Introduction

Bacterial urinary tract infections (UTIs) are estimated to affect 14% of dogs in their lifetime.¹ The incidence of UTIs increases in patients with neurologic dysfunction secondary to failure to store or eliminate urine.² Patients with spinal cord injury (SCI) secondary to intervertebral disc disease (IVDD) of the thoracolumbar spine develop a voiding dysfunction associated with urinary retention secondary to disruption of motor and sensory pathways of the detrusor muscle reflex.^3,4 This upper motor neuron spinal cord disease leads to incomplete or absent urinary bladder contraction and possible overflow incontinence.^3,4 The prevalence of UTIs in dogs with Type I intervertebral disc extrusion has been reported as high as 38% within the first 6 wk of surgery, increasing in up to 52% of patients with indwelling urinary catheters.^5,6 Risk factors for the development of a UTI in postsurgical canine patients with IVDD include female sex, paraplegia, inability to urinate voluntarily, duration of indwelling urinary catheterization, lack of perioperative antibiotics, lower intraoperative body temperature, antibiotic administration during urinary catheterization, and administration of dexamethasone less than 48 hr prior to surgery.^5,7–9 In one study, the risk of UTIs in patients with urinary catheters increased by 20% for each yr increase in age, 27% for each day increase in duration of catheterization, and 454% with antimicrobial administration.⁸

Recognition and effective treatment of UTIs in patients with myelopathy is crucial due to the consequence of persistent infection and reinfection in dogs.¹⁰ Poor owner compliance and inadequate dosing further complicate the ability to effectively clear UTIs in companion animals.¹¹ Cefovecin^a is a third generation cephalosporin labeled for use in dogs for treatment of UTIs. Its parenteral administration and ability to maintain peak plasma concentration for up to 14 days improve the rate of cure and reduce bacterial resistance.¹²

The purpose of this study was to validate the use of cefovecin in patients with SCI. The hypothesis was that cefovecin would significantly reduce UTIs when compared to patients who received cefazolin perioperatively for thoracolumbar hemilaminectomy. Secondary objectives included investigating incidence of UTIs with the following: urinary tract management, duration of neurologic signs, grade of neurologic deficit, and time to voluntary urination and ambulation.

Materials and Methods

Medical records were searched from client-owned dogs admitted to the Animal Specialty Center between January and August of 2012 who received either cefovecin or cefazolin as a prophylactic perioperative antibiotic for hemilaminectomy due to clinical spinal cord compression secondary to thoracolumbar intervertebral disc disease confirmed with MRI. Patients were excluded if they had a history of UTIs or recent antibiotic administration.

Data obtained from medical records of dogs selected for study included age, sex, breed, body weight, body condition score (BCS), duration of clinical signs, laboratory chemistry and complete blood count, and recent medications. BCS was recorded at the time of admission to the hospital using a previously described classification system (scale of 1 to 9).¹³ At the time of admission and at each subsequent re-evaluation, a complete neurologic exam was performed by a board-certified neurologist. Duration of neurologic signs for patients was retrospectively subdivided into <7 days and >7 days. Severity of neurologic deficit was retrospectively graded 0 to 5, adapted from a previously reported scale (Table 1).⁵

Table 1 Definitions of Neurologic Deficit Grading

View Fullscreen

Immediately after MRI^b and during concurrent general anesthesia, routine cystocentesis was performed for urinalysis and urine culture on all patients based on hospital protocol. Patients were then administered a prophylactic perioperative antibiotic as selected by the surgeon. Patients meeting the inclusion criteria were retrospectively selected for placement into one of the following groups: CEFOVECIN group (cefovecin 8 mg/kg subcutaneously once) or CEFAZOLIN group (cefazolin 22 mg/kg IV every 120 min from induction to the end of anesthesia). Hemilaminectomy was performed in all cases by either a board-certified specialist in neurology (J.M.B.) or surgery (J.J.B.). Those patients without voluntary control of urination received an indwelling Foley urinary catheter (subgroup, UCATH) or were managed with manual bladder expression (subgroup, NOUCATH). Those patients maintaining voluntary control of urination after surgery were also placed in the NOUCATH subgroup. Medical records were used to identify the duration of urinary catheterization (hr) and return to voluntary urination and ambulation if present prior to discharge.

