Leukogram patterns

Changes in total and differential leukocyte count are usually grouped into patterns, which facilitate interpretation. These patterns are:

• Stress leukogram
• Physiologic leukocytosis
• Inflammatory leukogram, including a table on how we use terminology to describe changes in mature (segmented) and immature (bands, metamyelocytes etc) neutrophils and how we use this terminology to understand what is going on to result in those changes (bone marrow response in relation to tissue demand or inflammation).
• Leukemia: More on leukemia is given in a separate section.

We have also provided useful tips on identifying the type of leukogram that is occurring (see below).

Stress leukogram

This is due to increased endogenous (or exogenous administered) corticosteroids. The classic leukogram pattern from increased corticosteroids (either exogenous or endogenous) includes a neutrophilia, lymphopenia, monocytosis, and eosinopenia. Not all of these changes will be present in any given animal. The most consistent finding is lymphopenia, followed by a mature neutrophilia (increase in segmented neutrophils, but not usually bands).  Monocytosis from corticosteroids is fairly common in dogs and is uncommon in cats, horses and cattle.

• Experimental studies in cats: Treatment with immunosuppressive doses of corticosteroids (dexamethasone at 0.55 mg/kg/day or prednisone at 4.4 mg/kg/day for 56 days given to two groups of 7 cats) increased mean neutrophil and monocyte counts and decreased lymphocyte and eosinophil counts in 14 cats. However, neutrophilia and monocytosis were uncommon with lymphopenia being the most common finding (Lowe et al 2008).  In another study of 10 allergic cats given 1-2 mg/kg/d prednisolone for 2 weeks, the mean neutrophil and monocyte counts peaked at 7 days, 1 hour after dosing, however it was not stated if the cats developed a neutrophilia or a monocytosis and lymphocyte and eosinophil counts were not provided (Khelik et al 2019). 
• Experimental studies in horses: In one study, 8 ponies were given 2 mg dexamethasone IV (likely not a good mimic of a stress response) and leukocyte kinetics were monitored over several days (Targowski 1975). The neutrophil count increased to a mean of 11,000/μL within 2 hours, to reach a peak at a mean of 16,000/μL at 7 hours. Counts were still mildly increased at 48 hours (mean of 7,000/μL) and had returned to baseline by 72 hours. Lymphocyte counts began to decrease at by 4 hours with the nadir (1.5-2,000/μL) at 7-24 hours and normalization by 48 hours. Eosinophil counts decreased by 4 hours (mean of 200/μL) with the nadir (0-100/μL) between 7 and 48 hours. The eosinophil count rebounded by 72 hours. A mild monocytosis (mean 800/μL) was seen only at 48 hours in this group of ponies. In another study, 2 horses given 200 mg prednisolone IV (roughly 0.4 mg/kg, assuming a 500 kg horse) doubled their neutrophil counts (up to 14,500 and 8,200/uL) within 3 hours of injection (Carakostas et al 1981), compatible with a shift from the marginating to the circulating pool (estimated at 1:1 in horses [Carakostas et al 1981]). In 8 thoroughbred horses given 1.25 mg ACTH IV (likely the best experimental mimic of stress), the mean neutrophil counts peaked at 8,000/uL at 4 hours (baseline of 4,800/uL) then declined to near baseline by 24 hours. The lymphocyte count decreased from a mean of 2,800/uL to a nadir of 2,100/uL at 4 hours and was at baseline by 6 hours (Rossdale et al 1982).
• Experimental studies in ruminants: In 3 calves given 0.2 mg/kg prednisolone, the neutrophil count increased by 4 hours to peak at 9-11,000/uL at 10 hours (the latest measured time point) after administration IV (Carlson and Kaneko 1976). In pregnant ewes given dexamethasone (0.2 mg/kg), total leukocyte and neutrophil counts increased with a concurrent decrease in lymphocyte and eosinophil counts, with no change in monocyte counts. Counts returned to normal within 48 hours (Edelstone et al 1978). 

