Cytology of effusions (Proceedings)

An effusion is defined as the escape of fluid into a part. Effusive processes can occur in essentially any tissue in the body. Obtaining this fluid for cytologic evaluation limits the types of samples that are amenable to examination. Synovial fluid, aqueous humor, cerebrospinal fluid and serous body cavity fluids are all routinely evaluated in veterinary medicine. This review will only include pleural, peritoneal and pericardial fluids. As is the case with any information derived from any test, the ultimate interpretation and application requires full knowledge of the case and findings from other sources such as diagnostic imaging.

The peritoneal cavity is covered by a serous membrane of mesothelial cells overlying a connective tissue stroma. The cavity is essentially a potential space, with a small amount (<1 ml/kg) of fluid present to lubricate the viscera and allow their unhindered motions. The equine abdomen is real and fluid can be obtained from healthy animals. The esophageal hiatus, vena caval hiatus and the aortic hiatus allow passage of the corresponding structures through the diaphragm. These coverings are covered by a thin layer of fused endothoracic and transverse fascia. The peritoneal lining is extremely absorbent. Stoma are present that lead to the lymphatic system. Four fifths of the lymph from the peritoneum drains into the sternal lymph nodes, primarily through the diaphragmatic peritoneum. Mesenteric lymph nodes drain the intestine, not the peritoneum, which is drained by the omental lymph nodes. The parietal layer covers the inner walls of the abdominal, pelvic and scrotal cavities, while the visceral portion envelops the abdominal viscera. The pressure within the capillaries of the visceral peritoneum is higher than elsewhere, making the peritoneal cavity particularly prone to developing accumulations.

The entire pleural cavity is lined by a serous membrane known as the pleura. In health this is a potential space that contains only enough fluid to facilitate movement of the organs (e.g. excursion of the lungs during inspiration and expiration) and transmission of forces. It is composed of a single layer of mesothelial cells supported by a delicate network of connective tissue rich in elastic fibers. The visceral pleura covers, and is adherent to, the lungs. Its capillary network is derived from the pulmonary circulation. The parietal pleura covers the remaining thoracic cavity and is sub-divided into the costal, diaphragmatic and mediastinal portions, depending on the part of the cavity involved. Its capillary network is derived from the systemic circulation; however, unlike the visceral pleura, there is a rich lymphatic network for drainage of the pleural space. The mediastinum is a cleft or wall between the right and left mediastinal pleura. It is complete in bovid, goats, pigs or in any young animal. A fenestrated mediastinum can be found in the dog, cat, horse and in sheep; however, with certain diseases (e.g. exudate with fibrin formation), these fenestrae can become occluded, effectively forming a complete mediastinum.

The pericardium is a fibroserous envelope around the heart. It is made up of an outer fibrous and inner serous part, which is composed of parietal and visceral layers. Approximately 0.3 to 1.0 milliliters of fluid resides in the space. The function of the pericardium is speculative; however, whatever the function is, it is not vital, as removal produces no untoward effects. Effusions that develop quickly here have the potential for serious and acutely deleterious clinical affects, as cardiac tamponade can rapidly be fatal.

Normal serous cavity fluid is essentially a filtrate of plasma. Normal effusions are due to a combination of factors including Starlings forces, pressure differential across the visceral pleura that favors absorption, lymphatic drainage, mesothelial cell activity and stoma between mesothelial cells. The different cavities can have a large, ongoing exchange of fluid and solute, yet a low, stable volume. Crudely, effusions form when there is an imbalance between formation and resorption. Forces that can impact the formation-resorption balance include:

Hypoproteinemia, increased hydrostatic pressure, increased vascular permeability and decreased lymphatic drainage. There are also multiple sources outside of Starlings forces including: ruptured urinary bladder, ruptured viscous, fractured liver, neoplastic cells producing fluid, iatrogenic, etc.