Standard hospital protocol for patient follow-up included recheck performed at wk 2 and 6 postoperatively for neurologic evaluation and acquisition of a sterile urine sample via cystocentesis for urinalysis and urine culture. Exam history and client communication records were used to determine final time to urination and ambulation after hospital discharge. All sterile urine samples obtained were submitted to a commonly used reference laboratory^c for urinalysis, urine culture, and antimicrobial susceptibility testing using standard automated techniques. Urine sample sediments and urine culture results were evaluated by a laboratory technician with results categorized according to criteria listed in Table 1. In samples with a positive urine culture, bacterial concentrations were >100,000 colony forming units /mL. All samples were tested with cefovecin discs (30 μg) with a mean inhibitory concentration reference range of 2–8 μg/mL, except for those with growth of Enterococcus spp. or Pseudomonas spp. as cefovecin was not anticipated to be effective.¹²

Patients receiving prophylactic cefazolin who demonstrated positive urine cultures at admit and wk 2 and 6 were dispensed and treated with a sensitive antibiotic based on susceptibility testing. Patients receiving prophylactic cefovecin who demonstrated positive urine cultures and sensitivity to the drug at wk 2 and 6 were administered a second dose of cefovecin within 7 days of test results. For patients whose positive urine culture demonstrated resistance to cefovecin, a sensitive antibiotic was dispensed based on susceptibility testing. Persistent infections were speculated when subsequent positive urine culture with identical bacteria and susceptibility results were identified.

Results

Antibiotic and Urinary Tract Management

Sixteen of 39 patients (41%) were placed in the CEFOVECIN group, of which 8/16 received an indwelling urinary catheter after hemilaminectomy (CEFOVECIN/UCATH). Twenty-three out of 39 patients (59%) were placed in the CEFAZOLIN group, of which 12/23 received a urinary catheter (CEFAZOLIN/UCATH). Patient groups and subgroups are listed in Table 2. Patients with urinary catheterization (n = 20) were managed for a mean of 62.6 hr (median, 63 hr; range, 34-84 hr). There was a significant difference in the mean age (P = 0.046) between the NOUCATH (7.3 yr [median, 6.5 yr; range, 3–14 yr]) and UCATH (5.4 yr [median, 5 yr; range, 3–10.5 yr]) subgroups. The mean grade of neurologic deficit of patients in the UCATH subgroup (4.1 [median, 4; range, 4–5]) was significantly worse (P = 0.004) than the NOUCATH subgroup (2.1 [median, 2; range, 1–4]). Patients in the CEFOVECIN group had a significantly worse mean grade of IVDD at admit (3.6 [median, 4; range 5–2]) than those in the CEFAZOLIN group (3.1 [median, 4; range 5–1]) (P = 0.042). Mean grade of neurologic deficit of all patients at presentation to the hospital was 3.3 (range, 1–5). The overall mean grade of neurologic deficit for study patients was 1.7 (range, 0–4) at wk 2. The overall mean grade of neurologic deficit for study patients was 1.2 (range, 0–3) at wk 6. There were no significant differences between the antibiotic groups and urinary tract management subgroups.

Table 2 Categorization of Patients Based on Antibiotic and Urinary Tract Management

View Fullscreen

Urinalysis and Urine Culture

One hundred and five urinalyses and 106 urine cultures were performed and retrospectively reviewed for the 39 patients in the study. Results of urinalysis and urine culture were available for 36/39 and 31/39 patients at wk 2 and 6, respectively. Patients had an overall prevalence of UTIs of 20.5% over the 6 wk study period. At wk 2, patients who received cefovecin trended towards positive urine cultures (P = 0.061) when compared to patients in the cefazolin group. Patients did not show any predilection toward UTIs between the two antibiotic groups at admit (P = 0.778) or wk 6 (P = 0.841). Overall, 25% of patients receiving cefovecin had positive urinary cultures compared to 17.4% in patients receiving cefazolin, which was not statistically significant. There was an equal distribution of patients with positive urine cultures across the main subgroups and group/subgroup combinations (Table 2).

Urine sediment was evaluated from 105 sterile urine samples. Numbers of cells detected in sediment were categorized on a sliding grouping of 1–4 (RBCs and WBCs) and 1–5 (bacteria) for data entry into the stats program to determine significance (Table 3). Hematuria was the most common microscopic abnormality, appreciated in 23/105 samples in 20 patients. The mean results for the 23 samples with hematuria were 2.0 or about 11–20 RBCs/high-powered field (median, 2; range, 1–4). Hematuria was present in 3/20 patients on successive urinalyses at admit and wk 2. All three patients had received perioperative cefovecin, and 1/3 patients were positive on urine culture at wk 2. The median hematuria results for these three patients was 3.0. Two of 20 patients with hematuria had a positive urine culture (mean, 1.5), which was not clinically significant.

