The collection of blood and urine samples is covered in Section 1.10.3. Parameters such as glucose, creatine kinase and aspartate aminotransferase (AST) can be altered by stress associated with handling and restraint, or tissue damage that has occurred during sample collection. For example, potassium results appear less reliable in samples taken with a plastic cannula as opposed to a hypodermic needle (Robson et al., 1981).

Rabbit blood haemolyses easily and clots quickly (Perry-Clark and Meunier, 1991). Small clots affect haematology results and haemolysis affects certain biochemistry results, especially potassium and serum inorganic phosphorus that are released from erythrocytes. Rapid clotting can affect the performance of some analysers and heparinized syringes and needles are required to avoid this.

2.1.2 Fasting and other physiological considerations

It is not possible to take a guaranteed fasting sample from a rabbit because they ingest caecotrophs. Parameters such as blood glucose are affected by digestion. Some parameters such as bile acids, cholesterol and urea follow a diurnal rhythm that also affects the total and differential white cell count (Fekete, 1989; Fox and Laird, 1970; Loeb and Quimby, 1989). Stress associated with car journeys or a period in unfamiliar surroundings will increase blood glucose and alter haematological parameters such as the distribution of neutrophils and lymphocytes. Pregnancy affects parameters such as protein, haematocrit, cholesterol, alkaline phosphatase, triglycerides (Viard-Drouet et al., 1984), glucose, sodium, calcium, phosphate and red cell indices (Palm, 1997). Serum cholesterol is the parameter most affected and can be up to 30% lower in pregnant than in non-pregnant animals (Palm, 1997).

Anaesthesia affects some blood parameters such as potassium (Robson et al., 1981). The effect of anaesthesia on biochemical parameters is minimized by taking samples within 5 minutes of induction. Intravenous or intraosseous fluids will also affect haematological findings (Ros et al., 1991) and blood samples should be taken prior to treatment.

2.1.3 Reference ranges

There are a number of published reference ranges for haematological and biochemical parameters in rabbits. Conversion factors for the variety of units used in reference ranges are given in Table 2.1. Differences in analytical techniques between laboratories can lead to disparity in results. Historically laboratory data were often derived from populations of rabbits of the same breed, sex and age that are not genetically diverse. In contrast, the pet rabbit population is made up of a variety of breeds and cross-breeds, and is composed of rabbits of all ages. With the increased interest of rabbit owners in pursuing diagnostics for their pets, many commercial labs have built up a vast database of samples from a genetically heterogeneous population. Significant breed and sex differences have been noted for some parameters in laboratory rabbits (Kozma et al., 1974).

Published reference ranges for pet rabbits are derived either from sets of data from laboratory rabbits or from data collected by the author. Some reference ranges are an amalgamation of other reference ranges often without addressing the fact that the ranges have been determined using different methodologies. This results in ranges so wide that almost any result will fall within the normal range. For example, the reference range for serum albumin concentrations of pet rabbits is given as 27–46 g/L (Malley, 1996) based on four published sets of data. Gillett (1994) gives an even wider range of 27–50 g/L for laboratory rabbits based on five sets of data. For some parameters there are big differences between published reference ranges. An example is blood calcium. Gillett (1994) gives a range of 1.5–3.4 mmol/L, compared to 3.4–4.0 mmol/L given by Harkness and Wagner (1995). Differences in calcium content of the diet between the groups of rabbits or differences in analytical technique could explain these discrepancies. Different laboratory methods can result in large differences between reference ranges. An example is alkaline phosphatase. Collins (1988) gives a reference range of 4.1–6.2 IU/L, compared to 10–70 IU/L (Gillett, 1994) or 112–350 IU/L (Harkness and Wagner, 1995).

Automated flow cytometers are designed to measure numbers of human blood cells. Rabbit erythrocytes are smaller in diameter than human erythrocytes and also vary in diameter. These differences cause problems with some automated analysers. Automated differential white cells counts cannot be relied upon for rabbit blood and accurate results can only be obtained using manual counting methods (Kabata et al., 1991).

