Septic peritonitis and uroperitoneum secondary to subclinical omphalitis and concurrent necrotizing cystitis in a colt

Abstract

A 15-day-old American Quarter horse colt was presented for depression and pyrexia. Peritonitis was diagnosed following peritoneal fluid analysis. Exploratory laparotomy revealed an area of focal necrosis over the dorsal wall of the urinary bladder leading to peritonitis and uroperitoneum. The affected area of the urinary bladder was resected and the peritonitis resolved with medical treatment.

A 15-day-old, 48-kg, intact male, American Quarter horse foal was referred to the Atlantic Veterinary College Veterinary Teaching Hospital (AVC) with a 12-hour history of depression, lethargy, decreased nursing activity, and fever. Foaling was observed by the owner and reported to have been uneventful. The umbilicus was dipped in dilute chlorhexidine during the first days of life. The foal nursed normally immediately after birth and was reported to be healthy for the first 2 wk of life with no abnormalities in attitude, appetite, defecation, urination, or in the appearance of the umbilicus.

Case description

Upon presentation to the AVC the foal was laterally recumbent and febrile (rectal temperature: 39.3°C). On thoracic auscultation, tachycardia (112 beats/min) and tachypnea (76 breaths/ min) accompanied by an increase respiratory effort were noted. Lung sounds were normal. Capillary refill time was 3 s. The oral mucous membranes and sclerae were injected and mildly icteric. On abdominal auscultation, intestinal borborygmi were absent in all abdominal quadrants. The abdomen was slightly distended and tense on palpation. The foal was estimated to be 8% dehydrated. Differential diagnoses for the clinical signs exhibited by the colt included gastrointestinal obstruction, uroabdomen, primary peritonitis or peritonitis secondary to an umbilical remnant infection, loss of integrity of the gastrointestinal tract, penetrating wounds, or intra-abdominal abscesses.

The initial diagnostic plan included a complete blood (cell) count (CBC), serum biochemical profile, and venous blood gas analysis. The CBC revealed a mild-to-moderate leukopenia [white blood cells, 3.1 × 10⁹/L; reference range (RR): 5.5 to 12.5 × 10⁹/L)], with mild left shift (band neutrophils, 0.76 × 10⁹/L; RR: 0.0 to 0.1 × 10⁹/L) and a plasma fibrinogen concentration within the reference range. These findings were consistent with a moderate acute inflammatory process. Abnormal findings on the biochemistry profile included a mild hyponatremia (sodium, 128 mmol/L; RR: 135 to 148 mmol/L) and hypochloremia (chloride, 92 mmol/L; RR: 98 to 110 mmol/L), which were attributed to decreased intake or increased loss (such as gastrointestinal losses, renal losses, or third space sequestration). The blood urea nitrogen concentration was mildly decreased (urea, 3.2 mmol/L; RR: 3.5 to 7 mmol/L). Additional blood biochemistry abnormalities were mild hyperglycemia (glucose, 8.8 mmol/L; RR: 3.6 to 5.6 mmol/L) attributed to a stress response, mild hypoproteinemia (protein, 52 g/L; RR: 60 to 77 g/L) and hypoalbuminemia (albumin, 19 g/L; RR: 25 to 36 g/L) attributed to the age of the patient and possibly increased loss. Lactate concentration was mildly increased (1.44 mmol/L; RR: 0.55 to 1.11 mmol/L) and venous blood gas analysis did not reveal significant abnormalities. Serum IgG concentration was measured and considered adequate (> 8 g/L). It was determined that the foal had a negative sepsis score of 10 according to the University of Florida system (1). A blood culture was not performed.

The clinical signs combined with an inflammatory leukogram were felt to be most consistent with a presumptive diagnosis of septic peritonitis. Due to the minor decreases in serum sodium and chloride concentration and the absence of azotemia or hyperkalemia, a diagnosis of uroperitoneum was considered unlikely. To confirm the tentative diagnosis transabdominal ultrasonography was performed and peritoneal fluid was collected via abdominocentesis for cytology and culture.