Table 3 Categorization of Laboratory Results for Statistical Analysis

View Fullscreen

Pyuria was found in 4/105 samples in three patients. The four samples with pyuria had a mean result of 1.3. All four urine samples with pyuria also had positive urine cultures. One patient in the CEFOVECIN/NOUCATH had pyuria on successive urinalyses at admit and wk 2, with a mean result of 1.5, which was not statistically significant. Time to voluntary urination and ambulation were each 12 days. Urine culture for the two samples were both positive for coagulase negative Staphylococcus spp. and sensitive to cefovecin. The patient was treated with a second dose of cefovecin. Recheck urinalysis and urine culture 46 days after surgery demonstrated normal urine sediment and a negative urine culture.

Bacteriuria was found in 7/105. The mean results for the seven samples with bacteriuria were 2.9 (median, 3; range, 1–5). Patients with bacteriuria had a mean WBC of 0.7 (median, 1; range, 1–2) and RBC of 0.4 (median, 0; range, 0–2). All patients with bacteriuria were positive for bacterial growth on urine culture.

Thirteen bacterial isolates were identified from 11/106 urine culture samples over 6 wk in eight different patients: 3/11 at admit, 5/11 at wk 2, and 3/11 at wk 6 (Table 4). The signalment of the three patients with greater than one positive urine culture included a 14 yr old female spayed bichon frise (patient 1), a 3 yr old male intact mixed–breed dog (patient 5), and a 3 yr old male castrated French bulldog (patient 7). The eight patients with positive urine cultures were evenly distributed amongst the four combinations of study groups and subgroups. All eight patients with positive urine cultures had asymptomatic UTIs as no patients with positive culture displayed dysuria, pollakiuria, and/or increased urgency of urination.¹⁴

Table 4 Characteristics of Patients with Urinary Tract Infections

View Fullscreen

The most common urinary tract isolate was Enterococcus spp., of which two were multi-drug sensitive and two were multi-drug resistant on susceptibility testing (Table 4). Five of 12 bacterial isolates were resistant to cefovecin (four Enterococcus and one beta-hemolytic Streptococcus). One isolate spp. was unknown (1 was gram negative bacilli, the sample was unable to be grown and speciated; no susceptibility was available). Two of the 13 isolates were suspected persistent infections as they were cultured on consecutive urine samples with equivalent antimicrobial susceptibility in two patients (Staphylococcus spp. in patient 1 and Enterococcus spp. in patient 7).

At admit, patients in the NOUCATH subgroup 3/19 (16%) trended toward positive urine cultures (P = 0.064) when compared to patients in the UCATH subgroup (0/20). Patients who received cefovecin trended toward positive urine cultures at wk 2 (P = 0.061). At wk 6, patients in the UCATH subgroup 3/15 (20%) trended toward positive urine cultures (P = 0.06) when compared to the patients in the NOUCATH subgroup (0/16).

Four of five patients that cultured positive at wk 2 had received preoperative cefovecin. Two of the four cultures were sensitive to cefovecin (patients 1 and 8); one culture did not speciate and susceptibility testing could not be performed (patient 2). Three of the four patients received a second dose of cefovecin and later demonstrated a negative urine culture at wk 6 (patients 1, 2, and 8). Patients 5 and 7 with Enterococcus spp. each received 2 wk courses of marbofloxacin at 2.75 mg/kg orally q 24 hr. Patient 5 was positive with Enterococcus spp. at wk 6 and 11; the antimicrobial susceptibility differed in enrofloxacin sensitivity. Patient 7 was positive on subsequent urine culture at wk 6 with Enterococcus spp. and beta-hemolytic Streptococcus. Patient 7 was lost to follow-up.

