Surgical site infection is an important postoperative complication of spinal surgery in human medicine.1,2 Several studies1–4 have been performed to estimate the incidence of surgical site infection in humans following various neurologic procedures. Infection rates vary with the type of surgical procedure and approach (anterior vs posterior) to the spine. The incidence of surgical site infection after decompressive laminectomy, diskectomy, and fusion is approximately 3%, but this can increase to as high as 12% when surgical implants are added.1,2
Surgical site infections may affect the incision or deep tissues and encompass skin infections, epidural abscesses, vertebral osteomyelitis, diskitis, and spondylodiskitis.5 In human medicine, the term spondylodiskitis indicates infection of the vertebral disk (diskitis) and adjacent vertebrae (spondylitis).6 Spondylodiskitis that develops as a result of surgical site infection is referred to as POS.7 In the past several years, POS in humans appears to have increased in frequency because of the increasing number of invasive spinal procedures as well as improvements in diagnostic capabilities.3 Despite undeniable improvements in diagnostic techniques, antimicrobial treatment, surgical technique, and postoperative care, POS continues to result in considerable morbidity and occasional death in humans.1,4 Factors associated with an increased risk of POS after spinal surgery in humans include age, obesity, diabetes, smoking, poor nutritional status, use of steroidal antiinflammatory drugs, posterior surgical approach, prolonged duration of surgery, and use of spinal instrumentation.8–13
Although termed differently, human spondylodiskitis and veterinary diskospondylitis likely refer to the same condition14,15; therefore, the 2 terms are used interchangeably throughout the present report. Similarly, given the similarities between POS and POD, POD is used to refer both conditions. To the authors’ knowledge, few data are available regarding development of POD in veterinary patients following neurosurgery.16,17 Early identification and appropriate management of this condition are important to avoid adverse outcomes, including additional neurologic deficits and death.
Given that the incidence of POD development in humans is increasing in parallel with the growing number of possible surgical and diagnostic procedures, we hypothesized that the incidence of POD development in dogs would also increase in parallel with the growing number of spinal surgeries performed. The purpose of the study reported here was to determine the incidence of and risk factors for development of POD at the surgical site in a large number of dogs with IVDH that underwent spinal decompression surgery.
Materials and Methods
Animals
Medical records of dogs that had undergone spinal decompression surgery (ventral slot procedure, hemilaminectomy, or minihemilaminectomy) for IVDH at the Portoni Rossi Veterinary Hospital in Bologna, Italy, between January 2007 and January 2011 were reviewed. Dogs that had undergone other surgical procedures such as vertebral fusion with use of instrumentation or fenestration (prophylactic or therapeutic) of the intervertebral disk were excluded from the study.
Data collection
Data obtained from medical records included signalment (dog breed, age, sex, and body weight), clinical history, use of corticosteroid drugs or antimicrobials, findings of complete physical and neurologic examinations, preoperative neurologic status, anatomic location and type of IVDH (extrusion vs protrusion), type of surgical procedure performed, and wound appearance and signs of spinal pain (present or absent) or other signs of infection (ie, pyrexia, weight loss, lethargy, dysrexia, or anorexia) at follow-up examinations. Dogs with incomplete medical records with respect to the aforementioned data or those that were lost to follow-up were excluded from the study. Additional information related to the anesthetic procedure (duration and mean esophageal temperature) and duration of hospitalization was recorded when available.
Clinical examination
All dogs had received a complete physical and neurologic examination performed by a board-certified neurologist (MB) or by a neurology resident (SC and PC). In the medical records, neurologic grade had been assigned in accordance with a 6-point scoring system for both thoracolumbar18 and cervical19 assessment in dogs. However, to assess the severity of clinical signs as a potential risk factor for development of POD, dogs were simply classified as ambulatory or nonambulatory.
Surgical techniques
All spinal decompression surgeries had been performed by the same surgeon in an operating suite dedicated to clean surgical procedures. Surgery was performed either immediately after the diagnostic imaging procedure (during the same anesthetic event) or within 7 days after diagnostic imaging, depending on the severity of the dog's neurologic condition. Thoracolumbar IVDH was treated by performance of hemilaminectomy or minihemilaminectomy with a standard dorsolateral approach.18 Cervical IVDH was treated by performance of a standard ventral-slot procedure or hemilaminectomy.20,21 When hemilaminectomy (cervical or thoracolumbar) or thoracolumbar minihemilaminectomy was performed, the surgical site was covered with a gelatin patch. Cefazolin sodium (30 mg/kg [13.6 mg/lb], IV) was administered at the time of anesthetic induction or within 30 minutes before skin incision. Additional doses were administered every 90 minutes during surgery until the surgical procedure and skin suturing had been completed. Antimicrobial administration at the same dose was continued 3 times/d for 48 hours after surgery. For statistical purposes, surgeries were categorized into 2 groups on the basis of anatomic approach to the spine (dorsal vs ventral).