On transabdominal ultrasonography (iU22 Ultrasound System; Philips Healthcare, Andover, Massachusetts, USA) using a 5–8 mHz curvilinear array transducer, there was a mild increase in volume and echogenicity of the peritoneal fluid within the abdominal cavity. The bladder contained a moderate amount of urine and the bladder wall appeared to be intact. Umbilical structures and abdominal viscera were evaluated and no other abnormalities were visualized. The ultrasonographic findings were consistent with mild to moderate peritoneal effusion.

Abdominocentesis yielded a small amount of peritoneal fluid which was turbid and serosanguineous. The nucleated cell count was markedly increased (237.84 × 10⁹/L; normal < 5.0 × 10⁹/L), as was the protein concentration (4.8 g/L; normal < 2.5 g/L). Peritoneal fluid creatinine concentration was not measured. Peritoneal fluid cytology revealed a predominance of nucleated cells and low-to-moderate numbers of erythrocytes on a clear to pale pink background. Based on a 400 cell count of the nucleated cells 87% were neutrophils occurring as a mixture of degenerative and non-degenerative cells, and 13% large mononuclear cells consistent with macrophages or mesothelial cells. In addition, low-to-moderate numbers of cocci to short rod-shaped bacteria were present both extracellularly and within neutrophils. The aforementioned findings supported a diagnosis of septic peritonitis. However, bacterial culture of the peritoneal fluid failed to yield growth.

The foal was hospitalized and an exploratory celiotomy was recommended to determine and possibly correct the primary cause of the septic peritonitis. Initial medical therapy was focused on correction of electrolyte abnormalities and dehydration to stabilize the patient before surgery. This initial therapy included intravenous fluid bolus (3 L of 0.9% saline) over a 2-hour period. The foal was observed to urinate normally on several occasions after the initial intravenous fluid therapy. In addition, the foal received ceftiofur sodium (Excenel; Pfizer Animal Health, Kirkland, Quebec), 5 mg/kg body weight (BW), IV, q12h and flunixin meglumine (Flunazine; Vétoquinol, Lavaltrie, Quebec), 0.5 mg/kg BW, IV, q12h.

Prior to general anesthesia, the foal was premedicated with diazepam (Valium; Sandoz Canada, Boucherville, Quebec), 0.2 mg/kg BW, IV. Anesthesia was induced with ketamine (Vetalar; Bioniche Animal Health, Belleville, Ontario), 0.2 mg/kg BW, IV, and maintained with 1.5% isoflurane (Pr Isoflurane USP; Pharmaceutical Partners of Canada, Richmond Hill, Ontario) in 3 L/min 100% oxygen.

The foal was placed in dorsal recumbency and the ventral aspect of the abdominal wall was aseptically prepared for surgery. A fusiform skin incision was centered over the umbilicus. Sharp dissection around the umbilical stump permitted access to the abdomen. Upon entering the abdomen, copious amounts of dark, yellow-colored fluid were visualized. Suction was used to drain approximately 1/2 L of peritoneal fluid. At this time, a sample of the fluid was submitted for determination of creatinine concentration. The creatinine concentration of the peritoneal fluid (438.5 μmol/L) was more than 2-fold the creatinine concentration of the serum (87.5 μmol/L).

The bladder was exteriorized by gentle traction from the umbilical stump. The bladder appeared moderately distended and its integrity was assessed by applying manual pressure on the bladder. A focal necrotic area (1.5 cm × 2.5 cm) with two small perforations (1 to 2 mm diameter) was visualized on the dorsocranial margin of the bladder. Urine leakage through the small perforations was evident. The remainder of the abdominal viscera and serosal surface of the peritoneum did not appear grossly inflamed.

