The generic term “cavitary effusion” defines the accumulation of fluid within a coelomic cavity. The pathogenesis of effusions may be explained by one of the following mechanisms.
Altered pressure balance
The volume of fluid present in cavities depends on the equilibrium between the hydrostatic pressure exerted by blood on vessel walls (Pcap), which tries to push fluids out of the vessel, and the oncotic or colloidal-osmotic pressure (POcap), generated by the macromolecules (albumins and other proteins) present in the plasma, which attracts fluids by osmosis into the vessel. These forces are opposed by the analogous pressure forces which are present within the interstitium (Pint and POint). An additional component that regulates the exchanges of fluids is the permeability of the vessel wall, or hydraulic conductivity (HC), which is high in subserous capillaries. These factors are related in accordance with Starling’s law:
Net fluid flux (ml/min) = HC x [(Pcap – Pint) - (POcap - POint)]
In normal subserous arterioles, Pcap – Pint is equal to 37 mmHg and POcap – POint to 25 mmHg: fluids, therefore, tend to leave the vessels leading to a progressive decrease of the Pcap. In capillaries, the two pressures tend to reach an equilibrium, while in venules POcap – POint (25 mmHg) becomes greater than Pcap – Pint (17 mmHg) and fluids, therefore, tend to return into the vessels and do not accumulate in cavities. This equilibrium may be altered in the conditions described below.
A reduction in the oncotic pressure: if POcap decreases, the equilibrium shifts in favour of the hydrostatic pressure and the exit of fluids from the vessels prevails over reabsorption. POcap decreases in cases of hypoproteinaemia (e.g. due to severe liver diseases such as liver failure or shunt, protein-losing nephropathies or enteropathies, malabsorption, intestinal parasitosis, chronic haemorrhage, neoplastic cachexia, malnutrition and dietary deficiencies). In chronic liver disease there is also an increase in the hydrostatic pressure caused by portal hypertension, as the hypovolaemia associated with the loss of fluids into cavities stimulates the activity of the renin- angiotensin-aldosterone system, which causes a rise in blood pressure. The effusion resulting from a decrease in POcap is poor in proteins and cells and is called a transudate.
An increase in the hydrostatic pressure: an increase in Pcap, not accompanied by variations of POcap, also alters the pressure equilibrium and favours the movement of fluids out of the vessels. The increase in Pcap may be systemic (e.g. due to heart failure) or localised (e.g. due to occlusion of vessels caused by external masses, strangulated hernias, torsion of organs; vascular obstruction caused by intravascular masses such as thrombi, parasites, tumours; liver fibrosis or cirrhosis with consequent portal hypertension). When the hydrostatic pressure increases, the fluid component as well as small quantities of proteins and/or cells can leave the vessels, giving rise to a so-called modified transudate. When the modified transudate is caused by the presence of a neoplastic mass the cellularity of the effusion may by elevated because of exfoliation of the tumour cells (neoplastic effusion). If the increase of Pcap occurs in lymphatic vessels instead of blood vessels (e.g. cardiopathies, neoplasms or other mediastinal masses, diaphragmatic hernias, torsion of lung lobes) lipid-rich fluids and small lymphocytes present in the lymph spill into the cavity, giving rise to a chylous effusion or chyle.
Vessel wall alterations
Effusions may result from the rupture or increased permeability of vessel walls.
Rupture of a vessel wall: effusions may occur following direct rupture of medium-large vessels (following trauma or diseases that weaken the vessel wall; e.g. aneurysms, angiomas) or of organs rich in blood (liver, spleen; also in this case the rupture is facilitated by diseases that make the organs more fragile; e.g. neoplasias, degenerations, blood stasis). In the latter case the effusion is haemorrhagic. If lymphatic vessels are ruptured the effusion is chylous. If fluid-containing organs are involved (urinary bladder, gallbladder), the haemorrhagic component is less than the fluid content of the organ, giving rise to uroperitoneum or chyloperitoneum, respectively.