Anesthesia
Dogs were sedated prior to spinal surgery with methadone (0.2 to 0.3 mg/kg (0.09 to 0.14 mg/lb), IM; n = 295 [79.3%]) or dexmedetomidine (1 to 2 μg/kg [0.45 to 0.91 μg/lb], IM; 77 [20.7%]). All dogs were anesthetized with propofol (1 mg/kg [0.45 mg/lb], IV), and anesthesia was maintained with oxygen and either isoflurane or sevoflurane delivered via an endotracheal tube. During the surgical procedure, analgesia was provided with a constant rate infusion of fentanyl (5 to 10 μg/kg/h [2.3 to 4.5 μg/lb/h], IV) alone or in combination with lidocaine (1 to 3 mg/kg/h [0.45 to 1.36 mg/lb/h], IV). Body temperature was measured with an esophageal probe and recorded every 15 minutes during the surgical procedure, and mean esophageal temperature was calculated. Total duration of anesthesia (defined as the interval from anesthetic induction to endotracheal extubation) was recorded when available.
Diagnosis of POD
Medical records of dogs that had developed POD were retrieved and examined to confirm the diagnosis. Medical and diagnostic imaging procedures performed to diagnose POD were reviewed. Radiographic findings considered indicative of POD included loss of definition of end plate margins, narrowing of the intervertebral disk space, lytic bony changes of the vertebrae adjacent to the intervertebral disk space, and sclerosis at the margins of bone lysis.16,22,23 Magnetic resonance imaging criteria considered indicative of POD were multifocal areas of hyperintensity of vertebral end plates, diffuse hyperintensity of the intervertebral disk, and changes in signal intensity of paravertebral soft tissue on T2-weighted and short tau inversion recovery sequences, associated with contrast enhancement of vertebral end plates, intervertebral disk, and paravertebral soft tissue on postcontrast T1-weighted images.24 All radiographs and MRI scans were reviewed independently by 1 author (MB) and an experienced radiologist. Diagnostic procedures and treatments used as a result of POD, interval from spinal decompression surgery to POD diagnosis, identities of bacterial agents isolated, and outcome were also recorded.
Postoperative assessment
All dogs had been examined daily during hospitalization and reexamined 10 to 14 days after surgery, at the time of suture removal. Dogs were reexamined on a monthly basis for 3 months after surgery by the authors or, in some situations, the referring veterinarian who performed physical and neurologic examinations. Dogs were also reexamined at any other times if requested by the owners, authors, or referring veterinarians. Telephone interviews with owners or referring veterinarians had been conducted and recorded 2 and 3 months after surgery (long-term follow-up) for dogs that were not directly assessed by the authors.
Statistical analysis
Statistical analyses were performed with statistical software packages.a,b Contingency tables were generated for the categorical variables (breed, age, sex, body weight, preoperative neurologic status [ambulatory vs nonambulatory], type and location of IVDH, and type of surgical procedure performed) and for the categorically converted variables duration of anesthesia, esophageal temperature during surgery, and duration of hospitalization. Receiver operating characteristic curve analysis was performed to determine appropriate cutoff values to reclassify the continuous variables body weight and age as categorical. Distributions of factors were compared between dogs with and without POD via the χ2 test. Odds ratios and 95% CIs were calculated for variables. Factors identified as having a liberal association with POD (ie, P < 0.10) were used to perform multivariate logistic regression. Factors were considered significant when the value of P was < 0.05 and when the 95% CI of the OR excluded 1.0.
Results
Animals
A total of 372 dogs met the inclusion criteria and were used in the study. A high proportion were mixed breeds (n = 119 [32.0%]) or Dachshunds (87 [23.4%]). Other breeds included German Shepherd Dog (n = 24); Beagle (18); French Bulldog (17); Shih-Tzu (11); Pekingese, Dalmatian, and English Cocker Spaniel (9 each); Maltese (6); Schnauzer, Miniature Pinscher, Doberman Pinscher, and Pug (4 each); Boxer, English Setter, Basset Hound, Standard Poodle, and Bolognese (3 each); Brittany Spaniel, Jack Russell Terrier, Labrador Retriever, Italian Hound, Border Collie, English Bulldog, Papillon, and Chihuahua (2 each); and various other breeds (1 each). There were 149 females (93 spayed and 56 sexually intact; 40.1%) and 223 males (30 neutered and 193 sexually intact; 59.9%). Mean age was 7.3 years (range, 1.6 to 14.3 years), and mean body weight was 14.2 kg (31.2 lb; range, 2.1 to 62 kg [4.6 to 136.4 lb]).
