Emphysematous Cystitis and Pneumoperitoneum in a Dog with Escherichia coli Urinary Tract Infection and Hyperadrenocorticism

Introduction

Emphysematous cystitis (EC) is a rare form of complicated urinary tract infection (UTI) reported in both human and veterinary medicine. It is characterized by gas presence intramurally within the bladder wall, the bladder lumen, or both.¹

Pneumoperitoneum occurs when there is free gas present within the peritoneum and is typically reported as a sequelae of hollow organ rupture with only ˜10% of cases in people being due to another reason.² It has been reported to occur secondary to EC in six people, of which two cases were managed nonoperatively.^3,4

This case represents a unique presentation of EC with pneumoperitoneum in a dog with an Escherichia coli UTI and hyperadrenocorticism (HAC) diagnosed with multiple imaging modalities and treated conservatively with an excellent clinical outcome.

Case Report

A 9 yr old male neutered Staffordshire bull terrier weighing 29.2 kg was referred to the veterinary teaching hospital for new clinical problems of vomiting, inappetence, and lethargy. The patient had an 18 mo history of HAC and a 1 yr history of multiple UTIs. These problems were being managed by two nonaffiliated veterinary hospitals. There was poor control of the HAC as evident by unresolved clinical signs (polyuria, polydipsia, and dermatologic abnormalities) and results of repeat adrenocorticotropic hormone (ACTH) stimulation tests. By the owners’ admission, this was in part due to poor compliance with administration of trilostane.

Five UTIs with E coli and Pseudomonas were diagnosed in the preceding 7 mo. They were initially susceptible to all tested antibiotics but showed increasing resistance over time. Glucosuria was noted on multiple urine analyses, with either concurrent normoglycemia or mild hyperglycemia.

EC was first noted ultrasonographically by the referring veterinarian 3 mo before referral and was confirmed with a radiograph. The patient had not undergone any procedures that would introduce air into the bladder by iatrogenic means.

Two days before presentation, the owner noticed pigmenturia. The day before presentation, the patient underwent an abdominal ultrasound with a board-certified radiologist. Findings included EC, pneumoperitoneum, bilaterally enlarged adrenal glands, and evidence of chronic inflammation of the pancreas. There was no peritoneal effusion.

Following the ultrasound, a urine culture was performed on a sample obtained via urinary catheterization. The patient had received no antimicrobials in the 3 wk preceding this sample collection. The trilostane dose was increased from 40 mg q 12 hr to 40 mg in the morning and 50 mg in the evening. This increase in dose was based on the last ACTH stimulation test performed 2 wk before showing a postsynacthen cortisol concentration of 5.1 mcg/dL, above the target range of 0.9 to 2.5 mcg/dL for dogs receiving trilostane therapy.

On presentation to the veterinary teaching hospital, the patient was found to be ˜5% dehydrated, with dry and injected mucous membranes. He was normothermic and mildly tachycardic with a heart rate of 120 beats per minute. He had generalized alopecia, thin, hyperpigmented skin, and a distended abdomen.

The patient was admitted to hospital. IV fluids were started (Hartmann’s solution^a + 40 mEq potassium chloride^b added to 1 L bag at 70 mL/hr), and the patient was administered maropitant^c 1 mg/kg IV, and amoxicillin-clavulanic acid^d 25 mg/kg IV. Complete blood count^e showed mild hemoconcentration with a packed red blood cell volume of 57% (reference interval [RI]: 37–55%), and leukocytosis of 12.70 × 10⁹/L (RI: 4.00–12.00 × 10⁹/L) comprising of a neutrophilia with no left shift (11.30 × 10⁹/L, RI: 4.06–9.36 × 10⁹/L). Serum biochemistry^f analysis found a mild hyperglycemia (131.52 mg/dL, RI: 59.45–115.30 mg/dL), hypokalemia (3.1 mEq/L, RI: 3.3–4.8 mEq/L), hyperlactatemia (3.5 mmol/L, RI: <2.5 mmol/L), hypercholesterolemia (380.85 mg/dL, RI: 54.13–289.99 mg/dL), an elevated alanine aminotransferase (166 U/L, RI: <60 U/L), and elevated alkaline phosphatase (2018 U/L, RI: <110 U/L). Total T4 was within normal limits (2.29 mcg/dL, RI: 0.93–3.88 mcg/dL).

A voided urine sample was red in color. Urine-specific gravity was 1.010. Dipstick^g revealed proteinuria, hematuria, and no glucose. Microscopic exam revealed hematuria (5–10 red cells/high-power field), pyuria (5 white blood cells/high-power field), and abundant rod-shaped bacteria.

Abdominal radiographs found a large volume of free peritoneal gas in the caudal abdomen and the retroperitoneal space. Multifocal stippled gas opacities were observed within the urinary bladder wall and within the bladder lumen (Figure 1). The remaining structures were unremarkable.

