The Case: Canine Urinary Obstruction

Initial Presentation

 A 5-year-old neutered male boxer was presented for a 24-hour history of lethargy and panting as well as 12 to 24 hours of straining to urinate. The patient, owned since he was a puppy, is an indoor/outdoor dog with a tendency to dig among backyard plants, and is current on vaccinations. He was neutered one year ago. He has no travel history. The owner reported that in the past the dog has had bouts of dietary indiscretion with concurrent lethargy and panting, usually resolving in 12 to 24 hours. However, this time the owner failed to see any improvement, was concerned, and brought him to the clinic.

Physical Examination

A large firm bladder (~10–15 cm) was palpated in the caudal abdomen; pulses were weak/synchronous. The prepuce was moist with a bloody red tinge. Other findings were relatively unremarkable.

Management

The patient was hospitalized for urinary catheterization/unblocking procedure.

8:30 am

Blood work

• Compensated respiratory alkalosis 

• Azotemia (blood urea nitrogen: 33 mg/dL [range 6–31], creatinine: 3.0 mg/dL [range, 0.5–1.6])

• Lactate: 1.2 mmol/L (range, < 1.5) 

Abdominal radiographs: large distended urinary bladder with evidence of a small radiopaque calculus within the bladder

Treatment 

General anesthesia was administered for urinary catheter placement and possible calculus retropulsion. Prior to anesthetic induction, patient vomited once (foul brown material). A 14 French red rubber catheter was difficult to place due to narrowing of the penile urethra proximal to the os penis. Dark, bloody urine (1.5 L) with a foul odor was obtained, with a sterile sample collected for analysis/culture and sensitivity testing. A 12 French Foley indwelling urinary catheter with a sterile closed collection bag was then placed. Patient was slow to recover from anesthesia. Abdominal ultrasound was elected to determine the cause of urethral narrowing.

5:10 pm

Imaging

Ultrasonography: moderate amount of anechoic fluid within abdomen; urinary bladder wall mildly to moderately thickened with no obvious tears or fluid flow into the abdomen when the bladder was actively distended. The prostatic/penile urethra was evaluated and showed no obvious lesions; full evaluation was limited by presence of urinary catheter. 

Contrast cystogram: mildly distended bladder with irregular luminal wall; no stones, masses, or overt rupture/leakage 

Urethrogram: narrowing of the penile urethra roughly 10–13 cm proximal to urethral orifice, consistent with area of difficulty when passing urinary catheter

Abdominal fluid analysis

• Creatinine: 7.6 mg/dL (range, 0.5–1.6)

• Potassium: 4.1 mEq/L (range, 3.6–5.5)

Blood work

• Lactate: 4.0 mmol/L (range, < 1.5)

• Progressive azotemia (blood urea nitrogen: 45 mg/dL [range, 6–31], creatinine: 4.1 mg/dL [range, 0.5–1.6])

Treatment

• Fluid therapy: lactated Ringer’s solution (7.5 mL/kg/hr)

• Enrofloxacin: 10 mg/kg IV q24h

• Ampicillin/sulbactam: 30 mg/kg IV q8h

9:35 pm 

Clinical decline continued; acute onset of seizure activity, which responded to bolus of diazepam (0.5 mg/kg IV). Blood glucose too low to read; administered dextrose bolus (0.5 gm/kg IV), followed by a 2.5% dextrose CRI in fluids. Hypoglycemia deemed most likely due to sepsis, potentially from leakage or rupture of the bladder.

10:20 pm

Patient became recumbent, minimally responsive, with poor femoral pulses and injected mucous membranes. Began blood pressure and blood gas monitoring. A coagulation profile was indicative of hypocoagulability.  

