Dog enrollment

Client-owned dogs that were presented to the University Wisconsin Veterinary Care (UWVC) between August 2018 and January 2019 with clinical signs of acute ambulatory or nonambulatory paraparesis or paraplegia were considered for study enrollment. Dogs were eligible for inclusion in the study if they had a clinical presentation of acute (≤ 6 days) ambulatory or nonambulatory paraparesis or paraplegia with neuroanatomic lesion localization to the T3-L3 spinal cord segments, spinal cord imaging via CT or MRI, and surgical decompression (hemilaminectomy) for an IVDH Hansen type I. Dogs were excluded from the study if they had been nonambulatory for ≥ 7 days. Additional exclusion criteria were concurrent systemic disease or illness (ie, chronic kidney disease, uroliths, liver disease, neoplasia), a positive urine bacterial culture at the time of surgical decompression, antimicrobial treatment within 3 months before surgery, or an indwelling urinary catheter placed during hospitalization. All study protocols were reviewed, approved, and conducted in accordance with the University of Wisconsin, School of Veterinary Medicine Animal Care and Use Committee. Informed consent was obtained from all dog owners before study enrollment.

Study design

Enrollment into this prospective, observational study included client-owned dogs diagnosed with IVDH Hansen type I and occurred at the time of surgical decompression (hemilaminectomy). Data collection time points occurred at the initial presentation to the clinic, at discharge, 2 weeks postoperatively, and between 4 and 6 weeks postoperatively. Data were recorded up to 6 to 8 months postoperatively in 3 cases and up to 2 years postoperatively in 1 case. At study enrollment, data recorded included age, breed, sex, body weight, body condition score (on a scale of 1 [emaciated] to 9 [obese]), clinical history including medications or supplements, physical and neurologic examination findings, rectal temperature, and urinary status. The results of perioperative CBC, biochemical profile, and urinalysis were also recorded.

After surgical decompression, the dogs were subjectively scored at each recheck (discharge, 2 weeks postoperatively, and between 4 and 6 weeks postoperatively). The owners provided an updated history and scored their dog’s behavior at home. At each reevaluation, the dogs had a general physical examination and a neurologic examination. Clinical assessments included assessing for urinary tract signs (hematuria, stranguria, pollakiuria, new-onset incontinence, or foul-smelling urine) or signs of systemic illness (fever, lethargy, malaise, vomiting, or signs of abdominal pain).²⁵ Laboratory diagnostics were performed at each recheck and included a CBC, renal profile (urea, creatinine, sodium, potassium, chloride, total CO², calcium, phosphorus, albumin, cholesterol, and anion gap), symmetric dimethylarginine (SDMA), urinalysis and urine culture.

In dogs with a positive urine bacterial culture, the decision to treat with antimicrobials was based on laboratory findings and any changes in the patient’s clinical condition suggestive of new-onset urinary tract signs or signs of systemic illness. Dogs with bacteriuria in the absence of clinical signs were considered to be dogs with SCB. These dogs were not treated with antimicrobials but were closely monitored for changes in their clinical condition, including developing lower urinary tract signs (pollakiuria, hematuria, stranguria, new-onset incontinence, foul-smelling urine) or signs of systemic illness suggestive of clinically relevant bacteriuria. Study dogs suspected to have clinically relevant bacteriuria were treated with antimicrobials based on antimicrobial culture and susceptibility testing. All dogs that developed SCB that persisted at the end of the study’s observation period were followed beyond the 6 weeks of the study. These dogs were monitored clinically and followed with urinalysis and urine cultures performed at varying intervals.

Subjective clinical scoring

For the neurologic evaluation and scoring, a modified scale was used.⁸ Neurologic status (NS) was scored on a 6-point scale, with 0 indicating a normal neurologic examination; 1 signs of back pain only; 2 ambulatory paraparesis and ataxia; 3 nonambulatory paraparesis; 4 paraplegia with pain perception; 5 paraplegia with absent pain perception. In this study, micturition was scored subjectively and determined by the dog’s observed ability to urinate. Micturition status (MS) was scored on a 4-point scale, with 0 indicating conscious voiding; 1 voiding with incontinence; 2a voiding required manual expression; 2b voiding required manual expression with incontinence; 3a urinary catheterization needed; 3b urinary catheterization needed with incontinence. In addition, a detailed history from the owner was attained at each evaluation to assess for any systemic or lower urinary tract signs supportive of a urinary tract infection.