Permeabilisation of a vessel wall: effusions are usually associated with a serosal inflammatory process or vasculitis. In most cases this takes place following systemic infections (especially viral) or immune-mediated disorders (caused by immune complexes), which are in fact often of infective aetiology. The infectious agents may, however, have been carried into the cavity by a perforating foreign body. An inflammatory response of variable degree may also develop when the cavity contains a neoplastic mass or an effusion of primarily non-inflammatory nature, especially if this contains potentially irritating chemical substances (e.g. chyle, urine, bile).
The above mentioned scenarios are followed by a typical inflammatory process, with its usual sequence of events (vasodilatation, vessel wall permeabilisation, extravasation of leukocytes). The effusion is consequently rich in cells and proteins and is classified as an exudate.
ANALYSIS OF EFFUSIONS
From a diagnostic point of view the analysis of effusions is based on a series of successive steps, with the intention of determining the:
- macroscopic appearance of the sample: specifically the colour, transparency or turbidity (possibly associated with the presence of particulate material in suspension, such as fibrin floccules), viscosity and odour of the effusion fluid;
- number of cells present: cells are counted with a haemocytometer or automated cell counter. If the effusion is particularly viscous, automated counting can be difficult (the fluid is not aspirated) or inaccurate (the adherent cells in the fluid are counted in groups instead of separately, causing an underestimate of the cell count). The automated count is also not accurate if the effusion contains a number of cells inferior to the analytical sensitivity of the cell counter. Immediately after the count a slide should be prepared for cytology: to do this the fluid can be smeared directly on the slide, or centrifuged first with subsequent smearing of the pellet obtained or, better still, it can be centrifuged with a special cytocentrifuge. The direct smear is preferable if the samples are extremely cellular, while cytocentrifugation is preferable for samples with a limited number of cells. It is always advisable to prepare several slides so that one can be stained with routine dyes (May Grünwald-Giemsa, rapid dyes) for traditional cytological analysis, while the other (or others) may be preserved for special stains or immunostains, should these techniques be necessary to complete the diagnosis;
- protein content and specific gravity: these can be estimated using a refractometer on the supernatant obtained by centrifuging the sample, although it should be kept in mind that when the protein concentration is low (under 20 g/l) and in cases of haemorrhagic, biliary or chylous effusions, the refractometric estimate is not accurate. For a more accurate assessment of the protein content it is better to use high-sensitivity biochemical assays (e.g. pyrogallol red);
- biochemical composition: should particular diseases be suspected, the following biochemical tests can be conducted on the supernatant: creatinine, urea or potassium (elevated in the case of uroperitoneum); albumin, with subsequent calculation of the albumin to globulin ratio (low in the case of feline infectious peritonitis [FIP]); globulin fractions, determined by electrophoresis (increased gamma-globulins in the case of FIP); bilirubin (high in choleperitoneum); lipase and/or amylase (elevated in the case of pancreatitis); glucose (low in neoplastic or septic effusions); lactate (elevated in septic forms); lactate dehydrogenase (elevated in neoplastic forms); triglycerides (higher in chyle than in serum and higher than cholesterol); cholesterol (higher in non-chylous effusions than in serum);
- presence of infectious agents or parasites: bacteriological tests should always be carried out, even in the case of phagocytosed bacteria, in order to be able to set up an antibiogram and start appropriate therapy. Should specific diseases be suspected (e.g. FIP, toxoplasmosis), the pathogens may be searched for using immunocytochemical techniques or polymerase chain reaction analysis, bearing in mind the limitations and advantages of each technique depending on the pathology under consideration.
CLASSIFICATION OF EFFUSIONS
Depending on the pathogenesis, effusions may have different macroscopic and physico-chemical characteristics. Based on these characteristics, effusions may be classified into transudates, modified transudates and exudates.
