neutrophils until the mitotic pool increases neutrophil production. It is uncommon in the dog and cat to have leukopenia during inflammatory diseases because this large storage pool is available for rapid release of neutrophils. The maturation and storage pools are combined in Figure 4-3. The most mature stages of neutrophil are preferentially released from the bone marrow into the blood first (segs, bands, metamyelocytes, myelocytes, and finally promyelocytes, in that order). As bone marrow stores of segs are depleted, nonsegs (e.g., bands, metamyelocytes, younger neutrophils) are released into the blood, and a left shift occurs: A left shift is a specific indicator of inflammation, and the severity of the left shift reflects the severity of inflammation.

FIGURE 4-3 Neutrophil compartments in the body. Neutrophils in various areas of the body are grouped into pools for evaluation. The bone marrow cells are divided into the mitotic, maturation, and storage pools (see text). Neutrophils in blood are either in the circulating pool, which is sampled by a complete blood count (CBC), or the marginal pool, which is hidden from sampling via the CBC. Neutrophils move in one direction into the tissues, where they can be evaluated by cytology or histopathology.

(Modified from Boggs DR, Winkelstein A: White cell manual, ed 3, Philadelphia, 1975, FA Davis.)

The maturation and storage pool constitutes 80% of the myeloid cell population, whereas the mitotic pool usually accounts for 20% of the myeloid series. In contrast with neutrophils, promonocytes and monocytes are released into blood at a relatively young age. This lack of monocyte maturation and storage in the bone marrow explains why monocytes are observed infrequently in most bone marrow aspirates, unless severe neutropenia is present.

Leukocyte numbers and morphology within the bone marrow can be evaluated by bone marrow aspiration for cytology and core biopsy for histopathology. Aspirate smears of marrow allow qualitative and quantitative observations regarding cell morphology and maturation. Core biopsy provides the best estimation of bone marrow cellularity and detects stromal reactions (e.g., myelofibrosis, granulomatous osteomyelitis). Chapter 2 discusses the interpretation of a bone marrow examination.

Neutrophil Circulation

When neutrophils are released into the blood, about half hesitatively stick and roll along the endothelial cells (i.e., in the marginal pool) and are not in the central axial flow of blood within vessels from which blood is taken during a venipuncture (i.e., in the circulating pool). Those neutrophils that are in the central axial flow of blood within vessels and are taken into a blood sample (and counted in the WBC count) are said to be in the circulating pool. The marginal neutrophil pool is a “hidden” population associated with the endothelial lining of capillaries, especially the lungs and spleen. The circulating and marginal cell pools make up the total blood neutrophil pool (TBNP). Neutrophils distribute between circulating and marginal cell pools, circulate for a brief period of time (half-life of 7.4 hours), and emigrate from blood vessels into tissues. Shifts between these pools can affect the WBC count, especially in cats. In dogs, circulating and marginal pools are about equal. In cats, the marginal pool is two to three times the size of the circulating pool (Table 4-1). Therefore, if neutrophils are mobilized from the marginal pool to the circulating pool in response to fear, excitement, or strenuous exercise, the neutrophil count can potentially double in dogs and triple in cats. This effect is called physiologic leukocytosis and is seen mainly in young healthy cats.

TABLE 4-1 TOTAL BLOOD NEUTROPHIL POOL, CIRCULATING NEUTROPHIL POOL, AND MARGINAL NEUTROPHIL POOL IN DOGS AND CATS

Neutrophil Emigration into Tissues

Neutrophils normally spend about 10 hours in the vascular system before emigrating from the blood vessels into the tissues. Emigration is a random (i.e., non–age ordered) and unidirectional (these cells do not return to the circulation) event. In health, neutrophils primarily migrate into the respiratory, digestive, and urinary tracts at a low rate in response to bacteria and other stimuli. Neutrophils lyse quickly in the septic environment of the lumen of the bowel. In inflammation, excessive tissue neutrophils may be visible as exudate or pus (see Chapter 16). In diseases such as enteritis, tissue neutrophils may be hidden from cytologic or gross observation; however, increased tissue demand for neutrophils usually is reflected in the leukogram.