The umbilical vein and the umbilical arteries were ligated and transected. The bladder was atraumatically grasped with 2 Babcock forceps. Then, two Rochester-Carmalt forceps were clamped across the apex of the bladder just proximal to the margin of the necrotic bladder wall. The umbilical remnants were resected and the necrotic apex of the urinary bladder was transected between the 2 Rochester-carmalt forceps. The resected umbilical remnants and necrotic apex of the bladder were submitted for gross anatomy pathological and histological evaluation. The bladder was closed using a routine 2–layer closure (Parker-Kerr suture pattern, with Lembert oversew) with 2-0 polyglactin suture material. The abdomen was thoroughly lavaged with sterile saline. The foal recovered uneventfully from anesthesia.

Gross anatomic examination of the formalin-fixed resected bladder and associated structures revealed marked thinning of the bladder wall along the dorsal midline with 2 small (1–2 mm) punctate perforations (Figure 1). On the lateral surfaces of the bladder, the wall was thick and reddened. Histological evaluation of sections taken from the area of bladder wall thinning disclosed a loss of the muscularis layer and extensive mucosal necrosis with replacement of normal tissues by a band of connective tissue and neutrophilic infiltrates. Necrosis, hemorrhage, and suppurative infiltrates accompanied by occasional macrophages extended into the adjacent tissues replacing large areas of the mucosa. Well-vascularized fibroblastic tissue with collagen deposition expanded the outer muscularis and the serosal layers. In addition, fibrin, numerous neutrophils, and fewer macrophages formed a thick layer covering the serosal surface. Histologically, similar exudate covered the serosal surface of the umbilical remnants. These pathological findings were interpreted as punctate rupture of the urinary bladder associated with a severe subacute to chronic necrotizing cystitis and localized necrosuppurative peritonitis. In addition, mild to moderate omphalitis was also diagnosed. Bacterial culture of the resected tissue was not performed.

Figure 1.

Photograph of the bladder wall with focal punctate perforation (red ellipse). This is a dorsal view of the formalin-fixed bladder with the apex of the bladder to the left and the surgical margin to the right. The blind ending stump at the top of the picture is the right umbilical artery.

For the first 24 h post surgery, the foal was depressed and recumbent for most of the time. However, the colt was able to rise without assistance, nursed when stimulated, and urinated and defecated normally. Postoperative treatment included administration of 1 L of plasma, IV to restore the serum protein levels. Antimicrobial therapy included ceftiofur sodium, 5 mg/kg BW, IV, q12h, for 10 d, and gentamicin (Gentocin; Schering Corporation, Kirkland, Quebec), 6.6 mg/kg BW, IV, q24h, for 5 d. Anti-inflammatory therapy and postoperative analgesia were provided by flunixin meglumine, 0.5 mg/kg BW, IV, q12h, for 3 d. Omeprazole (GastroGard; Mérial Canada, Baie d’Urfé, Quebec), 4 mg/kg BW, PO, q24h, for 10 d was also administered.

Three days after surgery, a sample of clear and colorless urine was collected (free catch) and submitted for laboratory analysis. Urinalysis (macroscopic and microscopic) results were unremarkable. Over the first post-operative week, blood samples were periodically submitted (approximately every 48 h) for CBC and chemical analysis. All blood laboratory values progressively returned to the normal reference ranges. Ten days after admission, the foal was discharged on sulphamethoxazole and trim-ethoprim (Novo-Trimel DS; Novopharm, Toronto, Ontario), 20 mg/kg BW, PO, q12h for 5 d and a recommendation for 2 wk of stall confinement. The skin staples were removed 2 wk after surgery by the referring veterinarian.