The power of the clincopathologic evaluation of effusions is undeniable. Using widely available tools and instruments, samples can be obtained with minimal risk to the patients, yet the diagnostic yield can be tremendous, depending on the specific disease entity present. Abdominocentesis is indicated when ascites is grossly evident or when fluid is identified via diagnostic imaging. A non-productive, "dry", tap does not rule out intra-abdominal pathology, as fluids and processes can compartmentalize and/or not produce a large quantity of fluid. A Diagnostic peritoneal lavage (DPL) can be performed when a sample cannot be obtained from the abdominal cavity using a straight abdominocentesis. Thoracocentesis typically will be performed when auscultation or radiographs indicate the location and extent of an effusion. Pericardiocentesis can be a diagnostic and therapeutic procedure that is performed when clinical signs, electrocardiographic, radiographic or echocardiographic findings indicate the presence of an effusion.

The proper handling of fluid can dramatically affect the diagnostic yield from samples properly obtained. If in doubt, contact your reference lab for details concerning submissions. Ethylenediaminetetraacetic acid (EDTA) is the ideal additive for cytologic analysis, as it provides the best morphology, with minimal distortion if smears are made relatively quickly. This tube can also be used for determination of the total nucleated cell count (TNCC) / white blood cell (WBC) count and total protein (TP), assessed by refractometry. If cells are left in EDTA tubes, they will age naturally and artifactually become vacuolated. EDTA is bacteriostatic and therefore not appropriate for culture. Tubes without additives can be utilized if any biochemical assays are to be performed on the sample. Although not ideal, they can be used for culture, mainly aerobic. Serum separator tubes (SST) should be avoided, as the gel can entrap cells and compromise cytologic evaluation.

If a delay is anticipated before processing, several air-dried direct smears (high cellularity) or line / concentrated smears (low cellularity) should be made, with the remainder being refrigerated. These smears will serve as a reference point for the cytopathologist to compare to the aged sample. Labeling can be very important, as is shipping overnight on ice and sending slides in slide transporters.

Long delays allow the macrophages to become vacuolated and begin to phagocytize erythrocytes, confusing the differentiation between in vitro vs. in vivo erythrophagia. In addition, hypersegmentation and pyknosis all begin to occur in neutrophils and other nucleated cells due to aging, respectively. In addition, any bacteria (pathologic or contaminant) can overgrow, making interpretation difficult.

Sample preparation is just as important as the microscopic evaluation. Direct smears can be made exactly like a peripheral blood smear technique (Figure 1). This is n extremely simple, fast technique to visualize nucleated cells in an effusion that has moderately high to high TNCC (>10,000/(l). This is the preparation that should be consulted if there are concerns about true, in vivo morphologic changes to cells and/or structures in the background (e.g. plant material, alleged ingesta, etc.). A drop of the fluid is placed on one end of a clean, new slide. The spreader slide is backed up into the drop, which then spreads towards the edges, but not to the edge. The spreader slide is advanced off the slide in one firm, smooth motion, producing a bullet shaped smear

Figure 1- Line diagram of a smear made from an effusion

Figure 1

The line technique is useful for making smears from samples with moderate nucleated cellularity. The technique is identical to the blood smear technique; however, the spreader slide is stopped short of the edge and lifted straight up, producing a line that contains a concentration of nucleated cells. In samples with TNCC's <5,000 / (l, there are two methods which can be employed to evaluate the nucleated cells. The first involves centrifuging a sample (1,000 - 1,500 rpm for 5 minutes), pouring off the supernatant and re-suspending the remaining pellet with 0.5 ml of fluid. A smear is made from this concentrated solution similar to making a blood smear. A second method is to take a portion of fluid and spin it in a micro-hematocrit tube. The tube is broken at the level of the buffy coat, which is then applied to a slide and streaked out as a blood smear. These preparations are useful in hemodilute samples. Samples with very low cellularity can be cytocentrifuged or prepared using gravity and a home made device.

In the practice setting, Diff-Quik or equivalent is often used. It is very easy to use, rapid, affordable and widely available. Good cellular and nuclear detail. Care must be taken to maintain the stains for consistent results. Gram's stain can be helpful for categorization of bacteria to aid in empiric selection of antibiotic therapy; however, cellular features are poor. A control slide is an absolute must.

