Peritonitis

INTRODUCTION

Peritonitis is defined as inflammation of the peritoneal cavity and may be classified according to the underlying etiology (primary or secondary), extent (localized or generalized), or the presence of infectious agents (septic or nonseptic). Primary peritonitis refers to a spontaneous inflammatory condition in the absence of underlying intraabdominal pathology. Secondary peritonitis occurs more commonly and is the consequence of a preexisting aseptic or septic pathologic, intraabdominal condition. Secondary septic peritonitis is the more common form in the dog and cat, most commonly resulting from leakage of gastrointestinal (GI) contents from a compromised GI tract. Because of the multitude of conditions that may lead to peritonitis, the types of clinical signs and their severity are varied.

Hematogenous dissemination of infectious agents has been postulated as the mechanism of development of primary peritonitis and is likely facilitated by impaired host immune defenses. The most common form of primary peritonitis is the effusive form of feline infectious peritonitis, caused by feline coronavirus, which should be included on any differential diagnosis list for cats with peritoneal effusion. Other infectious agents reported to have caused primary peritonitis in dogs and cats include Salmonella typhimurium, Chlamydia psittaci, Clostridium limosum, Mesocestoides spp, Blastomyces spp, and Candida spp.

Inflammation of the abdominal cavity in the absence of infectious pathogens (aseptic peritonitis) most commonly occurs in response to exposure of the peritoneum to sterile fluids (i.e., gastric, biliary, or urine), pancreatic enzymes, or foreign material. Aseptic bile and urine cause minimal peritoneal inflammation, and gastric fluid and pancreatic enzyme leakage lead to a more intense peritoneal reaction. Both microscopic and macroscopic foreign material, including surgical glove powder, surgical materials (suture, cotton swabs, surgical sponges), hair, and impaled objects (sticks, plant material, metal) may elicit a granulomatous response. To minimize iatrogenic causes of aseptic peritonitis, it is recommended that the surgeons rinse or wipe surgical gloves with sterile saline or use powder-free gloves, perform a surgical sponge count before opening and closing a celiotomy, and use surgical sponges with radiopaque markers.

More commonly, secondary peritonitis can be identified as a septic process, with the most frequent source of infection being the GI tract. Leakage of GI contents may occur through stomach and intestinal walls that have been compromised by ulceration, foreign body obstruction, neoplasia, trauma, ischemic damage, or dehiscence of a previous surgical incision. Spontaneous gastroduodenal perforation may be associated with nonsteroidal antiinflammatory drug administration but may also be seen with corticosteroid administration, neoplastic and nonneoplastic GI infiltrative disease, gastrinoma, and hepatic disease.^1,2 GI linear foreign bodies in dogs have been reported to lead to the development of peritonitis in 41% of cases, higher than that previously reported for cats.³ Dehiscence occurs in 7% to 16% of postoperative patients requiring intestinal enterotomy or anastomosis, with mortality rates of 75% to 85% in this population. One study identified dogs as being at high risk for leakage following intestinal anastomosis if they had two or more of the following conditions: preoperative peritonitis, intestinal foreign body, and a serum albumin concentration of 2.5 g/dl or less.⁴ Other causes of septic peritonitis can be found in Box 133-1 .

Box 133-1. Differential Diagnoses of Septic Peritonitis.

Primary

GDV, Gastric dilatation-volvulus.

• Feline coronavirus (feline infectious peritonitis)

• Salmonella typhimurium

• Chlamydia psittaci

• Clostridium limosum

• Mesocestoides spp

• Blastomyces spp

• Candidiasis spp

Secondary

• Penetrating abdominal wounds

• Surgical peritoneal contamination

• Peritoneal dialysis

• Gastrointestinal conditions

• Gastric rupture secondary to GDV, neoplasia, perforating ulcer

• Intestinal leakage

• Perforating foreign body, ulcer, or neoplasia

• Bacterial translocation secondary to obstruction (foreign body, neoplasia, intussusception, or bowel incarceration)

