Management of Peritonitis
David Holt, BVSc, Diplomate ACVS
The peritoneum is a serous membrane lining the abdominal cavity and reflecting around the abdominal organs. It is composed of a single layer of squamous cells of mesothelial origin, and an underlying connective tissue stroma. In the normal animal, a small amount of fluid separates the parietal and visceral peritoneal layers and decreases friction between the abdominal contents. Fluid (or contamination) disperses rapidly throughout the peritoneal cavity. Fluid in the peritoneal spaces drains by diaphragmatic lymphatics to sternal and mediastinal lymph nodes, and the thoracic duct.
Peritonitis is defined as any inflammatory process involving the peritoneum. Primary peritonitis occurs most commonly in cats with Coronavirus infection. In most cases, peritonitis occurs as a sequela to some other disease process. Secondary peritonitis can be aseptic or septic. Aseptic peritonitis may be secondary to foreign bodies (surgical sponges), ruptured neoplasms, or chemical agents such as pancreatic enzymes, bile (which can contain bacteria), urine, and stomach or proximal duodenal contents, in which bacterial concentrations are low.
Septic peritonitis results from bowel perforation distal to the duodenum, penetrating wounds, surgical contamination, or extension of a urogenital infection (ruptured pyometra or prostatic abscess). The peritoneum is exposed to large numbers of usually gram negative organisms as well as chemical bowel contents. Endotoxin is liberated and produced as bacteria grow in the peritoneal exudate.
Chemical injury results in inflammation of the peritoneum. Vasodilation and increased vascular permeability initially results in the loss of isotonic fluid. As vascular permeability increases, albumin is lost into the peritoneal space. Given the large peritoneal surface area, fluid and protein loss can be massive. White blood cells, fibronectin, and fibrin also enter the peritoneal space. Diaphragmatic lymphatics which normally return peritoneal fluid to the systemic circulation become overloaded and plugged with fibrin. Concurrent vomiting and diarrhea exacerbate fluid loss. Fluid loss decreases circulating blood volume and results in a decreased cardiac output and poor tissue perfusion. Poor tissue perfusion results in cellular hypoxia and anaerobic cellular metabolism. Cellular energy depletion causes loss of cell membranae integrity, cell death, and eventually organ failure.
Different etiological agents cause some variation in the pathophysiology of chemical peritonitis. For example, uroperitoneum rapidly causes a life-threatening hyperkalemia. Bile, although usually sterile, can cause permeability changes in the intestinal wall, allowing transmural bacterial migration. Gastric and pancreatic secretions are more irritating than bile, and produce a more rapid and severe peritonitis.
In septic peritonitis, bacteria are initially rapidly opsonized by white blood cells or absorbed by diaphragmatic lymphatics. Hemoglobin and mucus enhance the virulence of intraperitoneal organisms. Bacterial synergism occurs, meaning that the virulence of total bacterial load is greater than the sum of the individual organisms. Bacterial destruction liberates endotoxins, exotoxins and proteases. Endotoxin and cell membrane damage can both activate the arachidonic acid pathways, generating prostaglandins and leucotrienes.
The complement, clotting, and fibrinolytic systems are also activated. Macrophages are stimulated to release tumor necrosis factor, stimulating the release of other inflammatory cytokines. Ongoing absorption of bacteria and toxins, and generation of inflammatory mediators results in "sepsis" or the "systemic inflammatory response syndrome".
Systemically, the animal responds to these profound changes by trying to maintain perfusion to the heart and brain. Hypotension stimulates the carotid baroreceptors; subsequent inhibition of vagal tone and sympathetic stimulation increase heart rate and cause peripheral vasoconstriction. Vasoconstriction is augmented by angiotensin II, produced after hypotension-induced stimulation of the renin/angiotensin/aldosterone system. Angiotensin II stimulates aldosterone release from the adrenal cortex, resulting in sodium and water retention by the kidneys.
Peritonitis is often difficult to diagnose. The clinical signs are largely nonspecific. Depression and diffuse abdominal pain of a degree greater than that usually seen following abdominal surgery or trauma are often present. Most animals splint their abdominal wall at the slightest touch. Vomiting is also a prominent sign of peritonitis. Peritoneal inflammation often causes a paralytic ileus and intestinal dilatation, in addition to the massive effusion. In septic peritonitis, fever and leukocytosis are not consistent findings.