The mechanisms for corticosteroid-related leukogram changes are as follows:

• Neutrophilia is mainly due to a shift from MGP to CGP. A small portion of the neutrophilia may also be due to increased release from bone marrow and delayed apoptosis. Generally, the neutrophil count is usually not increased more than 1x above the reference limit (the MGP to CGP ratio is 1:1), but can be as high as 2-3x the upper reference limit in cats (since they have a higher MGP:CGP ratio than other species). In general, it would be uncommon for a stress leukogram to increase a segmented neutrophil count >10-12,000/μL in horses and ruminants. Note, that animals on chronic immunosuppressive (high) doses of steroids (not just a release of endogenous steroids due to “stress”), a mild left shift (band neutrophil count of 300 to 400/uL) may be seen without toxic change or an increase in mature segmented neutrophils (segmented neutrophil count is usually normal [personal observations]).
• Lymphopenia is due to multiple mechanisms, including decreased efflux from lymph nodes, decreased proliferative and activation cytokines (e.g. IL-2) for lymphocytes and lymphotoxic effects (induction of apoptosis, only more chronic conditions and likely with higher doses of steroids).
• The mechanism for the monocytosis is unknown, but it could be due to a shift from marginating to circulating pool, if these pools exist for this leukocyte.
• Eosinopenia is thought to be due to decreased release from the bone marrow, although the precise mechanism is unknown.

Care must be taken when using the term “stress response” to refer to a leukogram. A stress leukogram is specifically referring to the results of corticosteroids. It is tempting to use the term stress leukogram when describing a nervous animal, but these animals are likely to have an epinephrine response rather than a corticosteroid response. So only use the term “stress leukogram” when referring to changes that you are attributing to corticosteroids.

Causes of a stress leukogram are:

• Endogenous stress: Chronic or acute increases in corticosteroids. This is expected in ill animals.
• Hyperadrenocorticism: Excess endogenous corticosteroids.
• Exogenous corticosteroid administration: In dogs, this may be accompanied by increased liver enzymes (especially ALP).

Other changes in clinical pathologic results that may be seen with a stress leukogram are:

• Hyperglycemia: Due to corticosteroids stimulating gluconeogenesis (requires more chronic increase in steroids).
• Increase in ALP activity: Only in dogs and with chronic endogenous corticosteroids or exogenous corticosteroids.

Physiologic leukocytosis

This is due to epinephrine or norepinephrine and is called a “flight or fight” response. Changes on the leukogram due to epinephrine responses are most commonly seen in cats (of any age), young horses (1 year of age or less), and possibly younger animals of other breeds. The typical pattern is mild neutrophilia (mature cells with no left shift) and lymphocytosis (sometimes called a “physiologic lymphocytosis”). These changes are usually transient, and will diminish within about 30 minutes after the animal calms (Rose et al 1983). 

• The neutrophilia is due to a shift from the MGP to the CGP. A left shift will not be seen.
• The lymphocytosis is thought to be due to release from the spleen. It is uncommon for the lymphocyte count to exceed 10 thousand/μL, but we have seen rare cases in cats of an epinephrine-related lymphocytosis of about 30 thousand/μL and colleagues have reported higher values (very very rare). In 16 yearling unbroken quarterhorses, lymphocyte counts ranged from 5,600 to 14,000/uL, with higher counts seen in the horses caught last (Schalm’s Veterinary Hematology, 4th edition). A few yearlings had a mild neutrophilia (up to 14,000/uL) and monocytosis (up to 1,100/uL) as well. A mild increase in lymphocyte percentages occurred within 5 to 30 minutes of epinephrine injection (2 mg) in goats (from 60±4% to a peak of 70±4% at 5 minutes), which was lacking in splenectomized animals, however there was minimal change in the total leukocyte count (Abdelatif and Abdalla 2012).
• An eosinophilia and basophilia may accompany the above changes in cats.

Other changes in clinical pathologic results that may be seen with a physiologic leukocytosis are:

• Hyperglycemia: Due to epinephrine stimulating glycogenolysis (rapid response).

Inflammatory leukogram

The blood neutrophil count with inflammation represents the balance between tissue demand and bone marrow supply. The leukogram pattern can be variable depending on the source and severity of the inflammation. The changes bulleted below are often identified on the CBC from an animal with inflammation, but it is possible to have inflammation in an animal that does not show an inflammatory leukogram. This occurs particularly with localized inflammation that is not inciting a systemic inflammatory response (with associated clinical and clinical pathologic manifestations, such as fever, an inflammatory leukogram, hypoferremia and increased globulins). If there is other clinical or laboratory evidence of inflammation in an animal, do not let the lack of an inflammatory leukogram dissuade you from considering that an inflammatory process may be occurring. Also remember, that a single hemogram is but a “snapshot” in time. Since changes in the hemogram can occur frequently, monitoring hemogram results frequently (every 12 to 24 hours) can be very helpful in determining the course of an inflammatory response and response to treatment. Regardless, never treat laboratory data, always treat the patient.