• Dehiscence of intestinal surgical wound

• Ischemic intestinal injury

• Hepatobiliary condition

• Liver abscess

• Liver lobe torsion with abscess formation

• Ruptured biliary tract with bacterobilia

• Pancreatitis or pancreatic abscess

• Hemolymphatic conditions

• Splenic abscess

• Splenic torsion with anaerobic bacterial colonization

• Mesenteric lymph node abscess formation

• Urogenital conditions

• Renal abscess

• Septic uroabdomen

• Pyometra (ruptured or with mural bacterial translocation)

• Uterine torsion

• Prostatic abscess formation

CLINICAL SIGNS

Historical information may provide clues regarding the underlying cause of peritonitis. Previous and current maladies and surgical procedures (including neutering), current medications (particularly those which may predispose to GI ulceration), and duration of current clinical signs should be investigated. Owners should be questioned specifically regarding potential for trauma exposure and foreign body ingestion.

Clinical signs of dogs and cats with peritonitis vary in both type and intensity and may reflect the underlying disease process. Peritoneal effusion is a consistent finding but may be difficult to appreciate on physical examination if a small volume of fluid is present, and may even be difficult to detect sonographically in animals exhibiting dehydration. Abdominal pain may be appreciated on palpation, with a small number of dogs exhibiting the “prayer position” in an attempt to relieve abdominal discomfort. In a retrospective study focusing on cats with septic peritonitis, only 62% exhibited pain on palpation of the abdomen.⁵ Most animals with septic peritonitis are systemically ill and exhibit nonspecific clinical signs such as anorexia, vomiting, mental depression, and lethargy. It should be noted that animals with uroperitoneum may continue to urinate with a concurrent leakage into the peritoneal cavity. These patients may arrive in progressive states of hypovolemic and cardiovascular shock, with either injected or pale mucous membranes, prolonged capillary refill time, tachycardia with weak pulses, and with either hyperthermia or hypothermia reflecting poor peripheral perfusion. A significant number of cats (16%) with septic peritonitis exhibited bradycardia⁵ (see Chapter 106, Sepsis).

DIAGNOSTIC TESTS

Patients with suspected or confirmed peritonitis should have routine hematologic, biochemical, and coagulation analyses. A marked neutrophilia with a left shift is the predominant hematologic finding, although a normal or low neutrophil count may be present. It is anticipated that animals recovering without incident from GI surgery may also have a transient inflammatory leukogram; however, the overall peripheral white blood cell counts typically fall within normal limits.⁶ An increasingly left-shifted neutrophilia (or neutropenia) paired with clinical signs of peritonitis may raise the clinician's index of suspicion for postoperative intestinal dehiscence (which typically occurs 3 to 5 days postoperatively).

Furthermore, acid-base and electrolyte abnormalities may be noted. Hyperkalemia may indicate uroperitoneum, particularly if trauma or urinary tract dysfunction has been noted historically. Hypoproteinemia may be a result of the loss of protein within the peritoneal cavity. Patients with a concurrent septic process may be hypoglycemic. Hepatic enzymes, creatinine, and blood urea nitrogen may be elevated, indicating primary dysfunction of these organs or perhaps reflecting states of decreased perfusion or dehydration. The serum of patients with bile peritonitis may be icteric if the total bilirubin is elevated.

Patients with suspected peritonitis should be evaluated for peritoneal effusion. Little or no fluid may be detected initially if patients arrive early in the disease process or before fluid resuscitation if they are dehydrated. Large volumes of effusion may be obtained via blind abdominocentesis or, alternatively, via ultrasonographic guidance. Single paracentesis attempts are successful in only 20% of patients with low volumes of peritoneal effusion (3 ml/kg) and in only 80% with larger volumes (10 ml/kg). Ultrasonographic guidance will facilitate the retrieval of smaller volumes of peritoneal fluid. If single-site sampling is negative for fluid, four-quadrant sampling should be performed.