Sample collection and analysis

At the time of discharge, 2 weeks postoperatively, and between 4 and 6 weeks postoperatively, each dog had venous blood (3 mL; < 1% of body weight) and urine (10 mL, via cystocentesis or void) collected. A CBC, renal panel, and urinalysis were performed using standard protocols by the UWVC Clinical Pathology Laboratory. Serum SDMA was analyzed by Idexx Laboratory using a commercially available assay validated in dogs.²⁶

Urine was submitted to the UWVC Microbiology Laboratory for routine aerobic bacteriologic culture and antimicrobial susceptibility testing. Urine was submitted directly to the laboratory or stored at 4 °C if transport was delayed. Urine was plated for a quantitative bacteriologic culture within 24 hours after collection. The quantitative culture consisted of plating 10 or 100 μL of urine based on whether the sample was collected as a voided sample (10 µL) or by cystocentesis or catheterization (100 µL). Bacteriologic cultures of urine were evaluated for growth at 24 and 48 hours of incubation. Growth was quantitated as the number of colony-forming units (CFUs)/mL. Bacterial growth cutoffs were based on human and veterinary recommendations.^27–32 For this study, a positive result from a clean voided sample was defined as 1 to 3 pathogens present at > 100,000 CFUs/mL. The presence of > 3 species of microorganisms or < 10,000 CFUs/mL indicated probable urethral contamination, and the results of these cultures were considered negative. Any microorganism growth from urine collected by cystocentesis was considered clinically important unless the growth was suggestive of gastrointestinal contamination during cystocentesis. All relevant isolates were identified using standard microbiological methods. Minimal inhibitory concentrations were determined by broth microdilution for all important isolates. Methods and antimicrobial susceptibility breakpoint interpretations were performed in accordance with Clinical Laboratory Standards Institute guidelines.^33,34

Statistical analysis

Descriptive statistics were used to describe the results of this observational study. Continuous variables were reported as medians and ranges. Categorical data were tabulated for comparison. Statistical calculations were performed using GraphPad Prism 7.04.

Study population

Twenty-one dogs were evaluated and initially enrolled in the study. One dog was excluded from the study due to a positive urine bacterial culture with Escherichia coli (> 100,000 CFU/mL) at the time of surgical decompression, resulting in a total of 20 dogs studied. Sixteen dogs were purebred and included Dachshunds (n = 5), Cavalier King Charles Spaniels (3), Shih Tzu (2), French Bulldogs (2), and 1 each of the following breeds: Basset Hound, Miniature Schnauzer, Saint Bernard, and Toy Poodle. Four dogs were mixed-breed dogs. Nine dogs were spayed females, and 11 were neutered males. The median age and weight were 6 years (range, 2 to 12 years) and 6.5 kg (range, 4.5 to 57 kg). Two dogs required antimicrobial treatment for postoperative pneumonia. Their initial data were included, but they were excluded from further follow-up when antimicrobial treatment was started 1 week and 3 weeks postoperatively. A total of 18 dogs completed the study’s 4- to 6-week observation period. Longer-term follow-up ranged from 12 weeks to up to 2 years in the 4 dogs that developed bacteriuria.

All dogs were diagnosed with IVDH Hansen type I localized to spinal cord segments T3 to L3 identified by MRI (n = 14), CT (4), or CT myelogram (2). In 15 dogs, a single-site hemilaminectomy at T11-12 (n = 4), T12-13 (1), T13-L1 (5), L1-2 (3), L2-3 (1), and L3-4 (1) was done. A 2-level hemilaminectomy was done in 5 dogs and included T11-13 (n = 1), T12-L1 (1), T13-L2 (1), L2-4 (1), and T11-12 and L1-2 (1). Intraoperative analgesia included ketamine (2 to 10 µg/kg/min) and fentanyl (2 to 5 µg/kg/h) as a constant rate infusion (CRI) in all dogs. All dogs received perioperative prophylactic cefazolin (22 mg/kg) intravenously with repeated dosing every 90 minutes until the end of surgery. Surgical prophylaxis with cefazolin is considered extra-label drug use despite it being considered the standard of care for clean neurosurgeries in dogs.³⁵

The dogs were hospitalized for a median of 3 days (range, 2 to 8 days). Postoperative pain management consisted of fentanyl (2 to 5 µcg/kg/h CRI) in all dogs and the addition of ketamine (2 to 10 µg/kg/min CRI) in 6 dogs. One dog received a bupivacaine liposome injection (5.3 mg/kg intramuscularly; Nocita) during incision closure. Six dogs were transitioned to a fentanyl patch (12.5 to 125 µg/h), 15 to gabapentin (median, 10.7 mg/kg, PO q 8 to 12 hours; range 5.2 to 17.8 mg/kg, PO, q 8 to 12 hours), and 2 received pregabalin (0.98 or 1.9 mg/kg, PO, q 12 h). Six dogs were treated with trazadone (median, 4.3 mg/kg, PO, q 12 h; range, 2.1 to 5.6 mg/kg, PO, q 12 h). Fourteen dogs were treated with an NSAID (carprofen; median, 2.2 mg/kg, PO, q 12 h; range, 1.9 to 4.2 mg/kg, PO, q 12 h; n = 13) or meloxicam (0.11 mg/kg, PO, q 24 h; 1). One dog, not treated with an NSAID, was treated with prednisone (0.3 mg/kg, PO, q 48 h). Medications prescribed to facilitate bladder emptying included diazepam (median, 0.18 mg/kg, PO, q 8 h; range, 0.17 to 0.36 mg/kg, PO, q 8 h; n = 5) and prazosin (median, 0.2 mg/kg, PO, q 12 h; range, 0.04 to 0.2 mg/kg, PO, q 12 h; 4).