Left Shift

A left shift (increased absolute numbers of immature neutrophils) indicates inflammatory disease in the patient (see Figure 4-4). Identification of immature neutrophils requires blood smear evaluation. Immature neutrophils such bands, metamyelocytes, myelocytes, and younger neutrophils are termed nonsegmented neutrophils (nonsegs) in general. Identification of nonsegs is subjective and varies greatly among microscopists. Most experienced microscopists, who examine blood smears daily, clearly recognize increased immaturity and toxic changes in neutrophils during inflammation. However, the number of bands, metamyelocytes, and myelocytes reported will vary among observers. Exact numbers of these cells are often impossible to determine when severe toxic change is present, in which case a subjective description should be used; for example, “The blood had a severe degenerative left shift with severe toxic change in neutrophils, though exact numbers of each cell type could not be determined.” may be the most honest description. Even in nontoxic reactions the division between segs and nonsegs may be difficult, and the term hyposegmentation is then used to indicate that the neutrophils looked more immature than normal though the number of reported nonsegs was not increased.

Because of the effect of the observer’s impression in classification of nonsegs, one should use a definite increase in nonsegs for a certain conclusion of inflammation. A finding of greater than 1000 nonsegs/µl is a reasonable threshold for a left shift. (Note that 1000 nonsegs/µl equals 1.0 nonsegs × 10⁹/L.) Small changes from day to day or slight changes from the reference values should be interpreted with caution because manual differential leukocyte counts are especially imprecise with nonsegs and cells found in low numbers, such as basophils. Mild left shifts (i.e., 300 to 1000 nonsegs/µl) occur in hemorrhagic, chronic, or granulomatous diseases. No left shift may be noted in many patients with inflammatory disease, thus the absence of a left shift or a very mild left shift does not exclude inflammation from the diagnosis.

The absolute number of nonsegs and their state of immaturity indicate the severity of the left shift. Immature neutrophils (nonsegs) observed in blood include bands (stabs), metamyelocytes (juveniles), myelocytes, and promyelocytes. Bands usually constitute most of the left shift because the more mature neutrophils are released from bone marrow first. Neutrophils younger than bands indicate an increasingly severe left shift associated with increasingly intense inflammation. If several myelocytes and metamyelocytes are found, the number of metamyelocytes, myelocytes, and promyelocytes should be reported individually (not simply grouped as nonsegs) to indicate severity of the left shift. Detectable numbers of blast cells (i.e., myeloblasts) or irregular maturation patterns may suggest granulocytic leukemia. More nonsegs than segs indicates a degenerative left shift and a poor prognosis.

Leukogram Changes in Inflammation

Figure 4-5 shows a likely pattern of leukocyte changes through a typical inflammatory response, with a distinct onset of inflammation and then progressing uniformly to resolution. Changes in individual patients may vary from this typical pattern due to treatment, rupture of an abscess, or the like. The greatest left shift is expected in early stages of the disease process, because as the preexisting bone marrow storage pool is depleted of segs, then more bands and metamyelocytes are released to meet early intense demand. With time, myeloid hyperplasia within the bone marrow expands neutrophil production. When neutrophil production and maturation time are sufficient, mainly mature segs are released into the blood and the severity of the left shift should diminish or disappear. If tissue inflammation stabilizes at a persistent low to moderate level, the bone marrow should reach a production rate sufficient for most neutrophils to mature before release. Thus chronic inflammation may be characterized by little to no left shift and minimal to no leukocytosis. Therefore, chronic or mild inflammation may be difficult to document by leukogram data alone. Acute phase proteins, such as C-reactive protein or serum amyloid A, are sensitive indicators of inflammation and may be used to complement the leukogram. Increased rouleaux (see Chapter 2) in canine blood smears or fever in the patient suggests inflammation. Increased rouleaux formation usually is associated with production of acute phase proteins such as fibrinogen. A normal leukogram appearance does not exclude inflammatory diseases, especially if inflammation is mild or chronic, or only involves a surface (e.g., cystitis).

FIGURE 4-5 Expected leukocyte changes with resolving inflammation. The greatest leukocytosis and left shift are expected early in acute inflammation. This period is also accompanied by the lymphopenia of stress. During later phases of inflammation, a more mature form of neutrophilia is expected, because bone marrow hyperplasia and marrow production of neutrophils should be adequate to allow maturation of neutrophils before release into blood. Tissue demand for neutrophils also tends to decrease during recovery.