Ninety-five (25.5%) dogs had cervical IVDH, and all underwent ventral slot surgery except for 4 dogs, which underwent hemilaminectomy. Two hundred seventy-seven (74.5%) dogs had IVDH caudal to T3 and underwent hemilaminectomy (n = 267) or minihemilaminectomy (10). Surgery for 91 (24.5%) dogs was performed with a ventral approach, and surgery for 281 (75.5%) dogs was performed with a dorsal approach. Forty-six dogs had disk protrusion, and 326 had disk extrusion. Total duration of anesthesia was available for 292 dogs, and median duration was 3 hours (range, 1 to 10 hours). Duration of hospitalization was available for 366 dogs, and median duration was 4 days (range, 1 to 21 days).
Dogs with POD
Eight (2.2%) dogs developed POD, including 3 German Shepherd Dogs and 1 each of Dalmatian, Border Collie, Papillon, Doberman Pinscher, and mixed breed. German Shepherd Dogs were 9.8 times as likely to develop POD as were other breeds (Table 1). Dogs with POD included 3 females (2 spayed and 1 sexually intact) and 5 males (1 neutered and 4 sexually intact), and this distribution was not significantly different from that for dogs without POD.
Results of univariate analysis to identify factors unconditionally associated with development of POD in 372 dogs following spinal decompression surgery for IVDH.
| Factor | Percentage of dogs with POD with factor | Percentage of dogs without POD with factor | OR | 95% CI | P value |
|---|---|---|---|---|---|
| Sex | |||||
| Male | 62.5 | 59.9 | 1.12 | 0.26–4.74 | 0.88 |
| Female | 37.5 | 40.1 | Referent | — | — |
| Age | |||||
| > 8.8 y | 87.5 | 29.0 | 16.81 | 2.04–138.32 | < 0.001 |
| ≤8.8 y | 12.5 | 71.0 | Referent | — | — |
| Body weight | |||||
| > 20 kg | 87.5 | 21.0 | 26.46 | 3.20–218.54 | < 0.001 |
| < 20 kg | 12.5 | 79.0 | Referent | — | — |
| Breed | |||||
| German Shepherd Dog | 37.5 | 5.5 | 9.80 | 2.19–43.82 | < 0.001 |
| Other breeds | 62.5 | 94.5 | Referent | — | — |
| Site of IVDH | |||||
| Caudal | 75.0 | 74.4 | 1.03 | 0.20–5.19 | 0.97 |
| Cervical | 25.0 | 25.6 | Referent | — | — |
| Type of IVDH | |||||
| Protrusion | 37.5 | 11.8 | 4.48 | 1.03–19.41 | 0.03 |
| Extrusion | 62.5 | 88.2 | Referent | — | — |
| Presurgery neurologic status | |||||
| Nonambulatory | 37.5 | 53.0 | 0.56 | 0.13–2.36 | 0.42 |
| Ambulatory | 62.5 | 47.0 | Referent | — | — |
| Surgical approach | |||||
| Dorsal | 75.0 | 75.6 | 0.97 | 0.19–4.90 | 0.97 |
| Ventral | 25.0 | 24.4 | Referent | — | — |
| Duration of anesthesia | |||||
| > 2 h | 80.0 | 78.0 | 1.13 | 0.12–10.25 | 0.92 |
| ≤ 2 h | 20.0 | 22.0 | Referent | — | — |
| Esophageal temperature during surgery | |||||
| > 38°C | 40.0 | 32.0 | 1.37 | 0.23–8.36 | 0.73 |
| < 38°C | 60.0 | 68.0 | Referent | — | — |
— = Not applicable.
Values of P < 0.05 and 95% CIs excluding 1 were considered significant. To convert temperature from Celsius to Fahrenheit, multiply by 9/5 and add 32.
Mean age of dogs with POD was 10.1 years (range, 5.9 to 12.5 years), and mean body weight was 25.2 kg (55.4 lb; range, 6 to 37 kg [13.2 to 81.4 lb]). Receiver operating characteristic curve analysis performed for age and body weight revealed optimal cutoff values of 8.8 years and 20 kg (44 lb), respectively, for creation of categorical variables, corresponding with an area under the curve of 0.79 (sensitivity, 87%; specificity, 71%) and 0.76 (sensitivity, 87%; specificity, 79%), respectively. When these cutoffs were used, dogs > 8.8 years or weighing > 20 kg were at significantly (P < 0.001) higher odds of developing POD than were other dogs (Table 1).