Long-term follow-up contact with the client revealed that the foal was doing well up to 11 mo after discharge. However, at 1 year of age the horse developed traumatic septic arthritis of the right tarsocrural joint with associated osteomyelitis of the right distal tibia. At that point the horse was humanely euthanized. Necropsy confirmed the diagnosis of fibrinopu-rulent arthritis of the right tarsocrural joint and distal tibia suppurative osteomyelitis. In the abdominal cavity there were minor fibrous adhesions between the liver and diaphragm and to a lesser extent between the liver and stomach. Additionally, there were some fibrous adhesions between the serosal surface of the bladder and the surrounding structures in the pelvic cavity.

Discussion

Clinical disease resulting from bacterial colonization of the urinary tract including bacterial cystitis remains an uncommon occurrence in horses. In equine neonates, the umbilicus is a potential avenue of entry for pathogens, and infections of internal umbilical structures have been associated with neonatal septicemia and other septic diseases (1–3). In our case, based on histological evaluation of the umbilical cord stump, it seems highly plausible that the source of necrotizing bacterial cystitis that caused septic peritonitis and uroperitoneum, may have been an extension of chronic inflammation (ascending infection) from a compromised umbilical cord stump. In this case, the infection may well have started shortly after the foal was born.

Uroperitoneum is a long-recognized and well-documented disease that occurs mainly in newborn foals (4). Foals aged less than 1 wk are the most commonly affected, with an incidence of 0.2% to 2.5% (5). The condition may occur in either sex and may be associated with concurrent disease processes (5). The proposed pathogenesis of uroperitoneum includes high intrauterine pressure applied to the distended bladder during parturition, congenital defects, external trauma or strenuous exercise. Other reported causes include local necrotizing cystitis, and abscess of the urachus (4,5). In the present case, focal necrotizing cystitis rather than a discrete tear of the bladder wall was diagnosed. The necrotizing cystitis likely progressed to bladder rupture with associated urine leakage and subsequent establishment of uroperitoneum. This sequence of events has been previously described (4–7).

A tentative diagnosis of uroperitoneum may be made on the basis of signalment (< 1 wk of age), anamnesis, clinical signs (oliguria and abdominal straining), classic serum biochemical electrolyte abnormalities (hyponatremia, hyperkalemia, and hypochloremia) in conjunction with azotemia (4,8) and increased serum creatinine concentration. However, serum chemistry abnormalities characteristic of uroperitoneum have been reported to be absent in approximately 50% of foals with uroabdomen (5). This has been attributed to early diagnosis of uroperitoneum with the aid of ultrasonography prior to the equilibration of electrolytes between the peritoneal cavity and peripheral blood, which may have been the case with our foal (5). Septic foals with uroperitoneum, and receiving intravenous fluid therapy, may also exhibit normal serum electrolyte concentrations (4). Interestingly, the foal in this report was 2 wk old, was febrile and neither oliguria nor straining was present on admission. In addition, serum biochemical alterations were not supportive of uroperitoneum (8,9). A diagnosis of uroabdomen is also supported by a peritoneal-to-serum creatinine ratio of > 2:1. In the present case, although calculation of the peritoneal fluid to serum creatinine ratio during surgery confirmed a diagnosis of uroperitoneum, calculation of this ratio preoperatively may have provided a diagnosis assuming urine was leaking into the peritoneal cavity at that time despite not being reflected in alterations in serum biochemistry values.

Contrast cystography is a reliable diagnostic method for ruptured bladder in the dog, but it has been of limited usefulness in foals with uroperitoneum and was not performed in the current case (6,10,11). Transabdominal ultrasonographic findings of large amounts of free hypoechoic fluid within the abdominal cavity, a collapsed bladder in the caudal part of the abdominal cavity, or visualization of the defect in the urinary bladder support a diagnosis of uroperitoneum (5,12,13). In the present case, abdominal ultrasonography did not support a diagnosis of uroperitoneum, possibly due to the small size of the bladder wall defect and dehydration of the foal at presentation. In retrospect, a second ultrasonographic evaluation following IV fluid therapy may have been helpful in reaching a diagnosis of uroperitoneum.

