Postoperative Complications Following TECA-LBO in the Dog and Cat

Introduction

Total ear canal ablation combined with lateral bulla osteotomy (TECA-LBO) is a common surgical treatment for end-stage otitis externa and to remove ear canal masses in both dogs and cats. Many complications from those procedures have been documented in the veterinary literature, such as facial nerve paresis, Horner’s syndrome, persistent drainage, fistula formation, hemorrhage, vestibular disease, hypoglossal nerve dysfunction, and infection. Total complication rates range from 29% to 82%.^1–6 Despite the potential for morbidity with these procedures, TECA-LBO is highly effective in eradicating ear disease, providing pain relief, and treating advanced otitis that is no longer responsive to medical management. Previous retrospective studies report the incidence of postoperative facial nerve paresis as 13–46% in dogs and 56% in cats.^1–6 Temporary facial nerve deficits usually resolve within 2 wk postoperatively, but lasted up to 11 mo in one dog.³ Residual deficits in facial nerve function were reported in 4–23% of dogs and 28% in cats.^1–4

The purpose of this retrospective, multicenter study was to review TECA-LBO records, quantify the incidence of complications, and to identify potential risk factors that may underlie any disparity in clinical outcomes. It was hypothesized that the duration of preoperative ear disease would be associated with the incidence of postoperative facial nerve deficits and Horner’s syndrome.

Materials and Methods

Medical records of dogs and cats that underwent TECA-LBO surgeries between January 2004 and March 2010 at the Purdue University Veterinary Teaching Hospital, the VCA Veterinary Specialty Center, and the VCA Indiana Veterinary Specialists and Animal Emergency Center were reviewed. All procedures were performed by either a Diplomate of the American College of Veterinary Surgeons (DACVS) or a surgical resident supervised by a DACVS. All information pertaining to TECA-LBO surgery and the first 2 wk postoperatively was collected from the medical records, including signalment, history, clinical signs, location of otic disease (i.e., right, left, bilateral), previous treatments, preoperative diagnostic work-up (including radiographs, computed tomography (CT), bacterial culture and sensitivity, histopathology), and whether the primary surgeon was a DACVS or a surgical resident. If not available in the medical records, duration of preoperative clinical signs, reported in months, was determined by telephone follow-up conversation with the referring veterinarian.

TECA-LBO and curettage of the epithelial lining was performed routinely in each affected ear.⁷ There was variation in anesthetic protocols between institutions; however, all patients were administered injectable opioids both intraoperatively and perioperatively. Cefazolin (22 mg/kg IV) was administered q 2 hr intraoperatively and q 6–8 hr postoperatively for 12–24 hr, until the patients were transitioned to oral antibiotics. Lubricant ointment/drops were placed in the ipsilateral eye q 6–8 hr until a complete palpebral reflex returned. Use of either a Penrose or closed-suction drain was based on the surgeon’s preference. The type of drain placed was not consistently annotated in the records, so merely the presence or absence of a drain was recorded.

Postoperative medications and postoperative complications, such as facial nerve paresis, Horner’s syndrome, vestibular signs, or the development of a draining tract, were also documented. All biopsy samples, as well as bacterial cultures, were sent to a clinical pathology laboratory for analysis by a board-certified veterinary pathologist. Only one ear with a history of a lateral ear canal resection was included. Ears with evidence of lysis of the tympanic bulla on either radiographs or CT were included in the study. Exclusion criteria for the study included gross evidence of disease extending outside of the ear canal on CT scan, patients in which surgery was not recommended, or patients with a history of TECA without LBO.

Postoperative neurologic assessments, including cranial nerve examinations, were recorded in the medical record during the first 12–24 hr following TECA-LBO. Physical exam findings were listed for the following 2–3 days, depending on the duration of hospitalization, and again approximately 14 days postoperatively when the patients were examined at the time of suture removal.

Paralysis was defined as a loss of motor function due to a neural mechanism, and paresis was defined as incomplete paralysis.⁸ Facial nerve injury was graded using the following scale: 0 was defined as no deficits, 1 was defined as paresis, and 2 was defined as paralysis. Either an absent or diminished palpebral reflex was the primary physical exam finding used to gauge facial nerve injury. Deficits were considered residual if they persisted for a minimum of 1 yr from the first postoperative neurologic assessment.⁹

Other variables examined as potential risk factors for the development of residual facial nerve injury included age and weight of the patient at the time of surgery, a resident performing the surgery, and the concurrent histopathologic finding of malignant neoplasia.