Bone Marrow Response During Inflammation

Myeloid hyperplasia of the bone marrow is expected with inflammation of greater than 2 to 3 days’ duration. Bone marrow sampling is seldom indicated in patients with inflammatory diseases because a leukocytosis and left shift are usually present, and these indicate the marrow is active. Persistent, moderate to severe leukopenia in a dog or cat with inflammatory disease is unexpected and thus may be an indication to examine the bone marrow for the cause (see Chapter 2).

Prognosis

Obviously factors other than neutrophil counts affect prognosis, such as the cause of the disease and site of inflammation. However, the neutrophil counts reflect the current balance of effective bone marrow production and tissue demand. If the bone marrow is responding typically to an inflammatory process with a mild to moderate neutrophilia and mild to moderate left shift, the prognosis is relatively good. Leukocytosis in dogs is usually less than 40,000 WBCs/µl. In 182 canine CBC exams with leukocytosis, 151 (83%) had 17,500 to 39,990 WBCs/µl. Leukocytosis in this range is thus mild to moderate and suggests a favorable prognosis. Only 5% had marked to extreme leukocytosis of 61,050 to 127,500 cells/µl. Leukemoid reactions of over 50,000 to 60,000 WBCs indicate a poor prognosis because, even in the presence of excessive numbers of leukocytes (usually neutrophils), the cause of the inflammatory response is not corrected. The magnitude of feline leukocytosis is usually less than that in dogs (i.e., 70% of cases are < 30,000 WBCs/µl). The severity of eosinophilia in eosinophilic inflammation is less than neutrophilia during neutrophil inflammation. Therefore, a leukemoid reaction of eosinophils (due to strong eosinophil inflammation, not leukemia or hypereosinophilic syndrome) is diagnosed when there are more than 25,000 to 30,000 eosinophils/µl.

Leukocyte count criteria, suggesting a poor prognosis, are summarized in Table 4-2 (see also later discussion of Toxic Neutrophils). A degenerative left shift, leukopenia, neutropenia, lymphopenia, leukemoid reaction, or a combination thereof is an atypical, unexpected response to inflammatory disease indicating severe disease, toxemia, severe stress, inadequate bone marrow production, problems interfering with an effective response, or a combination of these. A degenerative left shift has more nonsegs than segs, regardless of total leukocyte count. Finding more immature than mature neutrophils indicates that the bone marrow cannot produce neutrophils at a rate sufficient for them to mature completely prior to release. Either tissue demand for neutrophils has escalated dramatically, cell production is decreased, or both. Severity of change and trend over daily CBCs are important in assessing prognosis. A degenerative left shift, severe neutropenia and leukopenia, lymphopenia, and marked toxic change in most neutrophils often suggest gram-negative sepsis, such as a ruptured intestine or parvovirus enteritis.

TABLE 4-2 LEUKOGRAM FINDINGS INDICATING A POOR PROGNOSIS

Both leukopenia and neutropenia are unfavorable prognostic signs. These findings suggest that bone marrow is incapable of producing sufficient numbers of neutrophils, that tissue consumption of neutrophils is overwhelming, or both. Neutropenia, whether primary (bone marrow disease) or secondary (excessive tissue consumption), severely predisposes the patient to infection and septicemia. Leukopenia is usually caused by neutropenia, but lymphopenia may also cause leukopenia despite normal neutrophil numbers. Lymphopenia usually indicates stress. Severe or persistent lymphopenia indicates severe or persistent stress. In severe leukopenia (e.g., <1000 WBCs/µl), a left shift is not concluded even if all neutrophils are nonsegs, because an increase in the absolute number of immature neutrophils greater than 1000 WBCs/µl is not possible. (Trying to perform a manual differential WBC count is also not necessary because the total WBCs already indicates a neutropenia and lymphopenia, and imprecision of a 25 to 50 ManDiff count with abnormally appearing cells is very great. Automated instrument differential counts are more accurate in severe leukopenia.)