TRANSUDATES
Transudates are colourless or have a serum-like colour, are transparent, non-viscous and odourless, do not contain particulate matter and have a low specific gravity (<1.017), low protein concentration (<25 g/l) and low cell count (<1,000 cells/µl). The rare cells that are present are mesothelial cells, macrophages or neutrophils. Pure transudates are found in cases of severe hypoproteinaemia or hypoalbuminaemia, caused mostly by protein-losing enteropathies and severe liver diseases, which, compared to other causes of hypoproteinaemia, more commonly cause panhypoproteinaemia.
MODIFIED TRANSUDATES
Modified transudates are colourless or have the appearance of bloodstained serum, are transparent or slightly turbid, only slightly viscous and odourless and do not contain particulate matter. The specific gravity ranges between 1.017 and 1.025, the protein concentration is over 25 g/l (usually by just a little), and the cell count is moderate (between 1,000 and 5,000 cells/ml) and characterized by the presence of macrophages, lymphocytes and reactive mesothelial cells (Fig. 1). Special attention should be paid to the presence of reactive mesothelial cells (Fig. 2), which are morphologically similar to neoplastic cells (bulky cells with abundant basophilic cytoplasm, single or multiple nuclei with dense chromatin and evident nucleoli). Modified transudates often occur as a consequence of pressure alterations (although in cardiogenic forms they frequently acquire special characteristics, as described later) or are caused by the presence of masses or chronic reactive phenomena within the cavity.
EXUDATES
Exudates are the result of an inflammatory process. The colour of these fluids is variable, usually being yellow or brownish yellow; in the case of sepsis, the exudates sometimes contain blood. An exudate may be transparent or, more frequently, turbid and/or contain particulate material, such as fibrin floccules (especially in some diseases, such as FIP). The viscosity is variable (low in non-septic forms, higher in septic forms, extremely high FIP), as is the odour (effusions caused by anaerobic bacteria or Nocardia spp. are foul-smelling, the others tend to be odourless). Exudates have a high specific gravity (>1.025), protein concentration (greater, and often by a lot, than 25 g/l) and cell count (>5,000 cells/ml). The cells present are usually neutrophils - which are particularly abundant and degenerated in septic forms (Fig 3) and may contain phagocytosed bacteria (Fig. 4) – as well as macrophages, lymphocytes, reactive mesothelial cells (these cells are more often found in chronic forms) and sometimes other cells in the case of exudates having a particular pathogenesis (e.g. eosinophils in parasitic forms). These effusions (pyothorax, pyoabdomen, etc.) are usually caused by pyogenic bacteria or by Nocardia spp, and more rarely by bacteria from perforated abdominal organs. In the case of a septic effusion it is always important to carry out bacteriological tests in order to identify the bacteria responsible or, if these are visible cytologically, to perform an antibiogram.
In cats, exudates with particular characteristics are found during cholangiohepatitis and FIP. In both cases the exudate is transparent, yellow (more rarely pinkish or chyle-like), dense, filamentous and tends to coagulate when exposed to air acquiring a gelatinous texture or giving rise to fibrin floccules. The specific gravity is high, as is the protein content (even up to 50-60 g/l); the main proteins are fibrinogen (responsible for the above-mentioned coagulation) and gamma-globulins. For this reason, for diagnostic purposes it is extremely important to do protein electrophoresis [3] of the effusion or, at least assay not only the total proteins but also albumin, from which the globulin concentration can be derived and, therefore, the albumin to globulin ratio, which is low in the case of FIP. The Rivalta test is an empirical method used to estimate the type of proteins present: if high molecular weight proteins are present, when a drop of the effusion is put into acidified water a small protein clot is formed. The proteins present are also responsible for the typical microscopic appearance of these exudates, in which, in addition to the cellular component (non-degenerated neutrophils, lymphocytes, macrophages, plasma cells and reactive mesothelial cells), there is also a granular background (Fig. 5).
In the presence of effusions with the above mentioned characteristics, the differentiation between cholangiohepatitis and FIP must be based on additional laboratory data (e.g. full blood count, serum electrophoresis and especially serum a1acid glycoprotein concentration, which is particularly high in the case of FIP), on imaging studies - which can show findings more compatible with either cholangiohepatitis or FIP - or on a liver biopsy, which is the only investigation that can really resolve which of the two diseases is present. Determination of anti-coronavirus antibody titres and polymerase chain reaction analysis for coronavirus in the blood or effusion give high rates of false positive and false negative results and are consequently of limited diagnostic utility.
