N. Clancey, BSc, DVM, MVSc, DACVP
Department of Pathology and Microbiology, Atlantic Veterinary College, Charlottetown, PEI, Canada
Routine evaluation of cavitary effusion fluid includes its gross appearance, total nucleated cell count (TNCC), protein concentration, and microscopic findings. Because a specific pathogenesis may be found, microscopic findings are of most importance. Cellular components in effusions typically include inflammatory cells, mesothelial cells, and neoplastic cells. The numbers, proportions, and types of cells help classify and determine the reason for the effusion. A specific cause may be determined (such as finding bacteria, fungi, parasites, lipid droplets, urine crystals, sperm, bile, plant material, and barium crystals). Additional tests (such as measurement of fluid triglyceride, cholesterol, creatinine, bilirubin, glucose, and lactate concentrations) may further characterize the fluid.
Transudates are often clear to medium yellow and transparent to mildly turbid. They have low TNCCs, with counts <5.0x109 cells/L and typically <1.5x109 cells/L, and are either protein-poor (<25 g/L) or protein-rich (>25 g/L). Variable proportions of macrophages and non-degenerate neutrophils are typical, with fewer lymphocytes and mesothelial cells. Mesothelial cells may be found individually or in single-layer sheets with bland uniform features. However, they frequently have dark-blue cytoplasm and variable degrees of anisocytosis, anisokaryosis, multinucleation, and other features found in malignant cells. Differentiating reactive mesothelial cells from exfoliated mesothelioma cells is extremely difficult to impossible; a diagnosis of mesothelioma should not be based solely on cytology. Erythrocytes from contamination, diapedesis associated with increased hydraulic pressure, or hemorrhage may be seen.
Exudates are typically cream to yellow and lightly to markedly turbid or flocculent, with TNCCs >5.0x109 cells/L and protein concentrations >25 g/L. Exudates are typically predominated by neutrophils (>70–85%), especially when effusions are caused by bacterial infection. Variable proportions of neutrophils, macrophages, and other inflammatory cells may also be seen. This is defined as mixed inflammation; it is typical of chronic inflammation caused by situations such as fungal infections, foreign material reactions, and neoplasia. A septic source, especially a bacterial one, is anticipated when neutrophils have degenerative changes.
These are characterized by cytoplasmic vacuolation and swollen nuclei, with loss of segmentation and pale staining chromatin. However, neutrophils may be degenerate with non-septic inflammation, particularly with bile peritonitis and uro-peritonitis. Exudative fluid warrants culture whether or not degenerative neutrophils or infectious organisms are observed microscopically. Effusions containing >10% eosinophils are characterized as eosinophilic effusions and are uncommon. When high numbers of eosinophils are present, these fluids may be grossly tinted green. Eosinophilic effusions are associated with neoplasia (particularly mast cell tumors and lymphoma), hypersensitivity conditions, dirofilariasis, sarcocystosis, Aelurostrongylus, and select neoplasms.1-3
A difference of serum to effusion glucose >1.11 mmol/L has been shown to support bacterial peritonitis.4 Lower effusion glucose concentrations may be due to utilization by bacteria and/or cells. Increased effusion concentrations of lactate, a product of anaerobic glycolysis, have been described with bacterial exudates.4,5 The increased lactate may be from bacteria, leukocytes, and erythrocytes. Increased effusion concentrations of lactate have also been observed in dogs with bowel strangulation6 and abdominal neoplasia7. As cellular metabolism continues post-sampling, glucose and lactate require prompt analysis. A portion of these fluids should be submitted in special tubes containing sodium fluoride to inhibit glycolysis.
Feline infectious peritonitis warrants special mention as effusion formation is exudative, yet fluids tend to have low cellularity (<5.0x109 cells/L) and markedly increased protein concentrations (often >45 g/L). This is because the inflammation is a vasculitis. Leukocyte chemotaxis to body cavities is not prominent, yet damaged vessels leak large amounts of protein. Fluids are grossly light to moderately yellow, transparent to mildly turbid, and may or may not contain fibrin strands or clots. Microscopically, fluids contain a mixture of non-degenerate neutrophils (∼60–80%) and macrophages (∼20–40%) with few lymphocytes/plasma cells and mesothelial cells. Given the high protein concentrations, the stained smear background is typically lavender to purple, stippled with precipitated protein. Protein crescents are often seen.