A telephone interview was performed to further characterize the postoperative complications and client satisfaction with the outcome of the surgical procedures. Several techniques were used to minimize recall bias. The same interviewer contacted all of the clients to aid in a systematic method of questioning. The questions were closed-ended, and asked in the same order to each client. Duration of complications was rounded to the nearest wk because the client was not expected to remember the duration of complications in days. The interviewer collected information from the clients without disclosing the hypothesis of the study until the end of the interview. Additional follow-up information was obtained from the records of the referring veterinarians.

Telephone interviews were conducted a minimum of 6 wk postoperatively. If there was any discrepancy between owner recall and the medical records, the information in the medical records took precedence. The owners were asked to comment on the duration of facial nerve dysfunction, corresponding to either a diminished or absent palpebral reflex. The owners were also asked to comment on the development of other postoperative complications, such as Horner’s syndrome, vestibular signs, pain, hearing loss, and draining tracts. The presence of Horner’s syndrome was established if the owners were able to appreciate a reduced pupil size, drooping of the upper eyelid, enophthalmos, and/or prominence of the third eyelid. Nystagmus was described to the owners as abnormal, rhythmic movement of the eyes in which they move slowly in one direction and quickly back in the other direction. Hearing loss was graded subjectively by clients where 0 was defined as apparent deafness, 1 was defined as diminished hearing, and 2 was defined as apparently normal hearing. If appropriate, the date and cause of death were noted. Client satisfaction with the procedure was also rated. A score of 2 indicated the owner was completely satisfied, a score of 1 indicated partially satisfied, and a score of 0 indicated dissatisfied.

Results

During the study period, 121 TECA-LBOs were performed on 82 dogs and an additional 12 TECA-LBOs were performed on 11 cats. Fifty-nine of the medical records (44.4%) were reviewed from one university veterinary teaching hospital, and 74 of the medical records (55.6%) were reviewed from two private practice veterinary referral centers. The mean postoperative follow-up time was 3.5 ± 1.81 yr (median, 3 yr; range, 6 wk to 6 yr). Two patients (2.2%) had follow-up times < 2 mo, and 21 patients (22.6%) had follow-up times < 1 yr. Of the 93 patients undergoing TECA-LBO, 39 patients (41.9%) had bilateral surgeries, including 38 dogs and 1 cat. Ten dogs had unstaged, bilateral TECA-LBOs. Twenty-eight patients (30.1%) had right-sided surgeries, including 23 dogs and 5 cats. Twenty-six patients (28%) had left-sided surgeries, including 21 dogs and 5 cats. Of the 82 dogs, 37 (45.1%) were spayed females, 37 (45.1%) were castrated males, 4 (4.9%) were females, and 4 (4.9%) were males. Of the 11 cats, 4 (36.4%) were spayed females and 7 (63.6%) were castrated males. Mean age of the dogs that underwent surgery was 8.27 ± 2.85 yr (median, 8 yr; range, 1.4–14.5 yr), and mean age of the cats was 8.88 ± 3.81 yr (median, 9 yr; range, 3–16 yr). Mean body weight of the dogs at the time of surgery was 20.7 ± 13.2 kg (median, 15.5 kg; range, 3.6–59.1 kg), and mean body weight of the cats was 6.7 ± 2.4 kg (median, 6.8 kg; range, 4.5–10 kg).

Twenty-five dog breeds were represented among the 82 canine patients (88.2%) included in the study. Those breeds included the cocker spaniel (n = 38; 46.3%); mixed-breed (n = 12; 14.6%); Labrador retriever (n = 4; 4.9%); Chinese shar pei (n = 3; 3.7%); 2 each of the following breeds (2.4%): German shepherd dog, golden retriever, and Pomeranian; and 1 each of the following breeds (1.2%): Airedale terrier, beagle, border collie, Boston terrier, Chesapeake Bay retriever, Chihuahua, chow chow, dachshund, English bulldog, Great Dane, Great Pyrenees, Maltese, miniature poodle, pug, Shetland sheepdog, shih tzu, standard poodle, West Highland white terrier, and Yorkshire terrier. For the 11 cats, 8 (72.7%) were domestic shorthairs, and the remaining 3 cats included 1 domestic mediumhair, 1 domestic longhair, and 1 Siamese.