A diagnostic peritoneal lavage should be performed when peritonitis is suspected despite the absence of detectable effusion or when a minimal volume of effusion makes it difficult to obtain a sample. Diagnostic peritoneal lavage ideally is performed using a peritoneal dialysis catheter but can also be performed using an over-the-needle, large-bore (14 to 16 gauge) catheter. The technique is performed by infusion of 22 ml/kg of a warmed, sterile isotonic saline solution through the catheter inserted in an aseptically prepared site just caudal to the umbilicus and retrieval of a sample for analysis and culture and sensitivity. It is important to remember that the lavage solution will dilute the sample and therefore may alter the analysis. A repeated diagnostic peritoneal lavage may increase accuracy of the technique when results of the first procedure are equivocal (see Chapter 156, Diagnostic Peritoneal Lavage).

Leukocyte counts in peritoneal fluid are normally less than 500 cells/μl. White blood cell counts between 1000 and 2000 cells/μl indicate mild to moderate inflammation, and a higher peritoneal fluid leukocytosis suggests marked peritonitis.^6,7 However, cell counts in peritoneal lavage fluid obtained from postoperative patients undergoing intestinal resection and anastomosis may also show evidence of significant inflammation in the absence of surgical complications. In the patient that has undergone a celiotomy, 7000 to 9000 cells/μl suggests mild to moderate peritonitis. In these patients, intracellular bacteria or increasing inflammation (numbers of neutrophils or morphologic features of toxicity in these cells) observed in serial samples correlated with clinical findings may prove more useful than single leukocyte counts in abdominocentesis samples when deciding whether reoperation is indicated. It is also of note that dogs receiving antibiotics may have no observable bacteria in peritoneal fluid samples, despite having peritoneal contamination.

In addition to the presence of bacteria and a high nucleated cell count, the glucose concentration of abdominal effusion is a useful predictor of bacterial peritonitis in dogs. A concentration difference of more than 20 mg/dl between paired samples for blood and peritoneal fluid glucose is a reliable predictor of a bacterial peritonitis. Additionally, a blood-to-fluid lactate difference less than 2 mmol/L was predictive of septic peritonitis in dogs but has not been as useful in cats.^8,9 Intravenous administration of dextrose or the presence of a hemoperitoneum may decrease the accuracy of this test.

Samples for aerobic and anaerobic cultures should be obtained at the time of initial sampling so that additional samples are not required after confirming the presence of a septic process and initiating antibiotic therapy.

The diagnosis of uroperitoneum in dogs can be made if the peritoneal fluid creatinine or potassium concentration exceeds that of the serum creatinine (>2:1) or potassium concentration (>1.4:1).¹⁰ Similarly, biliary rupture will lead to a bilirubin concentration that is higher in the peritoneal fluid than in the serum. In addition, bile pigment or crystals may be visible on cytologic examination of the peritoneal effusion in animals with bile peritonitis (Color Plate 133-1). These changes may not be seen in patients with bile peritonitis secondary to a ruptured gallbladder mucocele because the gelatinous bile often fails to disperse throughout the abdomen.

Plain radiographs may reveal a focal or generalized loss of detail that is otherwise known as the ground glass appearance. A pneumoperitoneum (Figure 133-1 ) suggests perforation of a hollow viscous organ, penetrating trauma (including recent abdominal surgery) or, less commonly, the presence of gas-producing anaerobic bacteria. Intestinal tract obstruction or bowel plication should be ruled out. Prostatomegaly in male dogs and evidence of uterine distention in female dogs should be noted. Thoracic radiographs should be performed to rule out concurrent illness (infectious, neoplastic, or traumatic). The presence of bicavitary effusions increased the mortality rate of patients 3.3-fold compared with that of patients with peritoneal effusions alone.¹¹ Ultrasonography may be useful for defining the underlying etiology of peritonitis, in addition to its use in localizing and aiding retrieval of peritoneal effusion. In the case of a confirmed uroabdomen, preoperative contrast radiography (excretory urography or cystourethrography) is recommended to localize the site of urine leakage and aid in surgical planning. It should be noted that all patients should be hemodynamically and medically stabilized before diagnostic imaging is carried out.