Treatment 

• Nasal cannula oxygen therapy: 5 L/min

• Continued enrofloxacin and ampicillin/sulbactam

• Metoclopramide: 0.01 mg/kg/hr CRI

• Pantoprazole: 0.7 mg/kg IV q24h

• Fresh frozen plasma: 6 mL/kg IV

10:50 pm

No clinical change

Blood work

• Potassium: 3.95 mEq/L (range, 3.6–5.5)

• Glucose: 50 mg/dL (range, 70–138)

• Lactate: 7.4 mmol/L (range, < 1.5)

• Blood urea nitrogen: 41 mg/dL (range, 6–31)

• Creatinine: 4.3 mg/dL (range, 0.5–1.6)

• pH: 7.221

• Packed cell volume: 61% (range, 37–55)

• Total protein: 4.0 g/dL (range, 5.0–7.4)

Treatment

• Dextrose: bolus (0.5 gm/kg IV), 2.5% dextrose CRI continued

12:15 am

Outcome

Patient experienced cardiopulmonary arrest. Intubation and cardiopulmonary resuscitation (15 minutes) produced no response. Urine culture 12 hours after plating showed abundant bacterial growth. The sample was not submitted for evaluation due to the outcome.

The Specialist’s Opinion

Amanda Cavanagh, DVM, DACVECC

The above case is a challenging one, as there are some findings consistent with uroabdomen and some that do not support this diagnosis. Ultimately, the rapid clinical decline and cardiopulmonary arrest are likely a result of uncorrected septic shock. Although there is no definitive diagnosis of an infectious source underlying septic shock, the urinary tract appears most likely.

Abdominal Fluid Evaluation

Evaluation of abdominal fluid can aid in the diagnosis of both urinary tract rupture, with subsequent uroabdomen as well as septic peritonitis. The fluid can be evaluated for the presence of urine. An abdominal fluid creatinine concentration that is greater than 2 times that of serum creatinine (ratio >2:1) is predictive of uroperitoneum with a sensitivity of 86% and a specificity of 100%.¹ In the retrospective study of dogs with uroperitoneum where the utility of this ratio was first reported, all dogs with uroperitoneium had an abdominal fluid creatinine concentration that was at least 4 times normal serum creatinine concentration.¹ Evaluation of the provided data shows that this dog had an abdominal fluid:serum creatinine ratio of 1.8:1; this elevated ratio is suggestive (yet not diagnostic) of possible uroperitoneum. With a false negative rate of 15%, other parameters should be used in the diagnosis of uroperitoneum.¹ Abdominal fluid potassium to peripheral potassium ratio can be used as a complementary diagnostic tool. Abdominal fluid potassium that is >1.4 times serum creatinine concentration (ratio >1.4:1) is also predictive of uroperitoneum with 100% sensitivity and specificity.¹ Peripheral potassium was not provided for comparison to abdominal fluid potassium. However, the abdominal fluid potassium within the normal range for peripheral potassium makes uroperitoneum less likely. Potassium is not absorbed across the peritoneal membrane as rapidly as renal potassium excretion occurs. Development of peripheral hyperkalemia can be delayed up to 48 hours, while abdominal fluid potassium increases more rapidly with the accumulation of potassium-rich urine. Therefore, a normal serum potassium concentration through the progression of this case cannot rule out uroperitoneum, but one would expect an increased abdominal fluid creatinine concentration if urine was accumulating in the peritoneal cavity.¹

With the rapid decline in this patient’s condition, re-evaluation of the abdominal fluid to rule out septic peritonitis with cytology, abdominal fluid glucose to blood glucose comparison, and abdominal fluid lactate to blood lactate comparison would have helped guide therapy. Abdominal fluid assessment should always include cytology performed with Romanowsky’s stain (Diff-Quik) to evaluate the number, type, and morphology of inflammatory cells as well as the presence of intra- or extracellular bacteria. Gram stain can further differentiate gram-negative and -positive bacteria to assist in empiric antimicrobial selection. Cytology consistent with septic peritonitis will reveal suppurative inflammation with intracellular bacteria. 

Abdominal fluid with a glucose concentration 20 mg/dL lower than blood glucose is reportedly 100% sensitive and specific for diagnosing septic peritonitis in dogs; this test is 86% sensitive and 100% specific for cats.² Similarly, abdominal fluid lactate at least 2 mmol/L higher than blood lactate is 100% sensitive and specific for diagnosing septic peritonitis in dogs; this technique is not accurate for the diagnosis of septic peritonitis in cats.^2,3 Abdominal fluid culture should always be obtained.