The primary complaint for 106 of the 133 included ears (79.7%) was chronic otitis. Twenty-seven of the 133 ears (20.3%) were referred for evaluation of a mass noted via either an otoscopic exam or direct visualization by the referring veterinarian. Two dogs had different types of pathology in each ear canal. The first dog had a ceruminous gland adenoma in the right ear and a ceruminous gland adenocarcinoma in the left ear. The second dog had squamous cell carcinoma in the right ear and otitis externa in the left ear. Of the included canine patients, 101 of the 121 ears (83.5%) presented for chronic otitis, and 20 ears (16.5%) presented for a mass in the ear canal. Of the included feline patients, 5 of the 12 ears (41.7%) presented for chronic otitis, and 7 ears (58.3%) presented for a mass.

Mean duration of preoperative clinical signs in all patients was 3.36 ± 2.79 yr (median, 2.75 yr; range, 1 wk to 11 yr). In dogs, the mean duration of preoperative clinical signs was 3.67 ± 2.77 yr (median, 3 yr; range, 1 wk to 11 yr). In cats, the mean duration of preoperative clinical signs was 0.59 ± 0.41 yr (median, 8 mo; range, 5 wk to 1 yr). Cats had a significantly shorter duration of preoperative clinical signs (P < 0.001). Historical findings and clinical signs recorded included head shaking, discharge from the ear canal, malodor, pain, scratching, aural hematoma, mineralization, ear canal stenosis, and previous bacterial cultures (when culture results were available). Preoperative neurologic signs included ipsilateral head tilt (n = 4 ears; 3%), ipsilateral facial nerve paresis (n = 2 ears; 1.5%), and horizontal nystagmus (n = 2 ears; 1.5%). Both of the ears that were associated with preoperative facial nerve paresis were in cocker spaniels. One of those dogs had severe otitis externa, and the other had a ceruminous gland adenocarcinoma and severe otitis. Both of those cocker spaniels had a similar degree of facial nerve paresis both pre- and postoperatively, and both had residual deficits. Two dogs had preoperative draining tracts. The underlying causes for the draining tracts were chronic otitis in one case and perforation of the auricular cartilage in the second case.

There was data pooled for TECA-LBOs performed in dogs and cats that were performed by 18 surgeons. Seven different histopathologic findings were noted. Patient age at the time of surgery, underlying disease process, surgeon, and perioperative complications were independently noted for each ear in bilateral procedures. The surgical technique for TECA-LBO was considered to be standardized between ears, regardless of the above-mentioned factors. Either a Penrose or closed-suction drain was used in 82 of the ears (61.2%). A drain was placed if there was debris/fluid encountered in the horizontal ear canal and osseous bulla intraoperatively.

The majority (91.7%) of the surgery reports described identification and protection of the facial nerve intraoperatively. Seven reports did not include that level of detail. Two reports described partial sacrifice of the facial nerve, either due to adhesions or proximity to a mass. Both of the dogs that had partial sacrifice had facial nerve paresis for 2 days postoperatively. One report indicated that the facial nerve was not visualized due to scar tissue in one dog, but no facial nerve deficits were subsequently noted. One report indicated complete sacrifice of the facial nerve in one cat, resulting in residual facial nerve paralysis.

Biopsy specimens were obtained from 99 of 133 ear canals (73.7%) as described in Table 1. Some surgeons elected not to submit a biopsy if there was a strong clinical suspicion for chronic otitis. Chronic otitis was the most common histopathologic finding, present in 69 of 99 canine and feline ear canals biopsied (69.7%). Ceruminous gland carcinoma/adenocarcinoma (CGA) was identified in 8 of 99 ears (8.1%), adenoma in 8 (8.1%), squamous cell carcinoma in 7 (7.1%), and inflammatory polyp in 5 (5.1%). In addition, 1 (1%) plasmacytoma and 1 (1%) neurofibrosarcoma were diagnosed. The overall rate of malignant neoplasms was 16.1%, and the rate of benign neoplasms was 9.1% when canine and feline biopsy samples were combined. Cats (75%) were significantly more likely than dogs (16.6%) to have a mass (either benign or malignant) in the ear canal (P < 0.001). In cats, 3 of 12 ears (25%) had histopathologic findings consistent with malignant neoplasia. In dogs, 13 of 87 ears biopsied (14.9%) had malignant neoplasms. Although the percentage of malignant neoplasia was higher in cats, this difference was not significant (P = 0.405).