View Figure

• Download as Powerpoint
• Download Figure

Abdominal ultrasound found a large volume of gas around the urinary bladder, obstructing the view of most of the urinary bladder wall and contents. There was a moderate volume of peritoneal gas and very small volume of peritoneal effusion. There was a small region of hyperechoic mesentery in the mid to caudal abdomen. The adrenal glands were bilaterally enlarged with the left caudal pole width measuring 1.7 cm and right caudal pole width measuring 1.4 cm. The pancreas was diffusely hyperechoic and normal in size and shape. The remainder of the ultrasound was unremarkable.

A pre- and post-IV iohexol^h contrast computed tomography (CT; 55 mL of iohexol with an injection time of 20 s) with retrograde cystogram (88 mL of iohexol and 132 mL of sterile water for injection injected via urinary catheter) was performed to assess the integrity of the bladder wall. A very small amount of gas was noted in the right renal pelvis. A moderate volume of gas was observed in the peritoneum and retroperitoneum, predominantly in the caudal abdomen and around the bladder (Figure 2). In the retrograde cystogram study, all contrast was retained within the urinary bladder.

View Figure

• Download as Powerpoint
• Download Figure

Based on all the diagnostic findings, a diagnosis of EC, pneumoperitoneum, and pneumoretroperitoneum secondary to UTI was made. Emphysematous pyelonephritis was a consideration. Bilateral adrenomegaly was consistent with the previous confirmed diagnosis of HAC.

The patient was kept in the hospital for observation on IV fluids (rate reduced to 40 mL/hr) and frequent walks for urination to keep the urinary bladder small. The results of the urine culture submitted by the referring veterinarian returned with a heavy growth (>10⁸ colony-forming units per liter) of E coli susceptible to both amoxicillin-clavulanate and trimethoprim-sulfate but resistant to nonpotentiated penicillins, tetracyclines, and fluoroquinolones. The owner elected to switch to trimethoprim-sulfate,ⁱ and the patient was started on 480 mg q 12 hr.

The following day, the patient was discharged with a plan for a minimum of 2 mo antimicrobial therapy and to continue the increased trilostane dosing recommended by the referring veterinarian.

The patient was presented to the teaching hospital for a recheck 5 days later. His owner reported that he had more energy and was drinking less water. Upon physical exam, a grade 3/6 systolic heart murmur was detected that had only been otherwise noted on two previous visits to the referring veterinarian. An echocardiogram was performed showing mild changes consistent with myxomatous mitral valve disease, with no evidence of endocarditis. A urine sample was collected via catheterization, and urine culture grew only a light growth (colony-forming units not quantified) of hemolytic Ecoli.

Almost a month after initial presentation, the patient presented for another recheck. His owner reported improvement in the patient’s demeanor and that his hair was growing back. Repeat urine culture found no bacterial growth, and the urinalysis was negative for glucose. The ultrasound found no evidence of EC. A repeat ACTH stimulation test found a postsynacthen^j cortisol concentration of 3.4 mcg/dL, within the target reference range of 50–200 nmol/L at this laboratory.

Another month later, the patient presented to his referring veterinarian who reported he was clinically well. A repeat urine culture was again negative for bacterial growth.

Discussion

EC is a rare condition in both human and veterinary medicine. In veterinary medicine, fewer than 100 cases have been published since the first reported case in 1926.^5,6 Reporting of human cases has increased over the last decade, and this is attributed to increased awareness of the disease and increased use of diagnostic imaging modalities.¹ Although this is also considered true for veterinary medicine, most dogs who have bacteriuria do not have imaging performed, and therefore, the prevalence of EC in dogs and cats is unknown and likely underreported.⁶ EC with subsequent pneumoperitoneum is very rare, having only been reported in six people, and to date has not been reported in veterinary medicine.^3,4

Of dogs with EC, over 95% are reported to have comorbidities that include diabetes mellitus (DM), chronic UTIs, neurologic disease (neurogenic bladder dysfunction, seizures/epilepsy, spinal disease), immunosuppression/impaired immune system (HAC, long-term glucocorticoid administration, hypoadrenocorticism, chemotherapy), chronic kidney disease, uroliths, Fanconi syndrome, and morphologic abnormalities.^5–10 Of people with EC, almost 70% also have DM, which is by far the most common comorbidity and considered to be a major risk factor for developing EC.¹ DM is also commonindogswithECbut does not dominate to the same degree. Before 2014, a little over half of dogs and cats reported to have EC had concurrent DM.⁷ Since then, other comorbid diseases have been reported in increasing frequency.⁷ In a retrospective study of 27 dogs, 9 had DM (33.3%), and 3 of them also had HAC.⁶ In another retrospective study of 38 animals (36 dogs), 10 (26.3%) cases reported an impaired immune system, with 3 of them (7.9%) having HAC. Of the 38 animals, only 4 (10.5%) had DM.⁵ It is therefore suspected that any condition causing impairment of the immune system and/or host defense mechanisms will predispose a patient to EC.^5–7 In the case reported here, poorly controlled HAC was considered to be the underlying disease predisposing the patient to recurrent UTIs and the subsequent development of EC.