Septic Shock?

This patient’s clinical and biochemical changes throughout the evening suggest septic shock, as there is evidence of new organ failure and hypotension that was unresponsive to fluid therapy. Differential diagnoses for the source of infection include urosepsis secondary to pyelonephritis (given his progressive azotemia), lower urinary tract infection with subsequent urinary tract rupture and septic peritonitis (given abdominal fluid changes possibly consistent with uroperitoneum), prostatic abscess (history of possible urethral obstruction and abdominal effusion), and gastrointestinal perforation (history of vomiting). Thoracic radiographs should be obtained to rule out intrathoracic sources of sepsis. A complete blood count to evaluate inflammatory changes, hematocrit, and platelet count is warranted as well. 

Upon diagnosing sepsis, it is essential to determine the source of infection, initiate appropriate antibiotics after obtaining samples for culture, and perform source control to remove infected tissue if present.⁴ Antibiotics should be initiated within 1 hour of diagnosing septic shock and must be effective against all possible pathogens.⁴ The combination of enrofloxacin and ampicillin/sulbactam in this case is appropriate for community-acquired bacterial sepsis, where there is a lower risk of antibiotic resistance. There is evidence that fluconazole (2.5 mg/kg IV or SC q24h) portends a survival benefit in abdominal sepsis due to enhanced bactericidal activity of neutrophils.⁵

Resuscitation Guidance

Changes in clinical and biochemical parameters can be used to guide resuscitation. A septic patient should be fully instrumented with equipment for continuous monitoring, including an electrocardiograph, pulse oximetry, rectal temperature probe, arterial catheter for direct blood pressure monitoring, and possibly a urinary catheter. Close attention must be paid to heart rate, perfusion parameters, mentation changes, urine output, and body temperature. Normalization of blood pressure in septic shock requires vasopressor use, as hypotension develops secondary to inappropriate vasodilation. Vasopressor dose is ideally guided by continuous direct arterial blood pressure readings and titrated to achieve a mean arterial pressure of 65 mmHg.^4,6 This patient’s indwelling urinary catheter is useful for monitoring urine output, which should remain greater than 0.5 mL/kg/hour.⁴ Oxygenation can be monitored via arterial blood gas analysis or by pulse oximetry; if PaO₂ is less than 90 mmHg, oxygen supplementation is warranted to maintain normoxemia. Mechanical ventilation can be considered for those patients with persistent hypoxemia despite oxygen supplementation, hypercapnia, or respiratory fatigue. 

Derangements in glucose, potassium, magnesium, and ionized calcium should also be aggressively corrected and monitored frequently. Guidelines also recommend transfusion of packed red blood cells to achieve a hematocrit of >30% in patients that have adequate intravascular volume, but persistent tissue hypoperfusion.⁴ Serial lactate measurement is used as a measure of tissue hypoperfusion and can be monitored to guide resuscitation.⁴ Failure to normalize lactate, or an increasing lactate despite intervention, is a potent predictor of mortality in both dogs and people.^7-9

Excision & Pain Control

After diagnosing the site of infection, source control via abdominal exploratory may be warranted once cardiovascular stability has been achieved. If there is an infected tissue source that can be excised, this intervention should ideally be performed within the first 12 hours.⁴ If the patient cannot be stabilized for general anesthesia, less invasive methods of source control should be considered. For example, ultrasound-guided placement of active closed suction drains can facilitate source control for a prostatic abscess or placement of closed suction peritoneal drains and an indwelling urinary catheter in the case of septic uroabdomen.⁴ Analgesia should also be considered in septic patients, as uncontrolled pain leads to increased morbidity and mortality. 

The patient’s cardiopulmonary arrest in this case was likely due to cardiovascular collapse secondary to uncorrected hypotension.

The Generalist’s Opinion

Barak Benaryeh, DVM, DABVP

This was an unfortunate case with a sad outcome. The dog presented at 8:30 am and advanced imaging was performed at 5:10 pm. By 9:35 pm the dog was experiencing seizures and within a few hours had a final episode of cardiac arrest. Considering the rapid downhill progression, it’s difficult to say if changing interventions along the way would have made a difference. Ultimately the dog likely died from septic shock. Whether this was a consequence of urosepsis or another cause remains an unanswered question. We can look at several different aspects of this case to try to gain insight into managing similar presentations.