Table 1 Histopathologic Diagnoses for Ear Canals Submitted After TECA-LBO in 87 Dog Ears and 12 Cat Ears

TECA-LBO, total ear canal ablation combined with lateral bulla osteotomy; CGA, ceruminous gland carcinoma/adenocarcinoma.

View Fullscreen

Samples for aerobic and anaerobic cultures of the tympanic bulla were obtained in 127 of 133 ears (95.5%). Staphylococcus spp. were the most common bacterial isolates, cultured in 55 ears (43.3%). Of the 16 cultures positive for Staphylococcus aureus, 16 (18.8%) were methicillin-resistant. Only 1 of 21 cases of Staphylococcus intermedius/Staphylococcus pseudintermedius (4.8%) was methicillin-resistant. Other common bacterial isolates included Enterococcus spp., Pseudomonas aeruginosa, Streptococcus spp., Escherichia coli, and Proteus mirabilis. There was no bacterial growth in 13 of 127 ears (10.2%). Seven of those 13 patients had biopsies consistent with otitis, 3 had malignant masses, 2 had benign masses, and 1 had no biopsy due to a clinical suspicion for otitis.

There were 10 TECA-LBOs for which the preoperative bacterial culture from the ipsilateral ear canals were available for comparison. In 4 of those 10 ears, the same bacteria were cultured both pre- and intraoperatively. The most common preoperative bacterial isolate was Pseudomonas aeruginosa (isolated in 5 of 10 ears).

All patients were discharged from the hospital with instructions to administer oral antibiotics and an oral analgesic/anti-inflammatory drug. Amoxicillin trihydrate/clavulanate potassium was prescribed following 48 surgeries, cephalexin was prescribed following 45 surgeries, and enrofloxacin was prescribed following 28 surgeries. Other antibiotics used postoperatively were marbofloxacin, orbifloxacin, metronidazole, chloramphenicol, trimethoprim/sulfonamide, cefpodoxime, and cefadroxil. Antibiotics were prescribed for a minimum of 1 wk postoperatively, pending the intraoperative bacterial culture results. If the bacterial culture was positive and the bacteria were sensitive to the empiric antibiotic choice, an additional 2–3 wk of that antibiotic was prescribed. This was documented in 23 of 127 cultures (18.1%). If the bacterial culture was positive and the bacteria were resistant to the empiric antibiotic choice, an appropriate antibiotic was prescribed for 3–4 wk based on the culture and sensitivity results. This was documented in 17 of 127 cultures (13.4%). Carprofen was prescribed following 59 surgeries, and meloxicam and deracoxib were each prescribed following 4 surgeries. Prednisone was prescribed following two surgeries. Tramadol was prescribed following 71 surgeries, and oral buprenorphine was prescribed following 4 surgeries. A transdermal fentanyl patch was applied in 1 patient.

All complications were ipsilateral, associated with the operated ear only. Sixty-eight TECA-LBOs (51.1%) had no facial nerve deficits, 36 (27.1%) had facial nerve paresis, and 29 (21.8%) had facial nerve paralysis. Of the TECA-LBOs performed on dogs, 64 of 121 ears (52.9%) had no postoperative facial nerve deficits, 33 (27.3%) had facial nerve paresis, and 24 (19.8%) had facial nerve paralysis. Of the TECA-LBOs performed on cats, 4 of 12 ears (33.3%) had no postoperative facial nerve deficits, 3(25%) had facial nerve paresis, and 5 (41.7%) had facial nerve paralysis. Although cats had a higher incidence of postoperative facial nerve deficits than dogs (66.7% versus 49.6%, respectively), that difference was not statistically significant (P = 0.259).

The presence and duration of facial nerve injury was documented in the medical records for a minimum of the first 2 wk postoperatively (Figure 1). Median duration of temporary facial nerve deficits in dogs was 2 wk (range, 2 days to 13 wk). Median duration of temporary facial nerve deficits in cats was 4 wk (range, 2–10 wk). A total of 14 of 133 ears (10.5%) had residual facial nerve deficits, including 4 of 12 feline ears (33.3%) and 10 of 121 canine ears (8.3%). The minimum follow-up period required for deficits to be considered residual was 1 yr.⁹ TECA-LBOs performed in cats had a significantly higher incidence of residual deficits than those performed in dogs (P = 0.02).