None of the dogs that developed POD had been treated with corticosteroid drugs prior to referral for IVDH, preventing the ability to evaluate any association between corticosteroid administration prior to surgery and POD. Regarding the type of IVDH, 3 of the 8 dogs with POD had disk protrusion (1 at the C1-C5 region and 2 at the T3-L3 region), compared with 34 of the 364 dogs without POD (P = 0.03; Table 1). Five dogs that developed POD had disk extrusion (1 at the C1-C5 region and 4 at the T3-L3 region). Dogs with disk protrusion were 4.5 times as likely to develop POD as were those with disk extrusion.
No association was identified between development of POD and any other variable evaluated through univariate analysis (Table 1). Multivariate analysis revealed that body weight > 20 kg was the only variable to maintain a significant (P = 0.04) association with development of POD, whereas age and type of IVDH were not found to be significant in this model (Table 2).
Results of multivariate logistic regression to identify factors significantly associated with POD for the dogs in Table 1, controlling for other factors.
| Factor | OR (95% CI) | P value |
|---|---|---|
| Age > 8.8 y (vs ≤ 8.8 y) | 6.52 (0.73–58.66) | 0.09 |
| Body weight > 20 kg (vs ≤ 20 kg) | 11.05 (1.12–108.62) | 0.04 |
| German Shepherd Dog (vs other breeds) | 1.58 (0.28–8.83) | 0.61 |
| Protrusion (vs extrusion) IVDH | 1.31 (0.25–6.94) | 0.75 |
Values of P < 0.05 and 95% CIs excluding I were considered significant.
Total duration of anesthesia was calculated for 5 dogs, yielding a median duration of 2.8 hours (range, 2.0 to 4.5 hours). Median duration of hospitalization was 5 days (range, 3 to 9 days). Clinical signs suggestive of infection developed a median of 29 days (range, 16 to 69 days) after surgery, and 5 dogs developed POD within 50 days after surgery. All POD diagnoses were made by the authors at recheck examinations.
Main clinical signs indicative of POD were recurrence of signs of spinal pain and reluctance to move, despite initial postsurgical improvement of the neurologic status and an apparent pain-free interval. Prior to diagnosis of POD, all dogs had received an NSAID alone or with other analgesic drugs. Surgical wounds of all dogs appeared clean, and signs of purulent infection such as erythema, drainage, or dehiscence were not documented for any dog.
Diagnosis of POD was achieved with survey radiographs alone (n = 4) or survey radiographs and MRI scans (4). Microbial culture of a blood sample was performed for 4 dogs, revealing Pseudomonas aeruginosa in samples from 2 dogs. For those 2 dogs, antimicrobial treatment was chosen on the basis of susceptibility testing. The remaining 6 dogs were empirically treated with broad-spectrum antimicrobials (cefazolin and enrofloxacin) and cage rest for a minimum of 6 weeks until clinical signs of infection had resolved and radiography revealed evidence of healing (ie, replacement of the previous lytic focus by bridging or fusion of the involved vertebrae).22 Radiography, performed a mean of 3 months after surgery, confirmed resolution of POD with bridging of the involved vertebrae visible for all 8 dogs.
Discussion
Postoperative diskospondylitis is a possible complication of spinal surgery. In human medicine, an incidence of POD between 0.1% and 3% has been reported, depending on the type of surgical procedure and anatomic site of the surgery.1,6 Conversely, only a few cases of iatrogenic diskospondylitis have been reported for dogs,16,17,25,26 and those reports are limited to a single case or sporadic cases included in retrospective studies of naturally occurring diskospondylitis. Procedures associated with development of diskospondylitis in dogs include epidural injections of analgesic, soft tissue and spinal surgery, and fenestration of intervertebral disks.16,17,25,26
To the authors’ knowledge, data on the incidence of and risk factors for POD after spinal decompression surgery in dogs are lacking. A retrospective study16 revealed an incidence of POD < 1% (4 dogs) after spinal surgery. However, the type of surgery performed and possible risk factors for POD were not reported. In the present study involving 372 dogs that underwent spinal decompression surgery, 8 (2.2%) developed POD.
The clinical picture of POD is usually nonspecific.3 In human medicine, the most common symptom is worsening pain recurring 1 to 4 weeks after a spinal procedure, and pain may or may not follow a period of initial relief.1 The incision is usually clean, and < 10% of surgical wounds have signs of purulent infection such as erythema, drainage, or dehiscence.1 Similarly, in the dogs of the present study, recurrent pain was the most prevalent sign and wound appearance was unremarkable for all dogs. People with POD can confuse this pain with a recurrence of their original back pain, thus mistaking the cause as mechanical.1 Similarly, owners of dogs surgically treated for IVDH with subsequent recurrence of signs of spinal pain may misinterpret this sign as a recurrence of the primary problem. This misinterpretation could lead to unwillingness of owners to return their dogs for examination and, consequently, may lead to underestimation of the incidence of POD,16 as may have occurred in the present retrospective study.