The beginning step of evaluation of an effusion is the gross examination. A clean, lined note card may be useful in assessing color and turbidity of the straight fluid as well as the supernatant after centrifugation. Normal fluid is pale yellow to straw colored, depending on diet. Color of a solution may be due to particulate matter such as: cells, fibrin, ingesta and clots; however, pigments such as hemoglobin, myoglobin and bilirubin can also cause a color change from normal. When the supernatant changes color after centrifugation, it implies that the color was due to something particulate. Different colors can suggest different things. Turbidity of the straight fluid as well as the supernatant -Normal fluid is clear to slightly turbid or opalescent. Odor is a characteristic that is often not reported; however, can offer insight into the composition of the fluid. Anaerobic infections tend to cause a foul odor, enterocentesis samples often have a fermented odor, while samples from a uroabdomen smell like ammonia, especially when warmed.

Chemical analysis is the next part of fluid evaluation. Total protein is one half of the criteria used to classify effusions into broad categories. It can be determined biochemically using the supernatant and a chemistry analyzer or can also easily be estimated using a refractometer. Other analytes can be measure for additional investigation into a fluid. Either the supernatant from a clot tube or a heparin containing tube should be used, with the exception of glucose and lactate. Triglycerides and Cholesterol are useful in characterizing the chylous composition of a fluid. Glucose is easily evaluated and may be decreased (<40 mg/dl) or have a serum-to-fluid glucose difference of >50 mg/dl in cases of septic processes. Samples for measuring glucose must be processed immediately or be placed in a fluoride-oxalate tube. Lactate, creatinine, pH, lactate dehydrogensase (LDH), amylase and lipase are all analytes that can be measured and have varying levels of utility.

The second half of the criteria use to classify effusions is the total nucleated cell count (TNCC), which some people refer to as the white blood cell count (WBC), even though there are many nucleated cells present that are not true leukocytes. The TNCC can be artifactually low if EDTA tube is less than 1/4 filled. The TNCC can be done manually with the Unopette (Becton Dickinson, Rutherford, NJ) system or by using an automated cell counter, as long as the solution isn't clumped. Cell clumping (especially with excess fibrin present), cell fragmentation and debris can be mis-interpreted as intact cells using both systems. Red Blood Cell count (RBC) or pack cell volume (PCV) can be easily performed using either a hematology analyzer or microhematocrit tubes, respectively.

Once the TP and TNCC are determined, the microscopic examination commences. RBCs are frequently present. Caution is strongly encouraged when interpreting RBC morphology; however, hemoparasites can rarely be detected. If there is suspicion of a specific poikilocyte, venipuncture and blood smear examination should be pursued. Mesothelial cells and the cells that line all serous cavities and are invariably present in effusions. They can be seen as singlets as well as in small to large cohesive clusters. These cells are generally grouped with macrophages into a group referred to as "Large mononuclear cells". When non-reactive, they appear as small to medium sized round cells. They generally contain a single, centrally place nucleus set in a moderate amount of light lavender cytoplasm, which occasionally blebs at the margin. The nuclei exhibit a finely reticular chromatin pattern and infrequently contain a few small nucleoli. These cells can have a bright pink, ruffled cell margin called the corona or glycocalyx which is an artifact of slow drying and is of no diagnostic significance other than identification. Mesothelial cells easily become reactive due to many things, primarily inflammation. When reactive, these cells become markedly atypical. The cytoplasm will turn dark blue to dark purple and can vacuolated. Polykaryosis is common as are marked anisocytosis and anisokaryosis (between cells as well as within multinucleated cells). The nuclei will exhibit a coarse chromatin pattern and frequently will contain one to several, variably sized, markedly pleomorphic, prominent nucleoli. Reactive mesothelium will become phagocytic and thus cells and debris can be seen within them. Differentiation between reactive / hyperplastic mesothelial cells and cells associated with malignant neoplasia (i.e. mesothelioma, carcinomatosis, etc.) presents a major diagnostic challenge that even the most seasoned cytopathologist cannot unravel. Macrophages are also frequently seen. When non-activated, they are small to large in size and generally contain a single round to somewhat pleomorphic nucleus set in a moderate amount of blue, smooth to lightly vacuolated cytoplasm. Occasional bi-nucleate and tri-nucleate cells can be found. The nuclei exhibit a lacy, fine chromatin pattern and overall look very similar to non-reactive mesothelial cells. Once activated, they have more abundant, more basophilic cytoplasm than their non-reactive counterparts. They also have ruffled cytoplasmic margins, prominent cytoplasmic vacuoles and/or inclusions. Lymphocytes are occasionally seen in effusions. Their percentage may increase in response to chronic inflammation, parasitism and in chylous effusions. They should look exactly as they do in peripheral blood, with younger cells being slightly larger. Reactive lymphocytes can also be seen. They are often larger than small lymphocytes. These cells have more cytoplasm than their non-reactive counterparts, which is dark blue in color and may have a small, perinuclear clearing. Neutrophils are frequently present in samples as a true in vivo entity (i.e. inflammation) as well as secondary to blood contamination. If present solely due to contamination, their morphology should be identical to that in the peripheral blood and should be present in the same proportion compared to RBCs (roughly 400:1). Further discussion can be found in the hemorrhagic effusion section. Every time neutrophils are assessed, their level of degeneration must be evaluated. Assessment of degeneration should be made in a monolayer portion of a direct smear to avoid the changes associated with mechanical stress.