Neurologic and micturition status monitoring

The neurologic scores ranged from 2 to 5 at enrollment with improving scores at 4 to 6 weeks after surgery, and micturition scores ranged from 0 to 2 for most of the dogs (Table 1). Five dogs had a MS of 0 before surgery and at all rechecks. Four dogs with a MS of 0 before surgery required manual bladder expression (n = 3) or urinary catheterization (1) postoperatively and returned to MS 0 after discharge. Two dogs that had a MS score of 0 before surgery became urinary incontinent postoperatively (MS 1) at all rechecks. Two dogs with a MS of 2a before surgery remained at 2a at all rechecks. Two dogs with a MS of 1 and 2a, respectively, became normal after discharge. One dog that had a MS of 1 before surgery went to a MS of 2a at 2 weeks and 2b at 4 to 6 weeks after surgery. Of the 5 dogs with a neurologic score of 5 before surgery, 2 dogs did not improve, 1 dog improved 1 grade and 2 dogs improved 2 grades. Of the 4 dogs with a NS of 4 before surgery, 1 improved 1 grade, and 3 improved 2 grades. Of the 6 dogs with a NS of 3 before surgery, 1 did not improve, and 5 improved 1 grade. All 4 dogs with a NS of 2 remained at a score of 2 postoperatively.

Clinical and laboratory monitoring

Dogs were monitored for the development of clinically important bacteriuria using changes in clinical signs and laboratory diagnostics. No dogs developed clinical signs of systemic illness, supported by a lack of marked increases in blood neutrophil count (median, 7,100 cells/µL; range, 3,200 to 36,600 cells/µL; reference interval, 5,000 to 14,000 cells/µL), the development of a left shift (bands; median, 0 cells/µL; range, 0 to 800 cells/µL; reference interval, 0 to 200 cells/µL), or evidence of fever (rectal temperature; median, 38.8 °C; range, 36.6 °C to 40.3 °C; reference interval, 38.3 °C to 39.1 °C). Some dogs’ rectal temperatures were > 39.1 °C at the time of re-evaluation despite any signs of clinical illness. Their clinical assessments were consistent with excitement-induced hyperthermia based on the dogs’ attitude, demeanor, and physical examinations.

No dogs had altered kidney function based on evaluations of urine-specific gravity (median, 1.037; range, 1.011 to 1.060; reference limit, ≥ 1.030), BUN (median, 14 mg/dL; range, 5 to 29 mg/dL; reference interval, 7 to 32 mg/dL), creatinine (median, 0.7 mg/dL; range, 0.5 to 1.6 mg/dL; reference interval, 0.5 to 1.5 mg/dL), and SDMA (median, 8.5 µg/dL; range, 4 12 µg/dL; reference interval, 0 to 14 µg/dL). Median urinary pH was 7 (range, 6.5 to 8) at the time of surgery, 6.5 (range, 6 to 8) at discharge, 7 (range, 6 to 8.5) at 2 weeks after surgery, and 7 (range, 6 to 8.5) at 4 to 6 weeks after surgery. Urine sediment findings documented the absence of pyuria at the time of surgery in 19 of the 20 dogs. The one dog (NS, 5; MS, 2a) with pyuria (1 to 5 WBC/hpf) had a urine-specific gravity of > 1.060 and a pH of 7. This dog’s pyuria had cleared by the time of hospital discharge.

Fourteen of the 18 dogs with 4 to 6 weeks of follow-up never developed bacteriuria despite persistent neurologic dysfunction (NS, 2 to 5), and in some dogs, continued abnormal micturition (MS, 1 to 2b; n = 4). Of these 14 dogs, 2 (one dog with an NS of 3 and an MS of 2a and another dog with an NS of 2 and an MS 2a) had documented transient pyuria (1 to 5 WBC/hpf) at hospital discharge. Pyuria had resolved by the 2-week recheck examination for one dog and by the 4- to 6-week recheck examination for the other dog.