With the classic inflammatory leukogram, the bone marrow senses an increased demand for neutrophils (through cytokine stimulation). The bone marrow initially responds to this demand by releasing the storage pool of post-mitotic mature and band neutrophils, which results in a neutrophilia with a left shift (with fewer band than segmented neutrophils, i.e. not degenerative), in most species, other than cattle (which release earlier forms). Cytokines (such as granulocyte and granulocyte-monocyte stimulating factor) also kick in and stimulate granulopoiesis within 3-5 days. Thus, the following changes are seen with a classic inflammatory leukogram:

• Neutrophilia (anything from very mild to very severe (rarely over 80 thousand/µL – mostly with sequestered inflammation that has induced a marked granulocytic hyperplasia), depending on the severity of the inflammation and effective response by the marrow. In general, dogs tend to have the highest degree of neutrophilia than other animals.
• Left shift in neutrophils: The more severe the left shift (with more immature neutrophils, such as metamyelocytes, myelocytes and rarely progranulocytes, not only bands, being released), the more severe the inflammatory stimulus.
• Toxic change: This usually, but does not always, accompany an inflammatory leukogram. Mild, chronic or resolving inflammation may not result in toxic change (which indicates immaturity and accelerated release versus a true toxic effect on these cells). Cytokine stimulation without infection, e.g. cytokine secretion stimulated by cancer (paraneoplastic response) and immune-mediated hemolytic anemia in dogs can have a moderate to marked neutrophilia with a  left shift (sometimes to metamyelocytes and myelocytes), with only mild or sometimes no evidence of toxic change. Usually, a moderate to severe inflammatory response, particularly that due to bacterial sepsis (but not limited to bacterial sepsis) will result in a left shift in neutrophils and toxic change.
• Monocytosis: This may or may not be part of an inflammatory leukogram. A monocytosis is usually seen with inflammation when it is more long-standing or resolving. However, in some animals, the increase in monocytes is quite mild.
• Concurrent lymphopenia (with or without eosinopenia): A lymphopenia (and eosinopenia) frequently accompany an inflammatory leukogram, but this does not always happen. We usually attribute this due to concurrent endogenous stress (corticosteroid release), although inflammatory cytokines may directly cause a lymphopenia (through shifting hematopoiesis to granulocytic lineages and downregulating lymphopoiesis). In some cases, a mild lymphocytosis may be seen with chronic inflammatory conditions that incite a decent antigenic response, e.g. pyometra, some protozoal infections (e.g. theileriosis, trypanosomiasis, young horses recovering from Potomac horse fever or Anaplasma phagocytophilum).

With overwhelming inflammation, the bone marrow is unable to release enough leukocytes to keep up with peripheral demand, leading to leukopenia. The typical leukogram pattern is:

• Neutropenia with a left shift, which may include more immature forms than bands (metamyelocytes, myelocytes). Some animals may have more immature than mature/segmented neutrophils (i.e. may see degenerative left shift). A severe neutropenia and degenerative left shift (often to myelocytes) is common in cattle with acute inflammation (e.g. acute mastitis or metritis). This is because cattle do not have large reserves of neutrophils in the storage marrow pool to release in respond to an acute inflammatory stimulus. In other species, a neutrophilia with a left shift is more commonly seen with acute inflammation and a neutropenia with a left shift only ensues when the inflammation is very severe and overwhelming.
• Toxic change: Moderate to severe toxic change in neutrophils should be seen in this situation.
• Concurrent lymphopenia (with or without eosinopenia): See above.

Sequestered inflammation, such as a closed pyometra or an abscess, can result in a moderate to marked neutrophilia (with or without a left shift) and usually with little or mild toxic change (mimicking a paraneoplastic leukocytosis). There may be a concurrent monocytosis. Rupture of the lesion will result in a more severe inflammatory leukogram and corresponding clinical signs (i.e. can be life-threatening).

Acute endotoxemia results in a different hematologic pattern, which may vary between species.