Figure 133-1.

Lateral abdominal radiograph showing free peritoneal gas and possibly ingesta free within the abdomen. Pneumoperitoneum without a history of recent surgery or open-needle abdominocentesis indicates the need for abdominal exploratory. This cat was diagnosed with a ruptured gastric mass during surgery.

TREATMENT

Medical Stabilization

The goals for animals with septic peritonitis are to identify and address the source of contamination in order to resolve the infection and treat the systemic consequences of such infection (i.e., fluid and electrolyte abnormalities and hypoperfusion). Before surgical intervention, a decision must be made whether additional hemodynamic and medical stabilization is indicated before proceeding, or whether this additional time and continued contamination of the abdominal cavity will result in further clinical decline that outweighs the benefits of further medical treatment.

The goals of medical treatment are to restore normal fluid and electrolyte balance and minimize ongoing contamination. Fluid resuscitation is initiated after obtaining pretherapy blood samples for a minimum database (packed cell volume, total solids, Azostix, dextrose), hematology, serum chemistry, and coagulation evaluation. Urine should be collected, if possible, for analysis with or without culture and sensitivity testing. Shock doses of crystalloids (90 ml/kg in the dog, 50 ml/kg in the cat) or a combination of isotonic crystalloids (20 to 40 ml/kg) and synthetic colloids (hydroxyethyl starch 10 to 20 ml/kg in the dog or 5 to 10 ml/kg in the cat; or 7% to 7.5% hypertonic saline in 6% dextran-70, 3 to 5 ml/kg IV over 5 to 15 minutes) should be administered to effect (see Chapter 65, Shock Fluids and Fluid Challenge). Because significant amounts of protein are lost into the peritoneal cavity, plasma and/or albumin administration may also be warranted. Electrolytes and glucose should be supplemented if indicated (see Chapters 55 and 69, Potassium Disorders and Hypoglycemia, respectively). After appropriate volume resuscitation, vasopressor therapy may be necessary to further alleviate hypotension. A urinary catheter may aid in diversion of infected urine in the case of a ruptured bladder or proximal urethra and allow for the necessary correction of any metabolic derangements (typically hyperkalemia and acidosis) (see Chapters 55 and 59, Potassium Disorders and Acid-Base Disturbances) before surgery.

Broad-spectrum antibiotic therapy should be initiated immediately after confirming the diagnosis of a septic peritonitis (seeChapters 108 and 109, Gram-Positive Infections and Gram-Negative Infections, respectively). Escherichia coli, Clostridium spp, and Enterococcus spp are common isolates. A second-generation cephalosporin such as cefoxitin (30 mg/kg IV q6-8h) may be used as a single agent or combination antibiotic therapy such as ampicillin or cefazolin (22 mg/kg IV q8h) administered concurrently with either enrofloxacin (10 to 20 mg/kg IV q24h [dog], 5 mg/kg q24h [cat]) or an aminoglycoside (amikacin 15 mg/kg IV q24h or gentamicin 6.6 mg/kg IV q24h). In the event that extended anaerobic coverage is necessary, metronidazole (10 mg/kg IV q12h) canbe used. Aminoglycosides should be avoided unless renal insufficiency has been ruled out and the patient is well hydrated. It is advisable to tailor antibiotic therapy to the results of culture and sensitivity testing when they become available.

Surgical Treatment

The goals of surgical treatment for septic peritonitis include resolving the cause of the infection, diminishing the infectious and foreign material load, and promoting patient recovery with enteral feeding, if indicated. A ventral midline celiotomy from xiphoid to pubis allows a thorough exploratory laparotomy to determine the underlying cause. Monofilament suture material is advocated in an animal with a septic process, and surgical gut is avoided because of its shortened half-life in this environment. Placement of nonabsorbable suture material or mesh within the abdominal cavity is not recommended in cases of septic peritonitis because either material may serve as a nidus for infection. If possible, the surgeon should isolate the offending organ from the rest of the abdomen with laparotomy sponges to prevent further contamination during correction of the problem.