Non-degenerate neutrophils look essentially just like they did in circulation. Non-degenerate neutrophils suggest an environment that is not particularly harsh; however, occasionally bacteria with weak toxins (e.g. Actinomyces spp.) can be present with non-degenerate neutrophils. Toxic changes may be observed, which are generally cytoplasmic in nature (e.g. Döhle bodies, blue foamy cytoplasm and toxic granulation). These changes are not assessed in cytologic preparations per se, as their significance is reserved for hematologic evaluation as they occurred during granulopoiesis. Degenerate changes indicate that a cell has been killed or sustained lethal injuries and therefore cannot voluntarily go into apoptosis or remain viable. The damage is due to local factors such as bacterial toxins, bile, mediators associated with pancreatitis, etc. These changes are nuclear and result from changes in permeability of the nuclear membrane, allowing water into the nucleus, which ultimately leads to single cell necrosis. Karyolysis is the swelling of the nucleus, with loosening of chromatin pattern as well as a less purple to magenta colored chromatin, rather than condensed and dark purple of a non-degenerate neutrophil (Romanowsky stains). Nuclear margins also tend to become fuzzy and indistinct. This proceeds to hyalinization of the nucleus and eventual dissolution. Aged neutrophils undergo apoptosis and eventually will be phagocytized by macrophages. Hypersegmentation is when greater than five lobes are present, which is often accompanied by thinning of the intra-nuclear bridges. This change is a precursor to pyknosis. When pyknosis occurs, the intranuclear bridges break down and the lobes condense into one to several hyperchromic structure(s).

The cells themselves will often condense and their cytoplasm will stain more intensely pink than their viable counterparts. Karyorrhexis is when the nucleus fragments into numerous hyperchromic balls.

Effusion classification is generally based upon the TP and TNCC. These broad classifications are used to help focus on a set of differential diagnoses; however, they are not all inclusive and have ample overlap. Specific effusions (e.g. chylous, bilious, hemorrhagic, etc.) may be named as such based on other characteristics such as cellular content, biochemical properties, etc.

A pure transudate / transudate is what is normally present within the serous cavity, if enough is present to obtain and analyze. The TNCC is low, generally <1,000 - 1,500/(l. Equine abdominal effusions have been described as normal with up to 10,000/(l; however, they are generally less than 5,000/(l. The cells present are usually a mixture of large mononuclear cells (mesothelial cells and macrophages), occasional small lymphocytes and rare non-degenerate neutrophils. The TP is generally low, <2.5 g/dl. Differential diagnoses include hypoproteinemia (especially hypoalbuminemia <1.0 g/dl) of a variety of causes, acute leakage of lymph from lymphatics and acute uroperitoneum.

Modified transudate- An increase in TP, without concomitant increase in TNCC OR vice-versa.