These clinical signs are not pathognomonic for peritonitis. Animals with peritonitis may have a leucocytosis with a left shift, or a neutropenia. In uroperitoneum, elevations in BUN, serum creatinine, and potassium are detected. SAP, SGPT, and total bilirubin levels are elevated in cases of bile peritonitis. Abdominal radiograph may show free gas or a lack of intestinal detail and a ground glass appearance from free fluid in the abdominal cavity. Recovery and examination of peritoneal exudate is extremely valuable in the diagnosis of peritonitis. A four quadrant tap is performed; if this does not yield fluid, peritoneal lavage is performed. The bladder is expressed and the ventral abdomen is prepared aseptically. An over the needle catheter or dialysis catheter is introduced into the abdomen 2 cm caudal to the umbilicus, and 20 ml/kg of warm lactated Ringer's solution is rapidly infused through the dialysis catheter. The fluid is distributed by abdominal massage and collected then collected. In normal animals, lavage fluid leucocyte counts before surgery are usually less than 1000 cells/mm. After uncomplicated intra-abdominal surgery, lavage fluid neutrophil numbers generally increase to 10,000 cells/mm or less. Toxic, degenerative neutrophils with intracellular bacteria indicate septic peritonitis. Creatinine and bilirubin levels may also be measured on the peritoneal exudate. The use of peritoneal fluid glucose levels and acid/base measurements to diagnose peritonitis is currently under investigation.
Initial Medical Treatment
Aggressive patient stabilization is required prior to anesthesia and surgery. Intravenous fluids are administered at shock doses (as high as 90ml/kg/hour in dogs and 45ml/kg/hour in cats). Capillary refill time, heart rate, arterial blood pressure, urine output, and central venous pressure are monitored to assess the response to therapy. The use of plasma or synthetic colloids (Hetastarch, dextran 70) is often required because of the massive loss of albumin into the peritoneal cavity. The choice of fluid type and electrolyte supplementation for subsequent treatment is based on the results of sequential blood gas and serum electrolyte measurements.
Broad spectrum, bacteriocidal antibiotics are administered intravenously as soon as the diagnosis of peritonitis is made. Antibiotics effective against gram positive and negative, aerobic and anaerobic bacteria are recommended. A combination of a penicillin or cephalosporin with an aminoglycocide antibiotic is recommended in animals without pre-existing renal disease. Cefazolin (Ancef, 20mg/kg QID) should be used in preference to cephalothin (Keflin) which does not reach adequate tissue levels in dogs. Gentamicin should be administered once daily at 6mg/kg, as the single high dose is more effective and less nephrotoxic. Metronidazole (20mg/kg BID) may be added for additional anaerobic coverage. Penicillins, cephalosporins, and aminoglycocides all reach intraperitoneal levels equivalent to their serum levels.
Corticosteroid administration in septic shock is controversial. Doses recommended are 15-30 mg/kg for methylprednisolone sodium succinate and 4-6 mg/kg for dexamethasone. Theoretically, corticosteroids increase myocardial function, block formation of several pro-inflammatory mediators, stabilize lysosomal membranes, and prevent complement activation. However, in some human clinical studies, corticosteroid administration either failed to improve or worsened clinical outcome. In experimental studies demonstrating benefit, steroid administration was often before or immediately after bacterial challenge, a situation rarely encountered in clinical practice. Once all of the proinflammatory cascades of sepsis are activated, steroids are likely to have limited beneficial effects.
Non-steroidal anti-inflammatory drugs (NSAIDs) are also a controversial in the treatment of septic shock. Again, in experimental septic shock models demonstrating therapeutic benefit, the NSAID (aspirin, indomethacin, phenylbutazone, flunixin meglumine) was administered prior to the onset of shock. Potential side effects of NSAIDs include gastrointestinal hemorrhage (especially if administered in conjunction with corticosteroids), nephrotoxicosis, and blood dyscrasias. Potential benefits include improved cardiac index, blood pressure, decreased microvascular damage and permeability, and improved survival. The suggested dose of ibuprofin is 1 mg/kg IV; the suggested dose of flunixin is 1-2 mg/kg IV. NSAIDs should not be administered to cats.
One of the most important aspects of treating peritonitis is prompt removal of the inciting cause. Whilst the animal should be stabilized before anesthesia and surgery, the underlying source of the peritonitis must be addressed to resolve the peritonitis. Exploratory laparotomy is mandatory to treat the source of the peritonitis, remove peritoneal contamination and exudate, and provide a source for enteral nutrition. A large ventral midline incision is used for exposure. A complete exploratory laparotomy is performed. The source of the peritonitis is identified and isolated from the remained of the abdomen using moistened laparotomy sponges. In animals with generalized peritoneal contamination, the author prefers to lavage the peritoneal cavity with a large volume of warm sterile saline before proceeding with definitive treatment. The fluid is immediately aspirated from the peritoneal cavity.
Definitive treatment often involves resection and anastomosis of damaged bowel. Omental wrapping or serosal patching are recommended to re-inforce anastomoses in the face of peritonitis. Serosal patching is a technique in which loops of healthy bowel are loosely sutured to the bowel adjacent to the anastomosis. The serosal surfaces of the healthy bowel are then in contact with the anastomosis site allowing a reinforcing fibrin seal to form.