View Figure

• Download as Powerpoint
• Download Figure

No association was detected between the severity of facial nerve injury assessed 12–24 hr postoperatively and the duration of those clinical signs in the overall population studied. In cats, there was a significant association between complete facial nerve paralysis postoperatively and residual facial nerve deficits (P = 0.047). Age and weight of the patient at the time of surgery, a resident performing the surgery, and the concurrent histopathologic finding of malignant neoplasia were not significant risk factors in the development of residual facial nerve injury.

Preoperative duration of clinical signs was not significantly different in patients with or without postoperative facial nerve deficits. No significant association was seen between duration of preoperative clinical signs and postoperative facial nerve deficits in either cats (P = 0.216) or dogs (P = 0.564). No significant association was seen between duration of preoperative clinical signs and residual facial nerve deficits in either cats (P = 0.216) or dogs (P = 0.366).

Horner’s syndrome was reported in 11 of 133 TECA-LBOs (8.2%) postoperatively. None of those patients had preoperative Horner’s syndrome. In dogs, Horner’s syndrome was seen after 4 of 121 TECA-LBOs (3.3%), whereas in cats, Horner’s syndrome was seen after 7 of 12 TECA-LBOs (58.3%). The incidence of Horner’s syndrome in cats was significantly greater than in dogs (P < 0.001). In addition, residual signs of Horner’s syndrome were only seen in cats (25%). The median duration of temporary Horner’s syndrome was 2 wk (range, 2 days to 4 wk). The median duration of preoperative clinical signs in cats with postoperative Horner’s syndrome was 10 mo (range, 5 wk to 1 yr). This was significantly longer (P = 0.035) than the median duration of preoperative clinical signs in cats with no postoperative Horner’s syndrome, which was 2 mo (range, 5 wk to 8 mo). In dogs, there was no significant difference in duration of preoperative clinical signs between patients that did or did not have postoperative Horner’s syndrome (P = 0.773).

Postoperative ipsilateral head tilt was reported in 15 of 133 TECA-LBOs (11.3%). Postoperative head tilt was transient in 8 of 15 patients (53.3%), 6 of which lasted ≤ 1 wk. One dog had a head tilt for 7 mo postoperatively. Median duration of postoperative head tilt was 2 mo (range, 1 day to permanent). Horizontal nystagmus was seen immediately postoperatively in 6 of 133 TECA-LBOs (4.5%). All patients with postoperative nystagmus also had a head tilt. All episodes of nystagmus were transient, lasting ≤ 1 wk.

General complications included incisional problems and exposure keratitis. Minor incisional complications, such as discharge or partial dehiscence, occurred in 7 of 133 ears (5.3%). Complications included 2 pinnal infections, 2 incisional infections, 1 incident of cellulitis, 1 incident of hemorrhagic incisional discharge 4 days postoperatively, and 1 incident of partial incisional dehiscence.

As part of the telephone questionnaire, clients were asked to comment on patients’ hearing following TECA-LBO. Preoperative hearing assessments were neither performed nor recorded in this cohort of patients for comparison with postoperative hearing client assessments. Only 2 of 77 clients (2.6%) reported apparently normal hearing in their pets postoperatively. Thirty-eight clients (49.4%) reported partial hearing, and 37 (48.1%) reported deafness postoperatively. Of the 45 animals undergoing a unilateral TECA-LBO, 27 (60%) were reported to have no hearing, 16 (35.6%) were reported to have partial hearing, and 2 (4.4%) were reported to have no change in hearing by their owners. Of the 32 patients undergoing bilateral TECA-LBO procedures, 10 (31.3%) were reported to have no hearing, and 22 (68.8%) were reported to have partial hearing. In dogs, 2 (2.9%) were reported to have normal hearing, 35 (51.5%) were reported to have partial hearing, and 31 (45.6%) were reported to have no hearing. In cats, 3 (33.3%) were reported to have partial hearing, and 6 (66.7%) were reported to have no hearing. When clients reported partial hearing, they commented that the pets heard high-pitched, deep-pitched, or loud noises, such as whistling, clapping, or stomping. Three clients with dogs that had unilateral TECA-LBOs mentioned that the pet seemed to have difficulty localizing sound postoperatively. Those clients reported their pets as having partial hearing postoperatively.