2.2.1 Morphological characteristics of blood cells

Some morphological characteristics of rabbit blood cells are different from other species. The red blood cells vary in diameter within a range of 5.0–7.8 μm (Sanderson and Phillips, 1981). This variation in diameter (anisocytosis) is a feature of blood smears from rabbits and is not a significant finding (see Figure 2.1). The red cell distribution width (RDW) is a measurement of the variation in size of erythrocytes and is higher in rabbits (11–15, IDEXX reference range) than in dogs and cats (8–10; Bush, 1991). In dogs and cats, anisocytosis is indicative of the presence of reticulocytes and a regenerative anaemia. In rabbits, 1–4% of circulating erythrocytes may be reticulotytes. Polychromasia and reticulocytes in rabbit blood smears have been attributed to the short life span and high turnover of erythrocytes (Kraus et al., 1984). Nucleated red cells and Howell–Jolly bodies can also be found occasionally (McLaughlin and Fish, 1994).

Rabbit neutrophils have an almost colourless cytoplasm and contain two types of granules. The smaller granules stain pink, giving a pinkish colour to the cytoplasm. Larger granules stain a deeper pinkish-red. The overall colour of the neutrophils varies according to the proportion of large and small granules. The granular appearance of the cytoplasm has led to different nomenclature. Rabbit neutrophils may be called heterophils, pseudoeosinophils, acidophils or amphophils, depending on the text (Benson and Paul-Murphy, 1999; Sanderson and Phillips, 1981). Neutrophils measure 10–15 μm in comparison with eosinophils, which measure 12–16 μm (Sanderson and Phillips, 1981).

Small lymphocytes are seen more commonly than large lymphocytes. The average cell diameter for small lymphocytes is 7–10 μm (Cooke, 2000). The lymphocytes are round cells with the typical morphology described for other species. An occasional large lymphocyte may be seen in the circulation or normal rabbits. These cells have a few azurophilic granules in the cytoplasm (Jain, 1986). Activated lymphocytes (immunocytes) typically have a more intensely stained cytoplasm.

Monocytes are large nucleated cells measuring 15–18 μm (Cooke, 2000). The nucleus has a diffuse lacey chromatin pattern that lightly stains a purple blue. The vacuolar cytoplasm stains light blue.

Eosinophils can be distinguished from neutrophils by their greater size and large acidophilic granules.

In contrast to other laboratory species, basophils are frequently found in the circulation of rabbits in small to modest numbers (Jain, 1986).

2.2.2 Interpretation of haematology results

A reference range for haematological parameters is given in Table 2.2. The haematological picture gives an indication of the general health status of a rabbit. Stress and a range of diseases will alter haematological parameters. Hinton et al. (1982) analysed the haematological findings in 117 healthy and diseased rabbits and found that blood cellularity was a good indicator of disease especially with regard to erythrocytes (these tended to be depressed) and lymphocyte counts (the counts were variable; however, heterophils tended to become dominant in acutely ill animals). These findings are in agreement with studies of experimental infections in rabbits (Toth and Krueger, 1988, 1989) and in a clinical study by Harcourt-Brown and Baker (2001). In this study, significantly higher red cell counts, haemoglobin values, haematocrits and lymphocyte counts were found in rabbits kept outside with unlimited access to grazing and exercise. A comparison was made with rabbits kept in hutches and those suffering from dental disease.

Reference ranges for packed cell volumes (PCV) vary between sources, with values between 30 and 50% (Malley, 1996). Pet rabbits tend to have PCV values at the lower end of the range, typically between 30 and 40% (Harcourt-Brown and Baker, 2001). Values greater than 45% are indicative of dehydration, especially in rabbits suffering from gut motility problems. Values of less than 30% indicate anaemia that can be classified into non-regenerative and regenerative in a manner similar to that of other species.