Time to Voluntary Urination

Mean time to voluntary urination was 8.3 days for all 39 patients (median, 3 days; range, 1–50 days). At admit, patients presenting with grade 5 IVDD had significantly longer mean time to voluntary urination (31 days [median, 31 days; range, 12–50 days] when compared to patients with grade 4 IVDD (7.2 days [median, 4 days; range, 1–19 days; P = 0.001]), grade 3 IVDD (7.3 days [median, 2 days; range, 1–22 days; P = 0.002]), and grade 2 IVDD (4.6 days [median, 1 day; range, 1–16 days; P = 0.001]) (Figure 1). At wk 2, patients with a grade 4 IVDD had significantly longer mean time to voluntary urination (33.5 days [median, 33.5 days; range 17–50 days]) when compared to patients with grade 3 IVDD (17.2 days [median, 17.5 days; range, 4–30 days; P = 0.009]), grade 2 IVDD (5.3 days [median, 2 days; range, 1–17 days; P = < 0.0001]), grade 1 IVDD (3.3 days [median, 2 days; range, 1–17 days; P = < 0.0001]), and grade 0 IVDD (4.8 days [median, 3 days; range, 1–16 days; P = < 0.0001]). At wk 6, patients with a grade 2 IVDD had significantly longer mean time to voluntary urination (19.1 days [median, 17 days; range, 1–50 days]) when compared to patients with grade 1 IVDD (5.2 days [median, 2 days; range, 1–17 days; P = 0.001]) and grade 0 IVDD (4.3 days [median, 2.5 days; range, 1–16 days; P = 0.005]). Patients with positive urine cultures at 6 wk had significantly higher mean time to voluntary urination (24.3 days [median, 21 days; range, 2–50 days]) when compared to patients with negative cultures (mean 7.9 days [median, 4 days; range, 1–30 days; P = 0.01]).

View Figure

• Download as Powerpoint
• Download Figure

Time to Ambulation

Mean time to ambulation was 11.2 days for 38/39 patients (median, 3.5 days; range, 1–63 days) who were available follow-up. At admit, patients presenting with grade 5 IVDD had significantly longer mean time to ambulation (28.5 days [median, 28.5 days; range,14–43 days]) when compared to patients with grade 2 IVDD (2.4 days [median, 1 day; range, 1–10 days; P = 0.034]) (Figure 2). At wk 2, patients with a grade 4 IVDD had significantly longer mean time to ambulation (36.5 days [median, 36.5 days; range 30–43 days]) when compared to patients with grade 2 IVDD (9.1 days [median, 5 days; range, 1–42 days; P = 0.003]) grade 1 IVDD (2.8 days [median, 1 days; range, 1–10 days; P = 0.0004]) and grade 0 IVDD (1.2 days [median, 1 days; range, 1–2 days; P = 0.001]). At wk 2, patients with grade 3 IVDD had a similar mean time to ambulation (30.2 days [median, 25.5 days; range, 10–63 days]) to those patients with grade 4 IVDD, and also had a significantly longer mean time to ambulation when compared to patients with grade 2 IVDD (P = 0.0004), grade 1 IVDD (P < 0.0001), and grade 0 IVDD (P = 0.0001). At wk 6, patients with a grade 3 IVDD had significantly longer mean time to ambulation (54 days [median, 54 days; range, 45–63 days]) when compared to patients with grade 2 IVDD (16.2 days [median, 12 days; range, 3–43 days; P = 0.0003]), grade 1 IVDD (9.5 days [median, 4 days; range, 1–42 days; P < 0.0001]), and grade 0 IVDD (1.2 days [median, 1 day; range, 1–2 days; P < 0.0001]). A positive correlation was found for time to voluntary urination and time to ambulation (P = 0.0002). Overall, the eight patients with positive urine cultures took 1.8 times longer to reach voluntary urination and over 2.5 times longer to reach ambulation then patients with negative culture results (Table 4).

View Figure

• Download as Powerpoint
• Download Figure

Discussion

At the time of publication, this was the first article to evaluate the use of cefovecin as a prophylactic perioperative antibiotic. Cefovecin has a reported efficacy rate in clinical cure (i.e., resolution of clinical UTI signs) of 88.1% and overall cure rate (i.e., confirmed with negative urine culture) of 73.5%, which was greater than the 70% clinical resolution rate and 45% rate of clinical cure seen with cephalexin-treated dogs.¹² Patients in that study were clinical for UTI and excluded if they received steroids (either short-acting within 7 days prior to day 0 or sustained-release formulations).¹²

In the present study, cefovecin proved safe with no reported anaphylactic or hypersensitivity reactions. No patients in the study developed a postoperative incisional infection. In a previous study, patients not administered perioperative cefazolin every 90 to 120 min were three times more likely to develop a UTI.⁷ The overall prevalence of UTIs in the present study in patients receiving prophylactic perioperative cefovecin for hemilaminectomy was 25% compared to 17.4% in patients receiving cefazolin. The overall prevalence of UTIs in all dogs over the 6 wk period was 20.5%, which was similar to a previous report (21%) and less than two other reports (27%, 38%).^5,7,8 Urinary tract management did not adversely impact the prevalence of UTIs. Therefore, this study did not support the use of cefovecin over cefazolin as a prophylactic perioperative antibiotic for reduction in UTIs.