In human medicine, patients with POD usually develop symptoms in the first few weeks following surgery, after an initial period of improvement.5 In 1 study,3 symptom onset occurred a mean of 27.7 days after surgery and 90% of people with POD had symptoms within the first 50 days after surgery. The situation was similar in the present study, in which POD was diagnosed in dogs a median of 29 days after spinal decompression surgery, with 5 of 8 affected dogs developing POD by 50 days after surgery. On the basis of these data, we recommend strict monitoring of dogs for neurologic condition for the first few weeks after spinal decompression surgery. Recurrence of signs of spinal pain after an initial period of pain relief should prompt clinicians to suspect POD.
In both human and veterinary medicine, the diagnosis of diskospondylitis, both naturally occurring or postoperative forms, can be achieved by various imaging modalities such as survey radiography, CT, and MRI.5,24 Standard radiography is performed first during workup of a suspected infection because of its wide accessibility.1,5 Nonetheless, findings are often insidious, and recently affected patients may have equivocal or even unremarkable radiographic findings.5,17 Indeed, radiographic evidence of diskospondylitis may lag behind the onset of clinical signs by as much as 2 to 4 weeks in both humans and dogs.1,5,24
In veterinary medicine, CT has been used to guide collection of fine-needle aspirates and tissue-core biopsy specimens, but no reports exist of the sensitivity and specificity of this technique for diagnosis of diskospondylitis and monitoring of associated treatment.27 Although MRI is the modality of choice for diagnosis of diskospondylitis, this approach is expensive and may not be financially feasible for all dog owners. Owners of 4 dogs with POD in the present study declined offers to perform MRI; consequently, the diagnosis was achieved solely through radiography.
The pathogenesis of surgical site infections after spinal surgery is multifactorial, and understanding the contributing factors is important to prevention and management of this complication.1 Risk factors can be classified into 3 broad categories: patient or host factors, surgical factors, and microbiological factors.1 Some patient factors are modifiable whereas others are not. In humans, advanced age, immunosuppression, spinal trauma, and diabetes mellitus are some of the most important established risk factors and are nonmodifiable.1,5 In the present study, age > 8.8 years and body weight > 20 kg were 2 potential risk factors that were significant in univariate analysis. Likewise, age, breed, and consequently body weight were significantly associated with development of naturally occurring diskospondylitis in dogs of other studies.16,24
In a study16 involving 513 dogs with diskospondylitis, dogs at highest odds of disease were those > 10 years of age. Elderly dogs might be at increased risk for POD because of a higher possibility of comorbid disease, immunocompromised state, or impaired healing processes, compared with the possibility in younger dogs. Although age was no longer significantly (P = 0.09) associated with POD after other factors were controlled for in the present study, this lack of significance may have been attributable to low statistical power given the small number of dogs that developed POD. Studies with greater numbers of dogs with POD would be required to rule out age as a risk factor for the disease.
In humans, obesity is an important risk factor for POD and surgical site infection in general.15,9 This may be related to biophysical alterations caused by obesity and technical issues with surgery,1 given that tissue dissection in obese human patients can be extensive. Wide retraction and use of electrocautery can lead to fat necrosis and large devitalized areas promoting bacterial proliferation.1 Body condition of dogs in the present study had not been assessed; therefore, no conclusions could be made regarding whether that factor played a role in the development of POD. Indeed, the association between body weight > 20 kg and POD could have merely reflected higher odds for large-breed dogs (eg, German Shepherd Dogs), rather than for dogs with obesity. Consequently, more studies are needed to better elucidate the effect of body size and body condition on development of POD in dogs.
Univariate analysis in the present study also revealed that German Shepherd Dogs had almost 10 times the odds of developing POD, compared with the odds for all other breeds combined, although this factor was no longer significant (P = 0.603) after other variables were controlled for. Great Danes, Boxers, Rottweilers, English Bulldogs, German Shepherd Dogs, and Doberman Pinschers are among the breeds at highest risk of naturally occurring diskospondylitis.16,24 Previous trauma to the cartilaginous end plates, vertebral bodies, and intervertebral disks may contribute to diskospondylitis development.28 The predominance of diskospondylitis in large-breed dogs has been linked to possible stresses placed on the vertebral column from increased body weight and greater exercise activity, compared with results in other breeds.29 Moreover, the immune response (particularly cell-mediated) of German Shepherd Dogs is reportedly reduced, compared with the immune response of other breeds,30 and appears to play a fundamental role in predisposing German Shepherd Dogs to various infectious diseases such as pyoderma, mycosis, and bacterial disease.31,32 All of these reasons support a predisposition of German Shepherd Dogs to POD, even though multivariate analysis failed to identify a significant association between that breed and POD. Again, the low incidence of POD in the dogs of the present study could have limited the statistical power to detect associations that truly were significant. Additional studies with larger numbers of dogs with and without POD are needed to determine with better certainty whether age and breed are or are not risk factors for POD in dogs with IVDH undergoing spinal decompression surgery.