Surgical treatment is tailored to the individual case and the underlying cause of the septic peritonitis. If a GI leakage is being treated, adjunctive procedures such as serosal patching or omental wrapping of the repaired site are recommended to reduce the incidence of postoperative intestinal leakage or dehiscence. Although heavily contaminated or necrotic omentum may necessitate partial omentectomy, preservation of as much omentum as possible is advised to promote drainage of the peritoneal cavity. In addition, surgical applications of the omentum relate to its immunogenic, angiogenic, and adhesive properties and include intracapsular prostatic omentalization for prostatic abscess formation,¹² pancreatic abscess omentalization,¹³ omentalization of enterotomy or intestinal resection and anastomosis sites, and around gastrostomy or enterostomy tube sites. Because enteral nutrition directly nourishes enterocytes and decreases bacterial translocation across the intestinal wall, feeding tube placement (gastrostomy or jejunostomy) should be considered during initial surgical exploration.

After addressing the underlying cause to prevent further contamination of the peritoneum, the infectious and foreign material load must be diminished through a combination of debridement and lavage. Localized peritonitis should be treated with lavage of the affected area only, to minimize dissemination of the infection. A thorough lavage of the entire abdominal cavity with body-temperature sterile isotonic fluid is essential if peritonitis is generalized. The addition of antiseptics and antibiotics to lavage fluid is not beneficial and may actually be detrimental by inducing a superimposed chemical peritonitis. Lavage of the abdominal cavity is continued until the retrieved fluid is clear. All lavage fluid should be retrieved because fluid accumulation in the abdominal cavity impairs bacterial opsonization and clearance.¹⁴

If debridement and lavage can resolve gross foreign material or GI spillage and the source of contamination can be controlled, the abdomen should be closed primarily because potential complications are associated with open abdominal drainage and closed suction drains. All patients with abdominal drainage are susceptible to superinfection with nosocomial bacteria and are subject to massive fluid and protein losses.

Open peritoneal drainage is accomplished with a simple continuous pattern of nonabsorbable suture material in the rectus abdominus muscle, loosely enough to allow drainage through a gap of 1 to 6 cm in the body wall (Color Plate 133-2). A preassembled, sterile bandage comprised of a nonadherent contact layer, laparotomy sponges or gauze pads, roll cotton or surgical towels, roll gauze, and an outer water-impermeable layer is placed to absorb fluid and protect the abdominal contents from the environment. Initially this bandage is replaced twice during the first 24 hours and daily thereafter, although the amount of drainage produced by an individual patient may dictate more frequent changes. A sterile-gloved finger may need to be inserted through the incision to break down adhesions and to allow thorough drainage of the peritoneal cavity. Alternatively, patients with severely contaminated tissues may be placed under general anesthesia and the abdomen explored and lavaged daily before reapplying the bandage. The quantity of fluid can be estimated by the difference in weight of the bandage before application and after removal. Abdominal closure typically can be performed 3 to 5 days following the initial surgery. The placement of a urinary catheter and collection system helps to limit contamination of the bandage and underlying exposed tissues.

Alternatively, the abdomen may be closed primarily and drainage accomplished with closed suction (Jackson-Pratt) drains.¹⁵ Closed suction drainage has been advocated for treatment of generalized peritonitis because it has several advantages over open abdominal drainage, including a decreased risk of nosocomial infection, less intensive nursing care and bandaging requirements, decreased risk for evisceration, and the need for only one surgical procedure.¹⁵ Disadvantages are that the drains may induce some fluid production and may become occluded, although in one study active drainage was maintained for up to 8 days with this technique in 30 dogs and 10 cats.¹⁵ Additionally, closed suction drains allow daily quantitative and qualitative assessment of retrieved fluid for assessing the resolution of peritonitis. Typically, one drain placed between the liver and diaphragm is sufficient for small dogs and cats, and two drains are more appropriate for larger dogs (the fenestrated portion of second drain is placed in the caudal abdomen along the ventral body wall). The drain tubes exit the body wall through a paramedian stab incision and are sutured to the abdominal skin with a pursestring and Chinese finger-trap sutures (Color Plate 133-3).