At wk 2, 4/16 patients who received cefovecin demonstrated positive urine cultures. Two of the isolates, Staphylococcus spp. (patient 1) and Proteus mirabilis (patient 8), were both sensitive to cefovecin. The persistent infection in patient 1 was considered secondary to protracted abnormal micturition as cefovecin successfully cleared the infection after second administration. Two of those four patients receiving cefovecin cultured Enterococcus spp., a bacteria resistant to the drug.¹² Both patients were managed with urinary catheters, leading to possibility for nosocomial spread. Patients with neurologic disease managed with an indwelling urinary catheter were at an increased risk of developing Enterococcus spp. in one study.⁸ It is possible that concurrent use of cefovecin during urinary catheterization could have selected for the drug-resistant uropathogen as both patients demonstrated negative culture at admit. Successive culture of Enterococcus spp. at wk 2 and 6 in patient 7 raise concern for persistent infection despite normal ambulation and urination observed within the first wk postoperatively. Seguin et al. identified 11/71 dogs with paraparesis as having reinfections, persistent UTIs, or both.¹⁰

The prevalence of UTIs in patients in the present study increased with patients with higher grades of neurologic deficit and longer times to voluntary urination and ambulation, supporting the findings found in other studies.^5,7 Although not significant, patients with urinary catheterization were found to have an increase in UTIs at wk 2 (3/5 patients). Short-term use of urinary catheterization in this study (mean, 62.6 hr) may have limited its impact as a risk factor.⁸ The five patients with initial UTIs at wk 2 and 6 had the highest mean days of time to voluntary urination (18 days) and ambulation (27 days), further highlighting these risk factors. The mean grade of neurologic deficit amongst this select group was highest at the respective study intervals.

The use of prophylactic antibiotics in humans with SCI is controversial. Early studies using trimethoprim-sulfamethoxazole failed to demonstrate a significant difference in the incidence of clinical infection.¹⁵ Furthermore, the treatment of asymptomatic bacteriuria (ASB) in humans with SCI results in early recurrence and multi-drug resistant strains in catheter-free patients and has shown no benefit in prevention of UTIs in patients with intermittent or indwelling catheterization.¹⁶ However, patients with concurrent immunosuppressive disease, frequent recurrent symptomatic UTIs, or long-term neurologic and subsequent urinary tract dysfunction may benefit from early diagnosis and treatment of asymptomatic bacteriuria. Future studies targeting this subset of patients with SCI could validate routine use of long-acting prophylactic antibiotics.

Patients that developed positive urine cultures in this study generally had asymptomatic bacteriuria. Treatment of human patients with ASB is controversial.^17–20 The avoidance in treating ASB in humans with SCI is more commonly due to chronicity of disease. Humans with long-term catheterization (>10 days) are at greater risk of chronic bacterial contamination of the urine with a polymicrobial spectrum.¹⁹ Prophylactic or long-term antimicrobial therapy in these patients is likely to lead to selection of multi-drug resistant strains.^18,20 Similarly, several studies in dogs support the association of urinary catheterization and UTIs with one study reporting the likelihood of infection to increase by 27% for each day of catheterization.^6,8,21 Therefore, optimizing bladder evacuation with the proper techniques (i.e., clean intermittent catheterization, short-duration indwelling catheterization, complete manual expression) may prove more beneficial, especially during periods of paraplegia.²² Although not significant, patients with positive urine cultures in the present study also took approximately 1.5 times longer to reach voluntary urination and 2 times longer to reach ambulation than the remainder of the study population. The effect of ambulation and voluntary urination on prevalence of UTIs was similar to findings in another study.⁷ Patients with higher-grade neurologic deficits at admit demonstrated a significantly higher risk of UTI. No associations were seen with the incidence of UTIs in relation to duration of neurologic signs or use of indwelling urinary catheters.

The power of the study was limited by low patient numbers and some minor expected attrition of patients by the 6 wk follow-up. A more appropriate prospective study design with a higher power and appropriate urine collection at the designated time points may have demonstrated a difference in UTIs in patients with perioperative cefovecin versus cefazolin.

Conclusion

The use of cefovecin showed no advantage when compared to conventional prophylactic antibiotics. Additional studies may demonstrate the utility of long-acting antibiotics in conditions where risk of chronic infection is present. Given the similar incidence of UTIs at each of the urine collection points and lack of accompanying urinary tract signs, long-term urinary surveillance by sterile urine culture in veterinary patients with SCI continues to be a strong recommendation, especially in patients with prolonged time to voluntary urination and ambulation.