TP is usually 2.5 - 7.5 g/dl. The TNCC is often slightly increased, 1,000 - 7,000/(l (non-equine). The proportions of cells present varies greatly, depending on the nature of the effusion. As one might suspect, this class of effusion is generally non-specific in origin. Cytologic evaluation, as well as anicllary tests are usually helpful in determining the cause. Chylous, bilious, hemorrhagic and neoplastic effusions often fall under this category as do diseases associated with increased hydrostatic pressure (e.g. liver disease, right-sided or bi-ventricular heart failure, etc.). Effusions secondary to FIP can also fall under this category. Uroperitoneum can present as a modified transudate, once inflammation has contributed to the effusion.

Exudate- These effusions have increased TP and TNCC. The TP is generally > 3.0 g/dl and the TNCC is >7,000/(l (non-equine). Neutrophils frequently predominant and can be degenerate if an agent (e.g. bacteria, toxin, etc.) is present to kill them. Non-degenerate neutrophils and/or lack of visualization of an agent do not rule out its presence. Other causes include bile peritonitis, chronic uroperitoneum, acute pancreatitis, celiotomy and necrosis.

Septic exudates are commonly seen in cases of colic, penetrating foreign bodies, extension of sepsis from a parenchymal organ, etc. "Sepsis" includes bacteria, fungal, rickettsial, protozoal and parasitic causes. The TP is invariably elevated (often >3.0 g/dl) due to influx of high-protein fluid, increased vascular permeability, introduction of peripheral blood and some local production of proteins. The TNCC can be >200,000/ul. Neutrophils generally predominate and are frequently degenerate; however, weak or small amounts of toxins may not produce degenerate neutrophils. Bacteria, if present, should be found within neutrophils; however, antibiotic use will make visualization very difficult. Macrophages are frequently markedly activated and phagocytic. They rarely contain bacteria. Mesothelial cells are nearly always markedly reactive. Eosinophils may be present in fungal, protozoal and parasitic processes.

Hemorrhagic effusions representing in vivo hemorrhage and can be a component of another process or by itself. Trauma, coagulopathies, necrosis and neoplasia are examples of differential diagnoses. The TP is variable, ranging from 2.5 - > 6.0 g/dl, depending on amount of blood and TP of existing effusion. The macrophages present will engage in erythrophagocytosis / erythrophagia after blood has been extravasated for 6hrs. In about 72hrs, macrophages will variably contain clumps of the blue-grey pigment hemosiderin and/ or a rhombus to amorphous shaped, translucent yellow crystal(s), which represent hematoidin.

Chylous effusions contain chylomicron-rich lymph fluid that is normally found within lymphatics. Effusions are most frequently in the thorax. Obstruction, dilation and/or rupture of lymph components are differential diagnoses. The TP often cannot be determined due to interference by the lipid present. Initially,small, mature lymphocytes predominate; however, as the irritating lymph is allowed to remain in the serous cavity, non-degenerate neutrophils will appear, followed by activated, lipid filled macrophages.

Uroperitoneum is occasionally seen secondary to bladder rupture, a ruptured urachal abscess or compromise in the urinary system. Initially, protein is very low (<0.5 g/dl); however, as inflammation appears due to the irritating urine, TP increases. TNCC will be very low at first, then increasing as inflammation increases. Neutrophils often predominate as the effusion becomes more inflammatory, with signs of aging / apoptosis occasionally seen. The macrophages become more activated as the duration of the effusion increases.

More specific case examples along with images will be covered during the lecture. In addition, biochemical tests, ancillary tests and other fine points will be illustrated.

References

Aumann M, Worth LT, Drobatz KJ. Uroperitoneum in Cats: 26 Cases (1986 - 1995). Journal of the American Animal Hospital Association. 34: 315-324. 1998.

Baker R, Lumsden JH. Pleural and Peritoneal Fluids in Color Atlas of Cytology of the Dog and Cat. Mosby, St. Louis. 2000.

Cowell RL, Tyler RD, Meinkoth JH. Abdominal and Thoracic Fluid in Diagnostic Cytology and Hematology of the Dog And Cat 2nd Ed. Cowell RL, Tyler RD, Meinkoth JH Eds. Mosby, St. Louis. 1999.

Crowe DT, Bjorling DE.Peritoneum and Peritoneal Cavity in Textbook of Small Animal Surgery 2nd Ed. Slatter D ed. W.B. Saunders. Philadelphia. 1993.