OTHER TYPES OF EFFUSION
The above basic classification has the limitation of including effusions with different pathogenises within the same class. Furthermore, it does not allow for forms with a mixed pathogenesis, which are actually the most common in clinical practice, and it does not include effusions which do not fall within the classical scenarios. For this reason, the following additional classes of effusions have been proposed, based on their macroscopic, cytological or biochemical characteristics, rather than on a simple determination of specific gravity, proteins and cellularity:
Uroperitoneum: this effusion is a citrus yellow colour, transparent, non-viscous and with the characteristic odour of urine. The specific gravity and the protein concentration are usually low, since water is attracted into the cavity by osmosis, with consequent dilution of the urine present within the abdomen. The effusion can contain cells (non-degenerated neutrophils and/or reactive mesothelial cells) and it has a high content of creatinine, urea and potassium ions.
Chyle (chylous effusion): the effusion is caused by lymphatic stasis (e.g. due to compression of intrathoracic lymphatic structures, torsion of lung lobes) or by the rupture of lymphatic ducts. The fluid is typically milky white or slightly pinkish in the case of contamination by blood. It is usually turbid, non-viscous and odourless. The cellularity is elevated and the effusion is characterized by the presence of small lymphocytes and macrophages which are often foamy because they phagocytose the lipids present in the lymph (Fig. 6). To confirm the diagnosis it may be useful to assay the concentrations of triglycerides, which are higher in chyle than in serum, and cholesterol (the cholesterol/triglycerides ratio is less than 1 in chyle). Cases with more abundant cholesterol and more numerous degenerated neutrophils were previously defined as pseudo-chylous, suggesting that in spite of the macroscopic analogies, the pathogenesis of the effusion is probably different (usually due to chronic pleural diseases). This distinction is, however, of limited practical relevance, and it is not even clear whether pseudo-chyle originates as an independent entity or whether, more probably, it is a type of evolution of a chylous effusion. For these reasons there is tendency is to abandon the use of the term pseudo-chylous effusion.
Haemorrhagic effusions: from a morphological point of view these effusions are very similar to blood samples (red, turbid, non-viscous, odourless) and following centrifugation they show the classical separation between plasma and the corpuscular component, although the haematocrit is 10-25% higher than that of blood. Cytology is of little help diagnostically since the appearance of the samples is often similar to that of a blood smear. There is frequently evidence of erythrophagocytosis, with macrophages containing red blood cells (recent erythrophagocytosis) or pigments resulting from the degradation of haemoglobin (haemosiderin, haematoidin). The presence of platelets indicates a recent haemorrhage or contamination by blood at the time of taking the sample. The finding of a haemorrhagic effusion (haemoperitoneum, haemothorax, etc.) must lead to a search for the cause of the haemorrhage (e.g. rupture of large vessels; organ rupture, especially if in the presence of stasis, neoplasms or degenerations; presence of coagulopathies). A special type of haemorrhagic effusion can be found during idiopathic pericarditis: apart from the haemorrhagic component, these effusions contain mesothelial cells that are morphologically extremely anomalous, again often containing haemoglobin residues (Fig. 7).
Choleperitoneum (biliary peritonitis): effusions due to rupture of the gallbladder have a brownish-green colour, are usually moderately viscous and odourless and appear turbid, sometimes with darker material in suspension. Following centrifugation the supernatant, which is also greenish, may appear slightly more transparent. In the case of choleperitoneum resulting from rupture of the liver, blood may be present and visible both macroscopically and microscopically. In doubtful cases the biochemical determination of bilirubin can confirm the biliary origin of the effusion.