Dogs with bacteriuria follow-up and outcome

Four dogs with IVDH Hansen type I had documented bacteriuria initially identified postoperatively at discharge (4 days; n = 1), 2 weeks postoperatively (2), and 4 to 6 weeks postoperatively (1). These 4 dogs included 3 spayed females and 1 neutered male with neurolocalization to spinal cord segments T11-12 (n = 3) and T13-L1 (1). The NS and MS for the 4 dogs with postoperative bacteriuria were 5 and 2a, respectively (n = 1); 3 and 2a, respectively (1); 4 and 0, respectively (1); and 2 and 0, respectively (1). In these 4 dogs, in-hospital postoperative pain management included fentanyl and ketamine (n = 3) or fentanyl alone (1). One dog continued fentanyl as a patch. In addition, pain management included oral administration of gabapentin (n = 4), meloxicam (1), and carprofen (3). Oral administration of diazepam was prescribed to facilitate bladder emptying (n = 1). Trazadone was prescribed to control anxiety and facilitate confinement during recovery (n = 1).

Three dogs experienced an improvement in neurologic function after surgery with no improvement in micturition score during the study observation period (dog 1: NS grade 5 to 3, MS 2a; dog 2: NS 3 to 2a, MS 2a, normal at 6 months; dog 3: NS 4 to 3, MS 0). The 1 remaining dog (dog 4) had no change in NS (2) or MS (0) after surgery and through the 4 to 6 week recheck but became neurologically normal at 6 months.

None of the dogs had clinical signs at the time bacteriuria was initially documented. Dog 1 had 2 different species cultured at discharge (Pseudomonas aeruginosa, > 100,000 CFU/mL; NS/MS 5/2a) and 2 weeks (Enterococcus faecalis, > 100,000 CFU/mL; NS/MS 4/2a) followed by 2 isolates cultured at 4 to 6 weeks (E faecalis, > 100,000 CFU/mL; Staphylococcus pseudintermedius, > 100,000 CFU/mL; NS/MS 3/2a), with both persisting at 16 weeks (NS/MS 3/2a), and the dog became clinical for a UTI. Dog 2 had growth of 2 isolates (Proteus mirabilis, > 100,000 CFU/mL and Citrobacter freundii, > 100,000 CFU/m; NS/MS 2c/2a) at 4 to 6 weeks, of which 1 isolate (P mirabilis, > 100,000 CFU/mL; NS/MS 2c/0) persisted at 16 weeks when clinical signs of a UTI became evident. Dog 3 had growth of E faecium (> 100,000 CFU/mL; NS/MS 3/0) at 2 weeks and S pseudintermedius (> 100,000 CFU/mL; NS/MS 3/0) at 4 to 6 weeks, then growth of 2 bacterial isolates (E faecalis, > 100,000 CFU/mL and S pseudintermedius, > 100,000 CFU/mL; NS/MS 2c/0) at 18 weeks when there were clinical signs of UTI. Dog 4 was repeatedly urine bacterial culture positive with S canis (> 100,000 CFU/mL) postoperatively at 2 (NS/MS 2/0) and 4 to 6 weeks (NS/MS 2/0) but did not show any clinical signs of UTI until the positive culture at postoperative week 26 (NS/MS 2/0).

All bacterial isolates associated with clinical bacteriuria in these 4 dogs maintained wide antimicrobial susceptibility. Clinical signs included new-onset urinary incontinence or urinary accidents (n = 3) or foul-smelling urine, stranguria, and pollakiuria (1). None of these 4 dogs developed systemic clinical signs or markers of systemic illness (Table 2). Changes noted on the urinalysis in these dogs at the time they developed clinical signs included an increase in urinary pH ≥ 8 (3/4), pyuria >5 WBC/hpf (3/4), and presence of bacteria (3/4) in the urine sediment analysis. Antimicrobial treatments included amoxicillin and clavulanic acid (11 to 22 mg/kg, PO, q 12 h for 14 days; n = 3) or amoxicillin (18 mg/kg, PO, q 12 h for 14 days; 1), with the resolution of clinical signs and negative urine bacterial cultures in all 4 dogs. Treatment with amoxicillin with clavulanic acid or amoxicillin was based on bacterial culture and susceptibility testing. Amoxicillin and amoxicillin with clavulanic acid are recommended as first-line treatments for urinary tract infections in dogs despite their use being considered extra-label in dogs.⁹ One dog, with longer-term follow-up, developed recurrent lower urinary tract infections that required intermittent antimicrobial treatment over the 2 years of follow-up. The organisms associated with this dog’s recurrent bacteriuria and clinical signs were variable (Escherichia coli, Klebsiella aerogenes, Enterococcus faecium, E faecalis, Staphylococcus pseudintermedius, Pseudomonas aeruginosa, and Corynebacterium urealyticum), but isolates maintained wide antimicrobial susceptibility.