D'Urso L.Thoracic and Pericardial Taps and Drainsin Textbook of Veterinary Internal Medicine 6th Ed. Ettinger SJ, Feldman EC eds. Elsevier Saunders, St. Louis. 2005.

DeHeer HL, Parry BW, Grindem CB. Peritoneal Fluid in Diagnostic Cytology and Hematology of the Horse 2nd Ed. Cowell RL, Tyler RD Eds. Mosby, St. Louis. 2002.

DeHeer HL, Parry BW, Grindem CB. Plerual Fluid in Diagnostic Cytology and Hematology of the Horse 2nd Ed. Cowell RL, Tyler RD Eds. Mosby, St. Louis. 2002.

Edwards NJ. The Diagnostic Value of Pericardial Fluid pH Determination. Journal of the American Animal Hospital Association. 32: 63-67. 1996.

Grindem CB, Fairley NM, Uhlinger CA, Crane SA. Peritoneal Fluid Values from Healthy Foals. Equine Veterinary Journal. 22(5): 359-361. 1990.

Hartmann K, Binder C, Hirschberger J, et al. Comparison of Different Tests to Diagnose Feline Infectious Peritonitis. Journal of Veterinary Internal Medicine. 17: 781-790. 2003.

Hirschberger J, DeNicola DB, Hermanns W, Kraft W. Sensitivity and Specificity of Cytologic Evaluation in the Diagnosis of Neoplasia in Body Fluids from Dogs and Cats. Veterinary Clinical Pathology. 28(4): 142-146. 1999.

Kjeldsberg C., Knight J. eds. Body Fluids: Laboratory Examination of Amniotic, Cerebrospinal, Seminal, Serous & Synovial Fluids 3rd Ed. ASCP Press, Chicago. 1993.

Kruth SA. Abdominal Distention, Ascites and Peritonitis in Textbook of Veterinary Internal Medicine 6th Ed. Ettinger SJ, Feldman EC eds. Elsevier Saunders, St. Louis. 2005.

Laforcade AM, Freeman LM, Rozanski EA, Rush JE. Biochemical Analysis of Pericardial Fluid and Whole Blood in Dogs with Pericardial Effusion. Journal of Veterinary Internal Medicine. 19: 833-836. 2005.

Mertens MM, Fossum TW, MacDonald KA. Pleural and Extrapleural Diseases in Textbook of Veterinary Internal Medicine 6th Ed. Ettinger SJ, Feldman EC eds. Elsevier Saunders, St. Louis. 2005.

Orton EC. Pleura and Pleural Space in Textbook of Small Animal Surgery 2nd Ed. Slatter D ed. W.B. Saunders. Philadelphia. 1993.

Rudloff E. Abdominocentesis and Diagnostic Peritoneal Lavage in Textbook of Veterinary Internal Medicine 6th Ed. Ettinger SJ, Feldman EC eds. Elsevier Saunders, St. Louis. 2005.

Shelly, SM. Body Cavity Fluids in Atlas of Canine and Feline Cytology. Raskin RE, Meyer DJ Eds. W.B Saunders, Philadelphia. 2001.

Tobias AH. Pericardial Disorders in Textbook of Veterinary Internal Medicine 6th Ed. Ettinger SJ, Feldman EC eds. Elsevier Saunders, St. Louis. 2005.

Tyler RD, Cowell RL. Evaluation of Pleural and Peritoneal Effusions. Veterinary Clinics of North America: Small Animal Practice. 19(4): 743-768. 1989

Van Hoogmoed L, Rodger LD, Spier SJ, et al. Evaluation of Peritoneal Fluid pH, Glucose Concentration, and Lactate Dehydrogenase Activity for Detection of Septic Peritonitis in Horses. Journal of the American Veterinary Medical Association. 214(7): 1032-1036. 1999.

Van Hoogmoed L, Snyder JR, Christopher M, Vatistas N. Peritoneal Fluid Analysis in Peripartum Mares. Journal of the American Veterinary Medical Association. 209(7) 1280-1282. 1996.

Zocchi L. Physiology and Pathophysiology of Pleural Fluid Turnover. European Respiratory Journal. 20: 1545-1558. 2002.