Cardiogenic effusions (caused by heart failure, “protein-rich transudates”): in view of the protein content (between 25 and 45 g/l), these are classified as modified transudates. However the macroscopic, cellular and physical characteristics of these effusions are very similar to those of pure transudates (transparent, minimally turbid, with a specific gravity that is variable but usually low, and practically cell-free, with only rare activated mesothelial cells and macrophages or, in the cat, lymphocytes).
Neoplastic effusions: these effusions are extremely variable in terms of their macroscopic characteristics, specific gravity, protein content and cellularity (often extremely marked). Effusions are defined as neoplastic when the microscopic examination confirms the presence of neoplastic cells, independently of the other physicochemical characteristics. Although the finding of neoplastic cells is of absolute diagnostic specificity, the diagnostic sensitivity is not high: in other words if neoplastic cells are not found, this does not exclude that the effusion (which is usually classified as a modified transudate) is from a patient with an intracavitary neoplasia not characterized by cell exfoliation. The tumours which are most commonly associated with effusions are:
1) Lymphomas: in multicentric lymphoma or in thymic or intestinal forms (occurring in young FeLV-positive cats and in elderly FeLV-negative cats, respectively), it is possible to find effusions characterized by large lymphoid cells, with characteristics typical of blasts (limited basophilic cytoplasm, diffuse chromatin nuclei, evident nucleoli, anisocytosis, anisokaryosis) (Fig 8). More rarely these effusions contain small lymphocytes and, sometimes, eosinophils (especially in T-cell lymphomas, which produce eosinophilopoietic cytokines).
2) Carcinomas: in the case of primary or metastatic neoplasms (pulmonary, pancreatic, gastrointestinal, ovarian), it is possible to find effusions characterized by groups (at times voluminous or organized in acinary or papillary structures) of polygonal cells, with rounded nuclei, dense chromatin and intensely basophilic cytoplasm, often with marked characteristics of malignancy (anisocytosis/anisokaryosis, pleiomorphism, multinucleation, multiple or bulky nucleoli, frequent mitoses) (Fig. 9). In some cases the tumour itself induces a neutrophilic inflammatory response which may confound the diagnostic interpretation if the malignant features of the epithelial cells are not particularly evident.
3) Mesotheliomas: cytologically these are very similar to carcinomas. The effusion is highly cellular and characterized by groups of polygonal cells with rounded nuclei and basophilic cytoplasm (Fig. 10). Binucleated or multinucleated cells and atypical mitoses are often present. The differential diagnosis of mesothelioma from carcinoma may be based on the finding of narrow, clear spaces between the mesothelial cells (mesothelial slits or “windows”), although these are not always present, or on immunohistochemical assays (e.g. simultaneous positivity for cytokeratins and vimentins). The differentiation between mesothelial reactivity and mesothelioma may also be difficult because reactive cells often have morphological characteristics similar to those of neoplastic cells.
Suggested readings
- Feldman B.F., Zinkl J.G., Jain N.C.: Schalm’s Veterinary Hematology, 5a edizione. Philadelphia: Lippincott Williams & Wilkins, 2000.
- Stockham S.L., Scott M.A.: Fundamentals of Veterinary Clinical Pathology, 2a edizione. Blackwell Publishing, 2008.
- Paltrinieri S., Bertazzolo W., Giordano A. Patologia Clinica del Cane e del gatto. Approccio pratico alla diagnostica di laboratorio. ISBN: 978-88-2143-159-3. Elsevier Masson, 2010.
- Tyler RD, Cowell RL. Evaluation of pleural and peritoneal effusions. Vet Clin North Am Small Anim Pract. 1989 Jul;19(4):743-68
- Fossum TW, Jacobs RM, Birchard SJ. Evaluation of cholesterol and triglyceride concentrations in differentiating chylous and nonchylous pleural effusions in dogs and cats. J Am Vet Med Assoc. 1986, 188(1):49-51
- Hirschberger J, DeNicola DB, Hermanns W, KraftW. Sensitivity and specificity of cytologic evaluation in the diagnosis of neoplasia in body fluids from dogs and cats. Vet Clin Pathol. 1999;28:142–146.