Because only 8 dogs developed POD in the present study, patients were broadly classified as to whether they were ambulatory or nonambulatory before surgery. However, this broad classification and small sample size precluded investigation of the effect of IVDH severity or specific clinical signs on the outcome. Additional studies are warranted to investigate preoperative neurologic status as a potential risk factor for development of POD. Univariate analysis revealed that dogs with protrusion- versus extrusion-type herniation had approximately 4.5 times the odds of developing POD. However, only 3 of 8 dogs with POD had protrusion and this factor was not significant (P = 0.751) after controlling for other factors. Additional studies with larger numbers of dogs with POD are needed to confirm or refute this finding. Although we were unable to identify any comparable data in the human medical literature, it is possible that intervertebral disk protrusions increase the odds of POD in dogs on the basis of chronicity of the lesions. Chronic compression might lead to prolonged ischemia of the spinal cord and surrounding tissue, resulting in a decrease in resistance to bacterial colonization of the surgical site after surgery.
Extensive emphasis is given to types of surgery and surgical approaches in the human literature regarding POD. Spinal procedures involving a posterior approach are associated with a higher risk of infection than are those involving an anterior approach.5 The lower infection rate for the anterior versus posterior approach is likely related to greater vascularity of the anterior portion of the spine and lesser extent of muscle dissection needed to achieve bone exposure during spinal surgery.33 Extensive soft tissue dissection, prolonged surgical duration, soft tissue devitalization, and dead space creation are important surgical risk factors for POD in humans.5 In the present study, no difference in the odds of developing POD was detected between ventral and dorsal approaches to spinal decompression surgery. Use of the fenestration technique is reportedly a risk factor for POD in dogs.16,17 Type of surgical procedure could not be evaluated for an association with POD in the present study because only dogs with certain types of surgical procedures were included.
Identification of causative organisms for diskospondylitis can be a challenging task. In both human and veterinary medicine, the combination of bacteriologic culture of urine, blood, and material collected from the affected intervertebral disk space is believed to provide the best chance for this purpose.24,34 Additionally, CT or fluoroscopic guidance can be used to obtain deep samples from the affected area for microbiological and histologic assessment.33 Bacterial culture of both urine and blood samples has an overall success rate that ranges widely (between 30% and 78%) for dogs with naturally occurring diskospondylitis.24 Pathogens most commonly recovered from affected dogs are coagulase-positive staphylococci (Staphylococcus intermedius and Staphylococcus aureus). Other bacterial organisms recovered include Escherichia coli, Corynebacterium spp, Streptococcus spp, Pseudomonas spp, and Proteus spp.16,24
In the present study, bacteria (Pseudomonas aeruginosa) were recovered for 2 out of 4 dogs with POD for which bacterial culture was performed. In human medicine, gram-positive cocci (S aureus, Staphylococcus epidermidis, and β-hemolytic streptococci) are the most common bacteria responsible for postoperative infections.1,5 However, among gram-negative infections, Pseudomonas aeruginosa is one of the most common organisms recovered.15
No evidence-based guidelines exist for the duration of antimicrobial treatment required to achieve satisfactory resolution of diskospondylitis.35 Until such guidelines become available, recommendations are to continue treatment until complete resolution of signs of vertebral column pain is apparent on deep palpation and no radiographic evidence remains of active disease.16,22,35 Radiographic evidence of resolution includes disappearance of the lytic focus, smoothing, and loss of the sclerotic margins formed around the lytic focus as well as replacement by bridging of the involved vertebrae.16,22,35 In the study reported here, the decision to stop antimicrobial treatment was based on clinical signs and radiographic findings.
A major limitation of the present study was its retrospective design and reliance on the availability and accuracy of the medical records. Despite this, most required data were retrieved without a substantial amount of missing information. A 3-month follow-up period was arbitrarily chosen after the medical records had already been reviewed. Although investigators in a previous study27 found that POD developed in dogs within 1 month following spinal surgery, we could not exclude the possibility that a longer follow-up period in the present study might have led to identification of more dogs with POD. Duration of surgery and anesthesia are 2 risk factors associated with surgical site infection in humans and small animals.1,8,11,36 Again, failure to detect such associations in the present study could have been attributable to the small sample size or could have reflected the lack of an influence of these factors on development of POD.