Regenerative anaemia is associated with chronic blood loss, where compensatory mechanisms are attempting to redress the balance. Common causes include heavy flea infestation or from sites that intermittently bleed such as a uterine endometrial aneurysm or uterine adenocarcinomas, a common finding in middle-aged unneutered does. Lead poisoning can result in a regenerative anaemia with the presence of nucleated erythrocytes, hypochromasia, poikilocytosis and cytoplasmic basophilic stippling (Fudge, 2000). Non-regenerative anaemia is caused by diseases such as lymphoma or chronic renal failure. Autoimmune haemolytic disease has not been described as a clinical phenomenon in pet rabbits although there are reports that it occurs. Autoimmune haemolytic anaemia has been reported in laboratory rabbits in association with lymphosarcoma (Weisbroth, 1994). Chronic debilitating disease such as dental disorders or abscesses often cause a mild anaemia in pet rabbits (Harcourt-Brown and Baker, 2001).

Nucleated red blood cells can be associated with acute infectious processes although a few may be present in normal blood films. Experimentally, infections with Escherichia coli and Staphylococcus aureus cause an increase in number of circulating nucleated red blood cells during the septicaemic phase of the disease (Toth and Krueger, 1988, 1989).

2.2.4 White blood cells

2.3.1 Glucose

Herbivores, such as rabbits, differ from carnivores in their carbohydrate metabolism. Carnivores eat periodically and have sudden large intakes of nutrients that must be stored for utilization during the fast between meals. Herbivores graze for long periods of the day and are continually absorbing nutrients from the digestive tract. In rabbits, volatile fatty acids are produced from bacterial fermentation in the caecum and are continually absorbed as an energy source. A fasting sample is difficult to obtain from a rabbit. Withholding food does not prevent caecotrophy and the digestion of caecotrophs provides a source of glucose. Blood samples taken after 96 h of food deprivation may show no alteration in blood glucose levels (Kozma et al., 1974).

Hyperglycaemia is a relatively common finding in rabbits and can be accompanied by glycosuria. Handling alone can cause an increase in blood glucose to the order of 8.5 mmol/L experimentally (Knudtzon, 1988) and 15 mmol/L or more anecdotally. Diabetes mellitus has not been described in pet rabbits and there is some difference of opinion about its importance as a clinical disease (Hoefer, 2000; Jenkins, 2000; Rosenthal, 2000). Herbivorous animals withstand the absence of insulin more readily than carnivorous ones (Bentley, 1998) and are therefore not so susceptible to diabetes mellitus. Diabetes mellitus has been induced in laboratory rabbits by the administration of alloxan. It has also been described as a hereditary disease in rabbits. A laboratory strain was selectively bred as an animal model of human diabetes mellitus (Roth and Conaway, 1982). Affected animals were polydipsic, polyuric and polyphagic with severely impaired insulin release. Elevated glycosylated haemoglobin values of 12.2% were observed in the overtly diabetic animals in comparison with 3.9% in normal animals. Increased glycosylated haemoglobin levels did not correlate with plasma glucose concentrations (Cannon and Conaway, 1981). Histologically there was hypergranulation of β-cells of the islets of Langerhans. Obesity and ketoacidosis were not features of diabetes mellitus in the laboratory rabbits. The hyperglycaemia was in the region of 540–590 mg/dL (30–33.4 mmol/L) and there was marked glycosuria. Roth and Conaway (1982) described the maintenance of one diabetic individual on insulin at a dose of up to 8 units per day for 3 years. Ketonuria was not observed.

In pet rabbits, a diagnosis of diabetes mellitus cannot be made on a single blood sample and requires serial blood and urine sampling to confirm the diagnosis, as well as an evaluation of fructosamine levels (performed using methodology validated for rabbits). In view of the physiological factors that can increase blood glucose levels it is advisable to take repeat blood samples from hyperglycaemic rabbits with time of day, phase of digestion, anaesthesia, influence of handling or car journeys in mind. Mild glycosuria is not a significant finding.