Hemorrhagic effusions are grossly bloody during collection and do not clot in non-anticoagulant tubes. The TNCC, erythrocyte count, and total protein concentration may be similar to but are usually lower than that of blood. Concurrent inflammation will increase the TNCC. Nucleated cells consist of leukocytes, low numbers of macrophages, and mesothelial cells.
Erythrocytes predominate and platelets are absent. Erythrocytophagia, macrophages containing hemosiderin and/or hematoidin crystals found intracellularly or extracellularly support previous hemorrhage and are key diagnostic features. When present, platelets support either very recent pathological hemorrhage or iatrogenic hemorrhage. Trauma, hemostatic disorders, and neoplasia are classic sources of hemorrhagic effusions. Various infectious agents also can result in hemothorax.8,9 Idiopathic pericardial effusion in dogs is well-recognised and may warrant further evaluation for vector-borne disease.10
Chylous effusions are grossly white, creamy to pale pink-red, and opaque. An upper creamy lipid layer may present if the fluid is refrigerated or left standing. The TNCC is highly variable but usually >2x109 cells/L and <15x109 cells/L in the author’s experience. Protein concentration is often >25 g/L but may be falsely increased by lipemic artifact. Small lymphocytes typically predominate with fewer non-degenerate neutrophils and macrophages. Macrophages usually contain numerous round, uniform, clear lipid vacuoles. With chronicity and irritation by chyle, neutrophils may outnumber lymphocytes. Fluid triglyceride concentrations >1.1 mmol/L confirm chylous effusion.11 However, anorexic patients may have lower triglyceride concentrations and the fluid may not be grossly chylous. Historically, trauma has been a commonly listed source for chylothorax. However, external trauma is a less common source than lymphorrhage associated with altered lymphatic drainage. Causes include increased systemic venous pressure such as seen with right-sided heart failure, heartworm disease, and pericardial effusion, or with lymphatic compression by neoplasia, granulomas, or lung lobe torsion. When an underlying source cannot be isolated, the chylous effusion is characterized as idiopathic. Chylous peritoneal effusions are rare and typically associated with lymphangiectasia or lymphatic obstruction.
Uro-peritoneal fluid may be clear to medium yellow and transparent to mildly turbid. Acutely, findings mimic that of a protein-poor transudate. However, with time and mesothelial irritation from urine, peritonitis develops. While clues such as urinary crystals and spermatozoa may aid diagnosis in select cases, uro-peritoneal fluid usually lacks specific findings. Clinical suspicion of uroperitoneum and creatinine evaluation are required. Fluid creatinine concentrations >2 times that of concurrent serum samples support uroperitoneum.12 However, as time and pre-existing azotemia will impact fluid creatinine concentration, lower values do not completely exclude uroperitoneum.
Bilious effusions are classically green to green-brown and moderately turbid to opaque. Depending on chronicity and leakage volume, effusions may resemble transudates. However, an exudate with mixed inflammation predominated by neutrophils and variable numbers of macrophages is typical. Presence of intra- or extracellular, amorphous to granular, yellow to green-brown bile pigment provides an excellent diagnostic clue. The fluid bilirubin concentration can also be assessed and is usually >2 times that of serum bilirubin concentration.13 Ultraviolet light degrades bilirubin, so samples should be protected from light. Bilirubin concentrations may be low in patients with ruptured gall bladder mucoceles. In such patients, the mucinous component of bile (referred to as white bile) may be observed in the effusion. It appears as amorphous to fibrillar, sky to deeper blue, extracellular material on Wright-Giemsa-stained smears; blue with Alcian blue-PAS stain; and deep red with mucicarmine stain.14
Bilious effusions are typically due to trauma, mucoceles, cholelithiasis, neoplasia, and necrotizing inflammation. Gut leakage or rupture generates effusions that are grossly tan to brown and variably turbid. Fluid may be flocculent with food or particulate matter. The TNCC and protein concentration vary, and volume may be low in acute cases. Automated TNCCs should not be trusted, as organisms and debris may be falsely counted as cells. Ultimately, a septic peritonitis develops. Concurrent hemorrhage may be seen. Finding phagocytized bacteria is a key differentiating factor from partial enterocentesis. Incidental enterocentesis should contain only intestinal contents, lacking inflammatory cells. Severe inflammation, trauma, obstruction, torsion, and strangulation are some sources.
