The physical appearance of a fluid together with its total nucleated cell concentration (TNCC), refractometric protein concentration, and microscopic appearance are used routinely to evaluate the potential causes of cavitary effusions. Microscopic evaluation is most important, because it can reveal specific pathogeneses. Cells that are routinely identified in effusions are hemic cells (neutrophils, macrophages, lymphocytes, eosinophils, mast cells, erythrocytes, and platelets), mesothelial cells, and neoplastic cells. The types, numbers, and proportions of cells reflect the pathologic processes, and the staining intensity of the spaces around the cells reflects the protein concentration. More specific information may be provided by other findings, including bacteria, fungi, larvae, urine crystals, sperm, bile, mucus, plant material, barium crystals, and lipid droplets. When specific findings are not present, TNCC, cell types, and protein concentrations help to classify the effusion and the pathologic process responsible for it. Other specific tests described below may further characterize the fluid and its cause. Case studies will be presented to illustrate the process and findings.
TNCCs are usually < about 1,500/µL in low-protein transudates and possibly up to 5,000/µL in high-protein transudates, thus requiring concentrated preparations. Most of the nucleated cells are macrophages and nondegenerate neutrophils (either may predominate). There are fewer lymphocytes, and very few other cells with the possible exception of erythrocytes and mesothelial cells. Few to many mesothelial cells may be present as individual round cells or sheets of epithelial cells. They are frequently reactive, having varying degrees of increased cytoplasmic basophilia, anisocytosis, anisokaryosis, multinucleation, and mitotic activity.
Exudates may be proteinaceous, cellular, or both. Neutrophils are the predominant nucleated cell in most exudates, but mixtures of mostly neutrophils and macrophages are common. Eosinophils occasionally predominate, suggesting parasitism, hypersensitivity, or neoplasia (e.g., mast cell neoplasia). Most exudates have a protein concentration > 2.0 g/dL and a TNCC > 5,000/µL, but low-grade inflammation may yield lower values and may be either the primary cause or a secondary contributor to the effusion. Whenever an exudate is present, a cause should be sought. Efforts to detect phagocytized bacteria should be increased when neutrophils appear degenerate (i.e., they have acquired a swollen and pale-staining nucleus and variable degrees of cytoplasmic vacuolization at the site of inflammation). Routine Romanowsky stains are better than Gram staining to detect bacteria. Culturing may better detect low numbers of organisms, but culture results may be negative because of inappropriate culture media or antibiotic treatment prior to fluid collection. Fungi, protozoa, and foreign material may also be found in neutrophils, but they are commonly found extracellularly or in macrophages. An exudate can be labeled septic or infectious once organisms are detected, but it should not be labeled nonseptic or noninfectious simply because organisms were not found microscopically.
Compared to most exudates, the high-protein exudate typical of feline infectious peritonitis (FIP) tends to have a low cellularity because the inflammatory process is in the blood vessels and not in the body cavity. Microscopic examination of a stained preparation may reveal pink granular staining of the fluid reflective of a high protein concentration. Proteins detected in effusions mostly represent those that leaked through capillary walls. The electrophoretic patterns of exudates and sera in cats with FIP are very similar because vasculitis enables plasma proteins to ooze from the blood. In people, a serum albumin-ascites gradient (serum albumin concentration--effusion albumin concentration) has been used to differentiate causes of peritoneal effusions; the gradient is greater in cirrhosis and congestive heart failure than in malignancies or inflammatory states. This gradient has been evaluated in dogs.1
Glucose and L-Lactate Concentrations
Because glucose in an effusion comes from plasma, it is important to interpret effusion glucose concentrations in relation to blood glucose concentrations. The glucose concentrations in peritoneal effusions were less in dogs and cats with bacterial exudates than in dogs and cats that had nonbacterial peritoneal effusions,2 and the glucose concentrations in bacterial exudates were less than the corresponding blood glucose concentrations in all dogs and nearly all cats. Low glucose concentrations in these effusions may be caused by the bacteria and/or a greater consumption of glucose by the greater number of cells in these fluids.
Increased effusion concentrations of L-lactate, a product of anaerobic glycolysis in mammalian cells and some bacteria, have been found in animals with strangulated and nonstrangulated intestinal obstructions,3 abdominal neoplasms,4 and bacterial exudates.2,5 Greater values may occur with neoplasia because of increased anaerobic glycolysis by neoplastic cells or interference of blood supply to nonneoplastic tissue. The source of L-lactate in bacterial effusions may include bacteria, leukocytes, and erythrocytes.
Because of cellular metabolism in vitro, samples for these analytes need to be analyzed quickly, placed in sodium fluoride tubes, or processed by removing supernatant from cells after centrifugation.
Erythrocytes are found in nearly all samples because of either iatrogenic (during fluid collection) or pathologic hemorrhage. Pathologic hemorrhage is supported by the presence of hemosiderin, hematoidin crystals, or erythrophagocytosis in fluid that appeared bloody throughout the collection process, together with the absence of platelets in EDTA samples. The presence of platelets suggests ongoing or iatrogenic hemorrhage. Erythrophagocytosis may also occur in vitro, so preparations from fresh samples should be evaluated. When the fluid Hct is greater than 3%, hemorrhage from any cause may be a significant contributor to the effusion.