Hyperglycaemia can be seen in the terminal stages of gut stasis and is a poor prognostic sign (Harcourt-Brown, 2011). It is associated with fatty degeneration of the liver at post-mortem examination. Marked hyperglycaemia is also seen in association with painful conditions such as acute intestinal obstruction. Blood glucose levels can rise to 20–25 mmol/L and return to normal once the condition is resolved (Harcourt-Brown, 2011). Experimental haemorrhagic or traumatic shock results in hyperglycaemia proportional to the severity of the condition. Hyperthermia also results in hyperglycaemia (McLaughlin and Fish, 1994). Diseases that elevate serum glucose levels in other species, such as hyperadrenocorticism or acute pancreatititis, have not been reported in pet rabbits, although they could occur.

Hypoglycaemia is a significant finding in rabbits and is associated with anorexia, starvation or disturbances in the digestion and absorption of carbohydrates. It can be a sign of hepatic dysfunction. A drop in blood glucose leads to mobilization of free fatty acids from adipose tissue and contributes to the development of ketoacidosis and fatty degeneration of the liver. Measurement of serum glucose is of value in moribund rabbits as a basis for the selection of appropriate fluid therapy. Septicaemia is another potential differential diagnosis for hypoglycaemia. Other causes of hypoglycaemia such as Addison’s disease or insulinomas have not been reported in pet rabbits although such conditions could occur.

2.3.2 Total protein

Interpretation of total protein concentrations is similar to other mammals. Artefactual increases in protein concentrations can result from excessive venous stasis during blood collection (for example, when it takes a long time to get a blood sample). Fluid and small molecules leave the plasma, resulting in a relative increase in proteins. This situation can occur in rabbits, especially in miniature breeds with small veins.

An increase in total protein indicates dehydration and chronic and immune-mediated disease. In rabbits dehydration due to water deprivation or gastrointestinal disturbances commonly occur. Examination of the haematocrit and albumin and globulin fractions can assist differential diagnosis.

Liver disease, chronic enteropathy, starvation or malnutrition may result in reduced protein levels. Glomerulonephropathy or protein-losing enteropathy are uncommon conditions that could cause low total protein levels in rabbits. A decrease in both albumin and globulin may be associated with haemorrhage or exudative skin lesions such as fly strike.

2.3.3 Albumin

The liver is the sole site of albumin synthesis and hypoalbuminaemia is a feature of advanced liver disease in all species. In rabbits, heavy parasitism can be a cause of liver disease. Eimeria stiedae causes hepatic coccidiosis (see Section 8.10.1.2). Cysticercus pisiformis, the larval stage of Taenia pisiformis, migrates through the liver and results in the development of fibrous tracks and necrotic foci (see Section 14.3.2). Severe infestations can result in low albumin levels. Non-hepatic causes of low serum albumin include glomerulonephropathy, protein-losing enteropathy, malabsorption and cardiogenic ascites.

Laboratory reference ranges for serum albumin levels in rabbits can be wide and vary between sources. Sex differences have been reported in laboratory rabbits. One study showed that female New Zealand white rabbits had higher serum albumin levels than males, although other studies have found no sex differences (Kozma et al., 1974).

In rabbits, hypoalbuminaemia is most likely to be associated with nutritional factors such as abnormal caecotrophy, incorrect diet, starvation or malnutrition associated with dental disease. Primary or secondary hepatic neoplasia occasionally occurs in pet rabbits. Hepatic coccidiosis is a cause of low serum albumin levels, especially in young rabbits that have been kept in colonies.

A high serum albumin concentration is not a feature of any specific disease, although an increased albumin level in conjunction with a raised PCV is indicative of dehydration. In a study by Harcourt-Brown and Baker (2001), pet rabbits kept under free-range conditions had significantly higher serum albumin concentrations than rabbits suffering from advanced dental disease. They were also significantly higher than rabbits kept in hutches that were not suffering from dental problems. The difference in albumin levels was attributed to differences in diet and husbandry. Caecotrophs are a source of amino acids for rabbits and normal caecotrophy is an important element of their protein metabolism. Low fibre diets, obesity, dental disease or skeletal abnormalities can prevent rabbits ingesting caecotrophs from the anus and reduce the available amino acids for protein synthesis. Rabbits kept in hutches are more likely to be eating a low fibre diet and to suffer from obesity and skeletal problems than rabbits living outside with unrestricted access to natural vegetation and exercise.