A leukemoid reaction is a marked leukocytosis (>50,000 to 100,000 WBCs/µl) due to inflammatory disease and not leukemia. Leukemoid means leukemia-like because of the magnitude of leukocytosis. A leukemoid reaction indicates a poor prognosis because, despite abundant (and actually excessive numbers of) neutrophils, the inflammatory disease is not being corrected. Causes of leukemoid reactions are often severe localized infections (e.g., pyometra, abscess). Additional causes are IMHA, paraneoplastic syndromes with bone marrow stimulation (e.g., metastatic fibrosarcoma, renal carcinoma, rectal adenoma), rare parasitism (e.g., Hepatozoon canis infection), and neutrophil functional defects (canine leukocyte adhesion protein deficiency [CLAD] of Irish setters).³⁹ With pyometra and some walled-off abscesses, there is an anatomic problem preventing healing. Pus and the infectious agent or other cause cannot drain out of the body, and antibiotics may not penetrate the lesion. With CLAD, dysfunctional neutrophils are incapable of correcting common infections even in high numbers. The leukemoid reaction in IMHA seems an exception, but acute massive destruction of erythrocytes and phagocytosis of debris by macrophages is a strong stimulus for an inflammatory reaction.

Differentiation of a leukemoid reaction from chronic granulocytic leukemia (CGL) is difficult. CGL is rare, so the odds favor diagnosis of a leukemoid reaction when massive neutrophilia exists. Both leukemoid reactions and CGL lack the blast cells or atypia seen in acute granulocytic leukemia (see Myeloid Neoplasms later in this chapter). Left shifts in both leukemoid reactions and CGL often involve mainly bands, and both can have a few more immature forms. Cytologic examination of lymphadenopathy is often a key in diagnosis of CGL. Lymph node aspirates (and even liver or spleen aspirates) with CGL look like bone marrow in having mixed hematopoiesis with a few to moderate number of immature myeloid cells, including myeloblasts. The number of blast cells is not obviously increased, so bone marrow aspirates in CGL are often not diagnostic. The leukocytosis of a leukemoid reaction should resolve after appropriate treatment (e.g., removal of pyometra uterus). Clinical signs of illness, toxic neutrophils, and evidence of diseases that cause leukemoid reactions (e.g., pyometra) suggest diagnosis of a leukemoid reaction. Dogs with CGL have persistent leukocytosis and often do not look sick.

Toxic Neutrophils

Toxic neutrophils were associated with increased fatality, length of hospitalization, and treatment costs in dogs.³ The prevalence of pyometra, parvovirus infection, acute renal failure, peritonitis, IMHA, disseminated intravascular coagulation (DIC), pancreatitis, septicemia, and neoplastic disorders was significantly higher among dogs with toxic neutrophils. Bacterial infections cause severe toxic changes in neutrophils, such as in secondary bacterial enteritis in parvovirus enteritis. However, toxic neutrophils occur in disease without infection (e.g., IMHA, pancreatitis, chemotherapeutic agents, renal failure). The presence of toxic neutrophils in cats was associated with a significantly higher prevalence of shock, sepsis, panleukopenia, peritonitis, pneumonia, and upper respiratory tract diseases, as were infectious (viral and bacterial) and metabolic disorders.³⁶ Negative findings in cats with toxic neutrophils were milder than with dogs, suggesting that cats form toxic neutrophils with milder diseases than dogs.

Classification of toxemia may be quite detailed³ or more simplified (see Chapter 2). The number of neutrophils that appeared toxic (percentage; or few, moderate, many) and severity of morphologic change should be reported. A few (1+) toxic neutrophils are of minimal importance. However, moderate to many (2+ to 4+) toxic neutrophils should not be ignored (see Figure 2-9). Toxic change in neutrophils, though subjective, is sometimes the only indicator of disease because numerical results of the leukogram appear normal in many cases. A blood smear evaluation by a competent observer should always be included during initial evaluation of a sick patient.

Parvovirus Infection

Parvovirus enteritis may be associated with leukopenia, neutropenia with severe toxic changes, and lymphopenia. Parvovirus infects and destroys rapidly dividing cells (e.g., intestinal crypt epithelium, lymphoid tissue, hematopoietic cells). Massive neutrophil exudation through the damaged mucosa contributes to leukopenia, as does hematopoietic cell damage in the bone marrow. Leukopenia is transient and usually early in canine parvovirus disease, so it may be missed without multiple CBCs. Mast cells from the inflamed gut may be detected in blood smears. The mast cells come from damaged intestinal mucosa and do not indicate mast cell neoplasia. Neutrophilic leukocytosis is expected with recovery. During periods of intense granulopoiesis, a leukemoid reaction or other disturbances of normal cell maturation may occur. With complete clinical recovery, baseline leukogram values are regained. Diagnosis of parvovirus enteritis is discussed in Chapter 9. Panleukopenia in cats has a similar hematologic response.