Although not part of the mechanistic classification of effusions, neoplastic effusions deserve recognition. They often allow an immediate, specific diagnosis and are created by one or more of the five effusion mechanisms. Neoplastic effusions are highly variable in gross appearance, TNCC, and protein concentration due to variation in duration, tumor type, and presence of secondary inflammation and/or hemorrhage. Round and epithelial cell neoplasms readily exfoliate cells. Sarcomas rarely cause effusions, and their cells rarely exfoliate cells. Common neoplasms include lymphoma, carcinomas, and adenocarcinomas, with mast cell neoplasia and mesothelioma less commonly observed. Lymphoma of granulated lymphocytes may be difficult to differentiate from mast cell neoplasia on cytology alone. Morphology may strongly overlap between some lymphomas, carcinomas, and mesotheliomas, requiring advanced modalities such as flow cytometry and clonality testing for diagnosis.
1. Fossum TW, Wellman M, Relford R, et al. Eosinophilic pleural or peritoneal effusions in dogs and cats: 14 cases (1986–1992). J Am Vet Med Assoc. 1993;202:1873–1876.
2. Allison R, Williams P, Lansdowne J, et al. Fatal hepatic sarcocystosis in a puppy with eosinophilia and eosinophilic peritoneal effusion. Vet Clin Pathol. 2006;35:353–357.
3. Miller BH, Roudebush P, Ward HG. Pleural effusion as a sequela to aelurostrongylosis in a cat. J Am Vet Med Assoc. 1984;185(5):556–557.
4. Bonczynski JJ, Ludwig LL, Barton LJ, et al. Comparison of peritoneal fluid and peripheral blood pH, bicarbonate, glucose, and lactate concentration as a diagnostic tool for septic peritonitis in dogs and cats. Vet Surg. 2003;32(2):161–166.
5. Levin GM, Bonczynski JJ, Ludwig LL, et al. Lactate as a diagnostic test for septic peritoneal effusions in dogs and cats. J Am Anim Hosp Assoc. 2004;40(5):364–371.
6. DeLaurier GA, Cannon RM, Johnson RH Jr, et al. Increased peritoneal fluid lactic acid values and progressive bowel strangulation in dogs. Am J Surg. 1989;158(1):32–35.
7. Nestor DD, McCullough SM, Schaeffer DJ. Biochemical analysis of neoplastic versus non-neoplastic abdominal effusions in dogs. J Am Anim Hosp Assoc. 2004;40:372–375.
8. Sasanelli M, Paradies P, Otranto D, et al. Haemothorax associated with Angiostrongylus vasorum infection in a dog. J Small Anim Pract. 2008;49(8):417–420.
9. Byun JW, Yoon SS, Woo GH, et al. An outbreak of fatal hemorrhagic pneumonia caused by Streptococcus equi subsp. zooepidemicus in shelter dogs. J Vet Sci. 2009;10(3):269–271.
10. Tabar MD, Movilla R, Serrano L, et al. PCR evaluation of selected vector-borne pathogens in dogs with pericardial effusion. J Small Anim Pract. 2018;59(4):248–252.
11. Waddle JR, Giger U. Lipoprotein electrophoresis differentiation of chylous and nonchylous pleural effusions in dogs and cats and its correlation with pleural effusion triglyceride concentration. Vet Clin Pathol. 1990;19(3):80–85.
12. Schmiedt C, Tobias K, Otto C. Evaluation of abdominal fluid: peripheral blood creatinine and potassium ratios for diagnosis of uroperitoneum in dogs. J Vet Emerg Crit Care. 2001;11(4):275–280.
13. Ludwig LL, McLoughlin MA, Graves TK, et al. Surgical treatment of bile peritonitis in 24 dogs and 2 cats: a retrospective study (1987–1994). Vet Surg. 1997;26(2):90–98.
14. Owens SD, Gossett R, McElhaney MR, et al. Three cases of canine bile peritonitis with mucinous material in abdominal fluid as the prominent cytologic finding. Vet Clin Pathol. 2003;32:114–120.