In the present case, the definitive diagnosis of uroperitoneum was not obtained by any method prior to surgery and only exploratory celiotomy permitted confirmation of the diagnosis. The difficulty experienced in detection of uroperitoneum was most likely due to the minuscule disruption of the bladder wall integrity, as previously reported (14). Histopathology of surgical biopsy samples of the bladder from foals with bladder rupture have demonstrated ischemia and necrosis with associated infection only in foals with a positive septic score of 12 or greater (5). In contrast, the case reported here had necrotizing cystitis and a negative septic score of 10 (15).

The prognosis for uroperitoneum in foals is favorable unless there are concurrent medical conditions such as neonatal septicemia and other septic diseases (5). However, the case presented in this report had concurrent bacterial peritonitis but responded well to therapy.

Many causes of peritonitis in the horse have been described, including systemic abscessation, parasitism, gastrointestinal perforation, neoplasia, penetrating trauma of the abdominal wall, intra-abdominal surgical contamination and damage to the urogenital tract as well as urinary bladder rupture associated with bacterial cystitis (7). Peritonitis is an uncommon complication in foals with uroabdomen (7). In the present case, the authors speculate that the uroabdomen and septic peritonitis in this colt may have arisen from a subclinical omphalitis and concurrent necrotizing cystitis with resultant leakage of bacteria and urine into the peritoneal cavity. To the authors’ knowledge the exact sequence of events leading to bacterial peritonitis in foals with uroperitoneum has not been reported.

Regardless of the cause, the pathophysiology of peritonitis is similar in each case. Bacterial contamination from a single point source can spread throughout the entire peritoneal cavity within 3 to 6 h by means of the normal contraction of the intestines, the abdominal wall, and the diaphragm (11). In comparison to other species horses appear to be at increased risk of developing diffuse peritonitis after bacterial contamination of the peritoneal cavity. This has been tentatively attributed to the small size of the equine omentum which decreases the ability to contain peritoneal contamination effectively (16).

Peritonitis may be suspected based on history and supportive clinical signs such as fever, abdominal pain, and ileus (16). A definitive diagnosis requires demonstration of an increased white blood cell count (> 10 000 cells/μL) and protein content (> 20 g/L) in peritoneal fluid, accompanied by cytological evidence of suppurative inflammation with presence of degenerative neutrophils and intracellular bacteria. Culture of peritoneal fluid samples may grow one or more bacterial isolates but has a low sensitivity, yielding growth in only 9.5% to 32.5% of samples (11,16,17). In the present case a diagnosis of septic peritonitis was made based on abdominal fluid analysis and cytology findings despite negative bacterial culture results. Culture of the resected tissues may have yielded bacteria, but was not performed. Similarly, blood culture was not performed. In retrospect, this was an omission in this 2-week-old foal with fever and leukopenia.

The prognosis for peritonitis depends on the ability to diagnose and treat the underlying cause and to prevent development of complications. Reported mortality rates for peritonitis range from 30% to 67% (17,18). Peritonitis associated with perforation of the intestinal tract in the horse is usually fatal (17). Peritonitis not associated with intestinal tract perforation had a lower mortality rate (42.9%) (18). However, peritonitis associated with urinary tract infection has been associated with poor outcomes (19). In addition, long-term complications like adhesions or internal abscesses may further reduce the survival rate (11). In the current case, septic peritonitis was attributed to a focal necrotizing cystitis which extended through all layers of the bladder wall. Based on the clinical presentation, laboratory data, and surgical findings of lack of significant gross abnormalities of the abdominal viscera and peritoneum, other than the lesion at the apex of the bladder, we considered the peritonitis to be an acute process. The acute nature of the peritonitis, in conjunction with institution of aggressive medical and surgical therapy, may have contributed to the favorable outcome. This horse was euthanized a year after admission for unrelated reasons and the necropsy finding of minor intra-abdominal adhesions, could have been a potential source of long-term complications such as intestinal incarceration. CVJ