Few animals with peritonitis will eat voluntarily during the postoperative period, so mechanisms for nutritional support should be considered during surgery. Placement of a gastrostomy or jejunostomy tube should be considered unless it interferes with the repair of leaking intestine. Enteral nutrition also improves enterocyte function, which may help minimize bacterial translocation from the intestines.
The peritoneal cavity should be thoroughly lavaged with a large volume of warm, sterile, balanced electrolyte solution to remove bacteria and debris. The volume of fluid required varies from 500 ml in a cat to several liters in a large dog. All lavage fluid must be aspirated. Lavage with inadequate aspiration merely spreads bacteria throughout the peritoneal cavity, and sequesters them from phagocytosis.
The addition of antiseptics to the lavage fluid is controversial. Several human studies have concluded that there is no benefit to adding povidone iodine to lavage fluid. Experimental studies have shown that 2ml/kg of povidone iodine (10% solution, 1% available iodine) instilled into the peritoneal cavity of dogs with peritonitis is lethal. Intraperitoneal povidone iodine also decreases the neutrophil percentage and increases bacterial numbers in the peritoneal cavity in rats with experimental peritonitis.
The addition of antibiotics to peritoneal lavage fluid is also debated. Most studies indicate that this treatment is not beneficial in patients receiving the appropriate antibiotics parenterally. However, a recent human study indicated that the addition of tetracycline to the lavage solution completely inhibited bacterial growth in the residual peritoneal fluid.
Contamination often remains within the peritoneal cavity even after extensive debridement and lavage. The clinician must decide which cases require postoperative peritoneal drainage. Local peritoneal drainage is important in cases in which the inflammation is confined to a specific area of the peritoneal cavity. Drainage tubes with or without suction can be used in such cases; examples include prostatic and pancreatic abscesses. Ideally, drain exit points should be covered with a sterile dressing.
Drains are ineffective for draining the entire peritoneal cavity. They are rapidly sealed by fibrin and omentum. The presence of the drains, which are effectively foreign bodies, increased bacterial translocation and histologic inflammation in an experimental peritonitis model. Drainage of the entire peritoneal cavity is best accomplished by incompletely closing the abdominal incision. The falciform ligament is removed, and the linea is sutured with a continuous monofilament suture leaving a gap of 2 to 4 cm between the wound edges. The wound is bandaged using a sterile dressing of vaseline impregnated gauze covered by sterile towels. This primary dressing is covered with an absorbent layer of combine material, then adhesive tape. The dressing usually requires changing every 6 hours initially, as fluid from the peritoneum soaks through. Bandage changes are performed under sterile conditions with the animal sedated. Fibrous adhesions which may entrap peritoneal exudate are gently freed. There are no exact criteria to help the clinician judge either which cases of peritonitis require open abdominal drainage, or when to close the open incision. In general, the incision is closed when drainage has decreased and the peritoneum appears grossly healthy at bandage changes. Closure should be performed as a complete laparotomy, and the peritoneal cavity examined for any evidence of residual infection.
The amount of new information available on peritonitis and sepsis is vast. Here are a few interesting samples of work which has the potential to impact our treatment of peritonitis in the future:
Experimentally, the use of amide local anesthetics within the peritoneal cavity of rats inhibited chemical peritonitis, and decreased the amount of albumin extravasated from the inflamed area. These effects appear to be due to properties of the anesthetic (lidocaine) itself, rather than dilutional or pH effects.
Oral tetracycline has been shown to protect mice from death in an endotoxin model. The study found that tetracycline inhibited the production of tumor necrosis factor alpha (TNF) and interleukin 1-beta (IL-1) from monocytes, which is normally associated with endotoxin stimulation. Proposed mechanisms for tetracycline's actions include inhibition of protein kinase C and inhibition of metalloproteinases.
Another inhibitor of TNF and IL-1, interleukin 10 (IL-10), improved survival in a mouse peritonitis model when administered 6 hours after the induction of sepsis. This effect was associated with a reduction in circulating TNF. There was no beneficial effect when the IL-10 was administered before or at the time of peritonitis induction.
Inhibition of an entirely different mediator, nitric oxide (NO), improved survival in a mouse model of peritonitis/sepsis. The authors suggest that nitric oxide synthase inhibition may improve the efficacy of conventional treatment of sepsis.
These and many other experiments emphasize the complexity of the body systems and inflammatory cascades involved in peritonitis. Improved treatment of peritonitis involves a more effective understanding and treatment of both the peritoneal cavity and the systemic inflammatory reaction resulting from peritonitis.