It has been hypothesized that glucosuria predisposes animals to EC via the fermentation of glucose by bacteria within the bladder to form carbon dioxide which then collects as small gas bubbles within the bladder lumen and wall.^10–13 In reports of nondiabetic and nonglucosuric cases, authors have theorized albumin as a breakdown product causing gas production. The colonization of these gas vesicles within the bladder wall and the subsequent coalescing and rupture is implicated in the propagation of bladder inflammation and damage.^10–12 A case was reported in a dog where a gas population was noted in the caudal abdomen on radiograph; however, necropsy revealed that the bubble remained within the external wall of the urinary bladder.¹⁴ The remainder of the bladder wall was found to have dissociation of smooth-muscle fascicles with irregular clear areas, assumed to be gas filled prior to processing, that were noted to be larger near the serosal bubble. The gas appeared to filter through the mucosal and muscularis layers in a transverse direction before creating the serosal bubble observed.¹⁴ The sequelae of the rupture of intramural blebs and tracking of gas into the peritoneum was described recently in a diabetic woman who, like the subject of this report, was proven to have no evidence of bladder rupture with contrast cystogram.³ How gas makes it outside the bladder wall is unknown; however, based on previous case findings, it is suspected this patient also had gas tracking both between muscle fascicles and along fascial planes, allowing for the development of pneumoperitoneum as well as pneumoretroperitoneum.¹⁵

The patient’s intermittent glucosuria is likely caused by uncontrolled HAC, resulting in transient hyperglycemia through increased circulating endogenous glucocorticoids and catecholamines, that was expected to be above the renal threshold at times, thus causing a glucosuria.¹⁶ Fanconi syndrome cannot be ruled out given the reported presence of glucosuria in the absence of hyperglycemia and the overlap of clinical signs of HAC. It was suspected that the absence of glucosuria at times could be due to either the intermittent nature of the hyperglycemia and mismatch in timing of blood and urine testing or perhaps bacteria fermenting the glucose within the urine lead to negative urine dip-stick readings. This case reported infection with Ecoli,the most common isolate, implicated in more than half of all UTIs and EC cases in both humans and animals.^5,17

The diagnosis of EC and pneumoperitoneum can be made with radiography, ultrasonography, or CT. Common use of radiography in humans has resulted in a radiographic classification scheme of three stages of EC. Stage one is characterized by a 1 mm zone surrounding the contrast medium representing a thin layer of gas presence within the bladder wall only; stage two is demonstrated by an irregularly thickened bladder wall with larger areas of intramural gas and no luminal gas; with stage three progressing to intraluminal gas also being present.¹⁸ This case falls into the category of stage three EC, with gas in both the bladder lumen and wall. Given the presence of extensive bladder wall disease with pneumoperitoneum and no evidence of bladder rupture, this case could possibly be considered to represent a further stage of EC classification with disease progression.

CT represents the gold standard as a more sensitive imaging modality. Although 40% of human EC cases undergo CT, in veterinary medicine CT is rarely performed.^5,19 The patient reported here had findings on CT such as gas within the right renal pelvis, along the length of the ureters, and within the retroperitoneum that was not detected by multiple board-certified imaging specialists on radiograph or ultrasound, demonstrating that CT may be required to diagnose the extent of the disease and to determine the appropriate treatment and prognostication.¹⁹

Pneumoperitoneum is most commonly reported secondary to hollow organ perforation or opening of the abdominal wall following surgery or penetrating trauma. Occasionally it can occur because of an extension of gas from the thorax via the mediastinal and retroperitoneal tissue planes, or rarely gas can travel directly from the pleural space across defects in the diaphragm.^2,20 Approximately 10% of people with pneumoperitoneum have a nonsurgical source of free abdominal air.² In 2003, a study of 39 animals with pneumoperitoneum found two cases secondary to bladder rupture but none due to EC and gas tracking outside of the bladder without rupture.²⁰ To the authors’ knowledge, only six cases in humans had been reported of patients with pneumoperitoneum secondary to EC. Of these cases, two have been managed nonoperatively, with both of these managed medically in a hospital setting with IV antibiotics and bladder drainage.^3,4 This case represents the third in any species to be reportedly managed medically but the first that was managed medically on antibiotics with no other aid in bladder drainage than natural voiding of urine.

Conclusion

This case demonstrates a presentation that could be considered progressive beyond previously reported stages of EC and reiterated the superiority of CT in detecting air within parenchyma and peritoneal space. The extensive destruction of bladder-wall layering in this patient would have posed a surgical dilemma should rupture have occurred because the layers typically targeted for suture strength were severely compromised. However, it showed a very favorable response to conservative management.