Imaging the Bladder

An ultrasound, contrast cystogram, and urethrogram were all performed to assess the integrity of the urinary tract and to diagnose any uroliths or urethroliths. Ultrasonography is not the most sensitive modality for detection of bladder leaks. We are told that a contrast cystogram was performed but not specifically which type. Any of the mentioned studies, if not performed properly, can lead to missed diagnoses.¹ 

For detection of bladder wall leaks, a positive contrast cystogram is ideal.² For detection of calculi, a double contrast cystogram is warranted.² These slightly different studies vary in amounts of contrast medium and air injected. They each must be done correctly with the correct amount and appropriate type of contrast agent. If there is a tear near the trigone, the inflated area of the balloon catheter can plug the leak, making it look like an intact bladder wall.¹ Be sure to deflate the balloon and shoot an additional image. It’s critical to have a familiarity and comfort level with any contrast imaging that is being performed. In addition, a bladder wall leak can seal itself with a fibrin clot. In cases such as these, although all tests are negative, there is urine in the abdomen with no active leakage.

Uroabdomen Diagnosis

An abdominal fluid to peripheral blood creatinine ratio of >2:1 was predictive of uroabdomen with a 100% specificity and 86 % sensitivity according to at least one study.³ Abdominal fluid potassium to serum potassium can also be used (the critical ratio here is 1.4)³ but we do not have the serum potassium in this case. If we look at our case, the numbers for creatinine are 7.6 mg/dL abdominal fluid/4.3 mg/dL blood, which calculates to a ratio of 1.8 (rounded). With these numbers, a definitive diagnosis of uroabdomen cannot be made. With a sensitivity of 86% we can only say that most of the time (86% of the time) a dog with these numbers would not have a uroabdomen but we can’t rule it out, either. 

Cytology of the fluid was not performed and would have been helpful in the diagnostic process. In addition to aiding to confirm uroabdomen, cytology helps guide surgical decisions for any cause of free abdominal fluid.

Treatment

This patient likely would have benefitted from some form of peritoneal lavage. Whether uroabdomen was present or not, there was free fluid in the abdomen. Even if the dog had a septic peritonitis from other causes, flushing the abdomen would have been of benefit. Abdominal lavage can be done surgically or with a peritoneal lavage catheter (not something available in most general practice settings). 

Looking back at this case, an early surgical intervention may have helped. An abdominal exploratory would have been a high risk procedure but would have provided the benefit of a definitive diagnosis as well as the ability to effectively lavage the abdominal cavity. At 5:30 pm, the lactate had increased from 1.2 to 4.0 mmol/L. Increasing lactate is associated with a worsening prognosis.⁴ Even so, the dog might have been stable enough to undergo a surgical procedure at that time. By 9:35 pm he was clearly in no condition to be taken to surgery. In general, any patient with a suspected bladder leak needs to be stabilized prior to any anesthetic procedures.⁵

Stabilizing the Patient

The clinicians in this case did a good job of trying to stabilize this dog. Enrofloxacin and ampicillin provided a complete spectrum of antibiotic coverage. It was noted that during initial catheterization the urine had a foul odor. A sample was obtained for analysis and culture but the antibiotics were not initiated at that point. While it is not generally indicated to start a catheterized patient on antibiotics immediately, this case may have been an exception: Foul-smelling urine usually indicates a UTI, and if there was an extravasation of urine into the abdomen it could have contributed to the development of sepsis. The culture did indeed grow bacteria but that was after the fact.

Alterations in Intervention

Additional interventions that could be recommended in hindsight are cytology of the abdominal fluid and starting earlier aggressive antibiotic therapy and earlier peritoneal lavage either surgically or via catheterization. We don’t know if any of these interventions would have made a difference: The patient decompensated in the course of 1 day. That sort of progression can be hard to stop, and it’s quite possible that, no matter what intervention was undertaken, this dog would not have survived.