• Dogs: Injection of endotoxin (1 -200 ug/kg) to 5-6 dogs resulted in an acute leukopenia (1-3 hours, nadir 2 hours), followed by a leukocytosis due to a neutrophilia by 24-48 hours. A left shift was evident at 8-24 hours. Mean leukocyte counts were still increased above baseline (but within reference intervals) by 5-7 days (Hoshiya et al 2001, Higgins et al 2003). Mean lymphocyte counts within 2-8 hours and were within reference intervals (but still lower than baseline) at 24 hours. Mean monocyte and eosinophil counts were decreased between 2 and 8 hours, with recovery of the monocyte count by 24 hours, whereas eosinophils were still suppressed (Higgins et al 2003). The mean platelet count decreased, with the nadir (mean, 195,000/uL with SD of 46,000/uL) at 24 hours. Albumin decreased by 24 hours (to a mean of 2.7 ± 0.1 mg/dL), whereas an increase in globulins and acute phase proteins (serum amyloid A and C-reactive protein) was seen at 8 and 2-4 hours, respectively (earlier increase in serum amyloid A versus C-reactive protein) (Higgins et al 2003). 
• Horses: Consistent changes appear to be a rapid neutropenia (1-3 hours), followed by a neutrophilia (6-24 hours), a rapid lymphopenia, which may persist for 24 hours, a rapid transient thrombocytopenia (<2 hours) and a rapid hyperglycemia (1-3 hours) after infusion of endotoxins in adult horses (Burrows and Cannon 1970, Burrows 1979, Lavoie et al 1990, Baskett et al 1997). Infusion of endotoxin into neonatal foals results in a rapid, transient neutropenia (within 1-4 hours, potentially without a left shift or toxic change, likely due to rapid margination of neutrophils) and lymphopenia (mechanism unclear) within a few hours, followed by a neutrophilia by 6-24 hours (likely due to release of bone marrow stores). The lymphopenia is too acute to be attributed to a stress response, being seen within 1-3 hours after endotoxin infusion (Lavoie et al 1990). Some foals had a mild transient thrombocytopenia (Lavoie et al 1990), which has been attributed to splenic or lung sequestration. A similar rapid development of neutropenia (<1,000/uL), by 60 minutes and persisting for 4 hours, was seen in 6 Shetland ponies given various doses of E coli endotoxin IV. In the latter studies, the ponies also developed a hyperglycemia (average 200 mg/dL) within 30 minutes of endotoxin administration, which persisted for an hour. The PCV increased from an average of 30% to 45% for up to 4 hours after endotoxin was given (attributed to splenic contraction, but could be secondary to fluid shifts into the extravascular space) (Burrows and Cannon 1970). Similar changes were seen in ponies given IV (150 ug/kg), a single high dose (300 ug/kg) of LPS intraperitoneally (IP) or multiple doses (100 ug/kg) IP at 3 hour intervals. The neutrophil counts decreased to a nadir between 1-4 hours after endotoxin and stayed low (less than 3,000/uL) for up to 10 hours in IP-dosed horses. This was followed by a neutrophilia at 24-48 hours (ranging from an average of 7,000/uL after a single IP dose to an average of 21,000/uL after a single IV dose). In contrast, the average neutrophil count remained around 3-4,000/ul after multiple IP doses for this time frame. The ponies also had an increase in PCV, which persisted for up to 10 hours after IP dosing but had returned to baseline by 8 hours after IV administration. The lymphocyte count decreased within 1 hour to a nadir (average 1,500-2,200/uL) between 6-24 hours after endotoxin (all doses and routes). The platelet count also decreased rapidly (resulting in a mild thrombocytopenia, with an average count of 80,000/uL) by 1 hour after administration, followed by a mild increase (but not to baseline levels). Mean glucose concentration spiked between 1-3 hours, whereas an obvious increase in mean plasma fibrinogen concentration was only seen at 48 hours after multiple IP doses versus a single IP dose (Burrows 1979).  
• Cattle: Internal studies at Cornell in dairy cattle in early lactation show that neutrophils decrease rapidly within 1 hour of LPS injection (250 ng/kg) and rebound by 24 hours. A left shift was evident at 4 and 24 hours. This data mimics that seen in other species, indicating consistency in the response to endotoxin injection in experimental studies.

Causes of inflammation are many and include:

• Infectious agents: Bacteria, viruses, protozoa, fungi. Some infectious agents, e.g. Theileria in cattle, can cause a concurrent lymphocytosis (which may reflect chronic antigenic stimulation).
• Immune-mediated conditions: Vasculitis, polyarthritis.
• Cancer: This could be due to cytokine production by the tumor cells (paraneoplastic response), a host response to the tumor, or tumor necrosis.
• Necrosis: Hypoxic injury from anemia, infarct, tissue burns.
• Foreign body: e.g. migrating grass awn.