A final point for consideration is that the antimicrobial prophylaxis protocol used in the present study may have affected the results. The choice to administer cefazolin for 48 hours after spinal decompression surgery was made because a venous catheter had been left in place for that period to allow postoperative provision of analgesia and fluids. In veterinary medicine, protocols for antimicrobial administration after spinal decompression surgery are inconsistent, and overall guidelines for antimicrobial prophylaxis do not exist. Antimicrobial administration is typically discontinued immediately after spinal surgery or is provided for a highly variable duration (0.5 to 14 days).37–40 Nevertheless, perioperative cefazolin administration did not result in any multidrug-resistant bacterial infections in the present study, and all dogs with POD had resolution of their infection without complications.
Knowledge of risk factors and accurate, standardized methods for monitoring dogs for development of POD after spinal surgery are important to ensure the best possible outcome. Understanding the factors that place a patient at risk for developing POD can help clinicians and surgeons identify those for which additional preventive measures might be beneficial. On the basis of the findings reported here, we recommend strict monitoring of dogs after spinal decompression surgery, particularly those weighing > 20 kg, for signs of recurrence of spinal pain after an initial improvement of neurologic status. Dogs with such signs should be strongly suspected of having POD and diagnostic testing and treatment pursued accordingly.
Acknowledgments
The authors thank Dr. Monica Alberti for assistance with review of all diagnostic images.
ABBREVIATIONS
| CI | Confidence interval |
| IVDH | Intervertebral disk herniation |
| POD | Postoperative diskospondylitis |
| POS | Postoperative spondylodiskitis |
Footnotes
SAS, version 9.3, SAS Institute Inc, Cary, NC.
MedCalc, version 12.4.0, MedCalc Software, Mariakerke, Belgium.
References
1. Chaudhary SB, Vives MJ, Basra SK, et al.Postoperative spinal wound infections and postprocedural diskitis. J Spinal Cord Med 2007, 30:441–451.
2. Beiner JM, Grauer J, Kwon BK, et al.Postoperative wound infections of the spine. Neurosurg Focus 2003, 15:1–4.
3. Jiménez-Mejías ME, de Dios Colmenero J, Sánchez-Lora FJ, et al.Postoperative spondylodiskitis: etiology, clinical findings, prognosis, and comparison with nonoperative pyogenic spondylodiskitis. Clin Infect Dis 1999, 29:339–345.
4. Weinstein MA, McCabe JP, Cammisa P. Postoperative spinal wound infection: a review of 2,391 consecutive index procedures. J Spinal Disord 2000, 13:422–426.
5. Nasto LA, Colangelo D, Rossi B, et al.Post-operative spondylodiscitis. Eur Rev Med Pharmacol Sci 2012, 16:50–57.
6. Van Goethem JWM, Parizel PM, Van Den Hauwe L, et al.The value of MRI in the diagnosis of postoperative spondylodiskitis. Neuroradiology 2000, 42:580–585.
7. Hernández-Palazon J, Puertas-García JP, Martinez-Lage JF, et al.Lumbar spondylodiscitis caused by Propionibacterium acnes after epidural obstetric analgesia. Anesth Analg 2003, 96:1486–1488.
8. Pull ter Gunne AF, Cohen DB. Incidence, prevalence, and analysis of risk factors for surgical site infection following adult spinal surgery. Spine 2009, 34:1422–1428.
9. Patel N, Bagan B, Vadera S, et al.Obesity and spine surgery: relation to perioperative complications. J Neurosurg Spine 2007, 6:291–297.
10. Fang A, Hu SS, Endres N, et al.Risk factors for infection after spinal surgery. Spine 2005, 30:1460–1465.
11. Olsen MA, Mayfield J, Lauryssen C, et al.Risk factors for surgical site infection in spinal surgery. J Neurosurg Spine 2003, 16:576–579.
12. Klein JD, Hey LA, Yu CS, et al.Perioperative nutrition and postoperative complications in patients undergoing spinal surgery. Spine 1996, 21:2676–2682.
13. Simpson JM, Silveri CP, Balderston RA, et al.The result of operations on the lumbar spine in patients who have diabetes mellitus. J Bone Joint Surg Am 1993, 75:1823–1829.
14. Foster JD, Trepanier LA, Ginn JA. Use of linezolid to treat MRSP bacteriemia and discospondylitis in a dog. J Am Anim Hosp Assoc 2014, 50:53–58.
15. Plessas IN, Jull P, Volk HA. A case of canine discospondylitis and epidural empyema due to Salmonella species. Can Vet J 2013, 54:595–598.
16. Burkert BA, Kerwin SC, Hosgood GL, et al.Signalment and clinical features of diskospondylitis in dogs: 513 cases (1980–2001). J Am Vet Med Assoc 2005, 227:268–275.
17. Schwartz M, Boettcher IC, Kramer S, et al.Two dogs with iatrogenic discospondylitis caused by methicillin-resistant Staphylococcus aureus. J Small Anim Pract 2009, 50:201–205.