After routine closure of the abdomen, the suction reservoir bulb is attached to the tubing with vacuum applied. A protective abdominal bandage is placed with sterile contact material around the tube-skin interface and is changed daily to allow assessment of this site. Fluid collected within the bulbs is emptied using aseptic technique, and the volume is recorded every 4 to 6 hours, or more frequently if needed. Drains are removed by applying gentle traction at a time when the volume of fluid production has decreased significantly and cytologic analysis suggests resolution of the peritonitis (decreasing numbers of nondegenerative neutrophils in the absence of bacteria). A sterile bandage is again reapplied to cover the drain exit site until the following day.

Postoperative Care

Postoperative care for patients with peritonitis is typically intense because these patients are critically ill and subject to a variety of complications.⁷ Aggressive fluid therapy is anecessity, particularly in patients with continued fluid losses from abdominal drainage. Electrolytes and acid-base status should be assessed routinely during the postoperative period and corrected as needed. Because anemia and hypoproteinemia are common complications in these patients, blood component therapy and synthetic colloidal support are often necessary, with a goal of maintaining a packed cell volume greater than 24%, serum protein over 3.5 g/dl, and colloid osmotic pressure higher than 15 mm Hg.

Proper nutrition will provide a source of protein that is greatly needed in these patients. Failing to meet nutritional demands, either with parenteral or enteral nutrition, may contribute to impaired wound healing and immune defenses. Enteral feeding is preferred over parenteral feeding but may be stymied by the anorectic patient unless GI feeding tubes were placed during the original surgery. If this was not done, nasoesophageal tubes can be placed in patients unable to tolerate more anesthesia or esophagostomy tubes in those that can. Animals with refractory vomiting typically require parenteral nutrition.

Postoperative hypotension may be treated with vasopressor therapy, but only after addressing any underlying hypovolemia (see Chapter 176, Vasoactive Catecholamines). Proper analgesia is required to ensure patient comfort and to diminish the negative cardiovascular effects associated with overactive sympathetic stimulation (see Chapter 164, Analgesia and Constant Rate Infusions). Other complications, including cardiac arrhythmias, disseminated intravascular coagulation, and systemic inflammatory response syndrome can be found in other chapters (see Chapters 11 and 107, Systemic Inflammatory Response Syndrome and Septic Shock, respectively).

PROGNOSIS

The prognosis for animals with peritonitis depends on the underlying etiology and whether or not infection is present. Studies in which patients have benefited from advances in critical care management cite overall survival rates of 50% to 70%.^1,4,16 Cats were reported to have a lower survival rate than dogs in two studies^1,16; however, another study focusing on 51 cats with septic peritonitis found a 70% survival in animals in which treatment was pursued.⁵ Poor prognostic indicators for animals with septic peritonitis have included refractory hypotension, cardiovascular collapse, disseminated intravascular coagulation, and respiratory disease.^15,16 Mortality rates in patients with septic peritonitis secondary to GI leakage have been reported to vary between 30% and 85%.^1,2,4,18,19 Bacterial contamination was highly associated with mortality in animals with bile peritonitis in one study, and only 27% (3 of 11) of animals with septic biliary effusion survived compared with 100% (6 of 6) with aseptic effusions.²⁰ Cats with uroperitoneum have an overall survival rate or 62%,²¹ whereas the survival rate in dogs is slightly lower at 43% to 56.2%.²² Survival rates appear to be similar in patients with septic peritonitis treated with primary closure, open peritoneal drainage, or closed suction drainage.^16,17