Other changes in clinical pathologic results that may be seen with an inflammatory leukogram are:

• Thrombocytosis: Due to inflammatory cytokines, such as interleukin-1 and 6. As noted above, acute endotoxemia can result in a transient thrombocytopenia.
• Changes in albumin and globulins: This is usually comprised of low albumin or high globulin concentrations or a combination of both. The low albumin concentration is due to a negative acute phase response (downregulation of production in hepatocytes) and the high globulin concentrations will be due to an acute phase response (increased production of α2 globulins by hepatocytes) or antigenic stimulation (polyclonal increase in immunoglobulins) or a combination of both.
• Changes in the iron panel: This could be hypoferremia with low % saturation or hypoferremia with low TIBC and normal % saturation.
• Changes in glucose: If there is concurrent stress, hyperglycemia may be seen. With sepsis, hypoglycemia may be evident.

Note that eosinopenia can be seen in experimental models of acute bacterial infection in rodents and rabbits, which decreased concentrations not being attributable to concurrent cortisol release (the decrease in eosinophils can precede the increase in cortisol concentrations, occurring quite rapidly within an hour, with a persistent decrease for up to 24 hours) (Bass 1975, Bass 1977, Bass et al 1980). Eosinopenia can also be a finding in severely ill patients with bacterial and viral infections (Davido et al 2017, Le Borgne et al 2021).

Left shift terms

Terminology that you may encounter associated with inflammatory leukograms can be confusing, including degenerative and regenerative left shift. The key concept is: Is the bone marrow responding to the tissue draw or inflammation (release of cells or hyperplasia), is the response sufficient to combat the inflammation (hyperplasia in longer-standing inflammation) or is the inflammation winning (with or without hyperplasia)? We have provided a table below that hopefully gives a useful guideline on how we interpret changes in mature and immature neutrophil numbers (remember, there are exceptions to the rules). Note also that immature applies to all immature neutrophil stages that can be seen in blood (bands > metamyelocytes > myelocytes > progranulocytes, i.e. a proportionate left shift where more mature stages outnumber the more immature stages).

Applied terminology for changes in neutrophil numbers with inflammation

Segmented (mature) neutrophil count	Relative proportions	Terminology	Hyperplastic marrow	Comments
Low	Immature > Mature	Degenerative left shift	NO	Acute (bovine in particular, also horses, less common in other species with acute inflammation due to release of neutrophils from bone marrow stores)
Low	Mature > Immature	Left shift*	Not likely	Not typical of inflammation
Normal	Immature > Mature	Degenerative left shift	Maybe	Uncommon, tissue draw (inflammation) likely or rebound from marrow injury – Monitor (who is winning?)
Normal	Mature > Immature	Left shift**	Likely	Substantial tissue draw remains – Monitor
High	Immature > Mature	Left shift***	Yes	Substantial tissue draw remains – Monitor
High	Mature > Immature	Regenerative left shift	YES	Marrow is meeting or exceeding demands
* This is not a typical pattern seen with inflammation, but is considered by some clinical pathologists to be a degenerative left shift. A production defect should be considered (which is why some avoid calling this pattern degenerative; as the term implies an inflammatory draw). If there is toxic change, a concurrent inflammatory stimulus is probably occurring, which could be a consequence of the low mature neutrophil count. ** In this scenario, the strict criteria of the immature forms outnumbering the mature forms is not met (hence the term degenerative may not be applied by some). However, this is akin to an anemic animal who has some reticulocytes, but not enough. At this point in time, we do not know if the marrow is undergoing hyperplasia or just releasing immature cells to meet demand and not keeping up. Alternatively, the marrow may not be producing enough cells to meet tissue demand for other reasons. *** Some clinical pathologists call this a degenerative left shift because immature forms are more numerous than mature or segmented forms. Others prefer not to call this a degenerative left shift, because the marrow is responding and hyperplastic. The term “degenerative left shift” has negative clinical implications, and in this scenario the body is responding better to the inflammation than an animal with a normal or low segmented neutrophil count. Realize in these cases, you have evidence that the marrow is hyperplastic, but you don’t know yet if the response is going to be adequate or if the tissue draw (inflammation) is stronger than the marrow’s response. Only serial monitoring will tell which way the animal is heading.
Table taken from teaching notes prepared by Dr. Erica Behling-Kelly for Cornell students.

 

Stress or inflammation?