18. Sharp NJ, Wheeler SJ. Thoracolumbar disc disease. In: Sharp NJ, Wheeler SJ, eds. Small animal spinal disorders: diagnosis and surgery. 2nd ed. Philadelphia: Elsevier Mosby, 2005;121–159.
19. De Risio L, Munana K, Murray M, et al.Dorsal laminectomy for caudal cervical spondylomyelopathy: postoperative recovery and long-term follow-up in 20 dogs. Vet Surg 2002, 31:418–427.
20. Sharp NJH, Wheeler SJ. Cervical disc disease. In: Sharp NJ, Wheeler SJ, eds. Small animal spinal disorders:diagnosis and surgery. 2nd ed. Philadelphia: Elsevier Mosby, 2005;106–116.
21. Sharp NJH, Wheeler SJ. Neoplasia. In: Sharp NJ, Wheeler SJ, eds. Small animal spinal disorders: diagnosis and surgery. 2nd ed. Philadelphia: Elsevier Mosby, 2005;266–270.
22. Shamir MH, Tavor N, Aizenberg T. Radiographic findings during recovery from discospondylitis. Vet Radiol Ultrasound 2001, 42:496–503.
23. Smith AS, Blaser SI. Infectious and inflammatory processes of the spine. Radiol Clin North Am 1991, 29:809–827.
24. Harris JM, Chen AV, Tucker RL, et al.Clinical features and magnetic resonance imaging characteristics of diskospondylitis in dogs: 23 cases (1997–2010). J Am Vet Med Assoc 2013, 242:359–365.
25. Finnen A, Blond L, Parent J. Cervical discospondylitis in 2 Great Dane puppies following routine surgery. Can Vet J 2012, 53:531–534.
26. Remedios AM, Wagner R, Caulkett NA, et al.Epidural abscess and discospondylitis in a dog after administration of a lumbosacral epidural analgesic. Can Vet J 1996, 37:106–107.
27. Gonzalo-Orden JM, Altonaga JR, Orden MA, et al.Magnetic resonance, computed tomographic and radiologic findings in a dog with discospondylitis. Vet Radiol Ultrasound 2000, 41:142–144.
28. Moore MP. Discospondylitis. Vet Clin North Am Small Anim Pract 1992, 22:1027–1034.
29. Kornegay JN. Diskospondylitis. In: Kirk WR, ed. Current veterinary therapy IX small animal practice. Philadelphia: WB Saunders Co, 1986;810–814.
30. Faldyna M, Levà L, Knotigovà P, et al.Lymphocyte subsets in peripheral blood of dogs—a flow cytometric study. Vet Immunol Immunopathol 2001, 82:23–37.
31. Harrus S, Waner T, Neer TM, et al.Ehrlichia and Anaplasma infections. In: Greene C, ed. Infectious diseases of the dog and cat. 4th ed. Philadelphia: Elsevier Saunders, 2012;227–259.
32. Rosser EJ. German Shepherd Dog pyoderma. Vet Clin North Am Small Anim Pract 2006, 36:203–211.
33. Petherick ES, Dalton JE, Moore PJ, et al.Methods for identifying surgical wound infection after discharge from hospital: a systematic review. BMC Infect Dis 2006, 6:170.
34. Duarte RM, Vaccaro RA. Spinal infection: state of the art and management algorithm. Eur Spine J 2013, 22:2787–2799.
35. Kerwin S. Discospondylitis and related spinal infections in the dog and cat. In: Fingeroth JM, Thomas WB, eds. Advances in intervertebral disc disease in dogs and cats. Ames, Iowa: Wiley Blackwell, 2015;161–167.
36. Nelson LL. Surgical site infections in small animal surgery. Vet Clin North Am Small Anim 2011, 41:1041–1056.
37. Konar M, Lang J, Fluhmann G, et al.Ventral intraspinal cysts associated with the intervertebral disc: magnetic resonance imaging observations in seven dogs. Vet Surg 2008, 37:94–101.
38. Ingram EA, Kale DC, Balfour RJ. Hemilaminectomy for thoracolumbar Hansen type I intervertebral disk disease in ambulatory dogs with or without neurologic deficits: 39 cases (2008–2010). Vet Surg 2013, 42:924–931.
39. Suwankong N, Meij BP, Van Klaveren NJ, et al.Assessment of decompressive surgery in dogs with degenerative lumbosacral stenosis using force plate analysis and questionnaires. Vet Surg 2007, 36:423–431.
40. Naudé SH, Lambrechts NE, Wagner WM, et al.Association of preoperative magnetic resonance imaging findings with surgical features in Dachshunds with thoracolumbar intervertebral disk extrusion. J Am Vet Med Assoc 2008, 232:702–708.