The two most common causes of a lymphocyte-rich effusion are lymphoid neoplasia and an accumulation of lymph. These can usually be differentiated microscopically. Neoplastic lymphocytes are usually intermediate or large with basophilic cytoplasm, finely stippled chromatin, and prominent nucleoli. Nonneoplastic accumulations consist mostly of small lymphocytes typical of those seen in blood films; reactive forms may be present. Additional diagnostic methods might include clonality studies of cells in the fluid or histologic examination of excised tissue. With chylous effusions, chylomicrons may be abundant and thus create a creamy white fluid that contains macrophages with punctate cytoplasmic Sudanophilic lipid vacuoles.
Cholesterol and Triglyceride Concentrations
Measurement of cholesterol and triglyceride concentrations in the animal's serum and effusion may be used to help identify chylous effusions.6,7 Triglyceride concentrations are greater in chylous than in nonchylous effusions, and thus effusion ratios of cholesterol to triglyceride are lower (< 1 for chylous; > 1 for most nonchylous). Disorders associated with chylothorax include lymphangiectasia, thoracic neoplasia, venous thrombosis, feline cardiomyopathy, diaphragmatic hernia, and idiopathic chylothorax.
Effusions formed from rupture of the biliary tract, the urinary tract, or the alimentary tract may contain microscopic clues of their origin (bile, crystals or sperm, ingesta and organisms, respectively). An early effusion caused by rupture of the urinary tract may have a low TNCC because of urine entering the cavity, but initiation of an inflammatory response leads to features of an exudate. Effusions caused by alimentary tract rupture are usually septic and have high TNCCs. Bile often induces a low-grade exudate.
Urea and Creatinine Concentrations
Paired measurements of serum and peritoneal creatinine or urea concentrations may be helpful in identifying uroperitoneum. When there is a recent addition of urine to peritoneal fluid, the resulting fluid will have greater urea and creatinine concentrations than will the plasma. With time and diffusion, concentrations in extravascular and intravascular fluids will become similar again, and both will be increased. Creatinine diffuses and equilibrates more slowly than urea, and its effusion concentration is often at least twice the serum concentration.8 However, the ratio of serum and effusion values will be affected by factors such as the duration of the uroperitoneum or pre-existing azotemia, so lower ratios do not exclude uroperitoneum.
When blunt or penetrating trauma, necrotizing cholecystitis or cholangitis, cholelithiasis, or biliary neoplasia cause leakage of bile into the peritoneal cavity, peritonitis develops and the resultant exudate typically has a protein concentration > 2.5 g/dL, a TNCC > 5,000/µL, and > 80 % neutrophils. When microscopic examination does not detect intracellular or extracellular amorphous yellow to brown bile pigment, the effusion can be analyzed for bilirubin. The bilirubin concentration in bile-containing effusions is typically at least twice an animal's serum bilirubin concentration.9.10 However, effusion bilirubin concentrations may not be increased despite gallbladder damage. Bilirubin is degraded by ultraviolet light, and thus effusion samples should be protected from light.
Most neoplasms that cause effusions do not slough cells into cavitary fluid; they typically cause an effusion by other processes (exudation, transudation, hemorrhage, or damaged lymphatic system). The features of the effusions can vary from only rare neoplastic cells among many inflammatory cells to extremely high concentrations of neoplastic cells. Neoplastic cells that may be detected in effusions include lymphocytes, carcinoma cells, mesothelial cells, mast cells, melanoma cells, and rarely spindle cells. Cells exfoliating from carcinomas frequently have pleomorphic features as in tissue, but it may be impossible to differentiate carcinoma cells from neoplastic or highly reactive mesothelial cells found in nonneoplastic effusions. Sarcomas typically do not exfoliate cells or cause effusions, but hemangiosarcomas might rupture and cause a hemorrhagic effusion. A few mast cells are not unusual in exudates and are also found in other effusions. Unless the cells have prominent microscopic features of malignancy, neoplastic and nonneoplastic mast cells cannot be reliably differentiated via microscopy.
Successful analysis and interpretation of cavitary effusions may be limited by sample volume or quality (e.g., aged sample), poor slide preparations, suboptimal staining, an inadequate microscope, microscopy expertise, omission of potentially useful special tests, or understanding. Nonspecific findings are common, but they may narrow the hunt and provide diagnostic direction. Specific findings are common enough that every effort should be made to thoroughly evaluate all cavitary effusions.
1. Pembleton-Corbett JR, et al. J Vet Intern Med 2000;14:613.
2. Bonczynski JJ, et al. Vet Surg 2003;32:161.
3. DeLaurier GA, et al. Am J Surg 1989;158:32.
4. Nestor DD, et al. J Am Anim Hosp Assoc 2004;40:372.
5. Levin GM, et al. J Am Anim Hosp Assoc 2004;40:364.
6. Fossum TW, et al. J Am Vet Med Assoc 1986;188:49.
7. Waddle JR, et al. Vet Clin Pathol 1990;19:80.
8. Schmiedt C, et al. J Vet Emerg Crit Care 2001;11:275.
9. Ludwig LL, et al. Vet Surg 1997;26:90.
10. Owens SD, et al. Vet Clin Pathol 2003;32:114.