Results of the telephone questionnaire indicated that 69 of the 77 clients (89.6%) were completely satisfied with the procedure. Only 7 clients (9.1%) were partially satisfied with the procedure, and 1 client (1.3%) was dissatisfied with the procedure. One client was not contacted because the pet had died of cardiac arrest 4 days postoperatively, before discharge from the hospital; thus, all relevant information for that case was attained through the medical record. In total, 77 of 93 clients (82.8%) were available for follow-up questioning, which minimized nonresponse bias.¹⁰ The remaining 15 patients were lost to follow-up. Unsuccessful attempts were made to either telephone those clients or to determine follow-up information by contacting the referring veterinarians.

The three clients who were either dissatisfied or partially dissatisfied with the hearing loss associated with the procedure owned pets that either had bilateral TECA-LBOs or a unilateral TECA-LBO in one ear and chronic otitis in the other. One client stated that the primary problem of pruritus was relieved, but would not perform bilateral TECA-LBOs again because it was too drastic. One client was satisfied with the surgery, itself, but was partially dissatisfied with the fact that the pet still had pruritus and pinnal dermatitis. One patient acutely collapsed 2 wk postoperatively (despite doing well in the perioperative period), and the client was concerned that the collapse might have been a sequela to surgery. The only client who was fully dissatisfied with the procedure stated that the pet became more aggressive postoperatively. The client noticed that the pet had become more easily startled.

Discussion

To the authors’ knowledge, this is the largest retrospective analysis of TECA-LBOs in dogs and cats to date. It has been more than 20 yr since the procedure was first evaluated in the literature, and it has been nearly that long since any reports on the neurologic complications of the procedure in dogs have been described.¹¹ In this study, the overall incidence of facial nerve deficits following TECA-LBO surgery was 48.9% (65 of 133 ears), comparable to the findings of Smeak et al. (1986), Mason et al. (1988), and Matthiesen et al. (1990), and greater than those reported by White et al. (1990).^2,3,5,6 Only 14 of the TECA-LBOs (10.5%) were associated with residual deficits. The incidence of residual facial nerve deficits in this report was lower than previous reports, except for White et al. (1990).²

In the veterinary literature, describing the complications of TECA-LBO, dysfunction of the facial nerve, is generally referred to as “paralysis.”^1,2,4–6 That terminology has been maintained for consistency, although “paralysis” technically refers to an inability to move the muscles innervated by the facial nerve. Physical examination, such as a palpebral reflex, evaluates the muscle group, rather than the nerve, directly. In animals that are described as having facial nerve paralysis, the facial nerve has likely undergone either axonotomesis (involving focal destruction of axons and myelin sheaths with an intact endoneurium and Schwann cell sheath) or neurotmesis (involving complete transection of the nerve from its cell body).¹² In animals that are described as having facial nerve paresis, the facial nerve has likely undergone neuropraxia, which is a transient interruption of nerve function and conduction without associated axonal degeneration.¹³

In both dogs and cats, Horner’s syndrome can be divided into first-, second-, and third-order. First-order Horner’s syndrome is caused by a disruption in central nerve fibers running from the hypothalamus, tectum, and tegmentum to the first three thoracic spinal cord segments. Second-order Horner’s syndrome is caused by a disruption in preganglionic nerve fibers that run from the first three thoracic spinal cord segments to the cranial ganglia, where they synapse on postganglionic neurons.¹⁴ The tympanic plexus, arising from the tympanic, superficial petrosal, and caroticotympanic nerves, contains postganglionic sympathetic nerve fibers that distribute and course over the bony promontory of the middle ear.¹² Iatrogenic disruption of those nerve fibers can result in third-order Horner’s syndrome. The distribution of those nerve fibers is similar in dogs and cats, despite the more distinct chambers seen in the feline middle ear.⁷ In this study, postoperative Horner’s syndrome was seen in a significantly higher percentage of cats than dogs. Although there is no previous study that directly compares this complication rate between species, higher complication rates among cats (42%) have been described in the literature, with a greater tendency toward residual deficits (14%).¹ Bacon et al. (2003) theorized that the relatively high incidence of Horner’s syndrome following LBO in the cat was due to either increased exposure or sensitivity of the tympanic plexus to iatrogenic trauma, rather than a species-specific difference in neuroanatomic structures.¹ The results of the current study also support the finding of increased incidence of nerve dysfunction in cats.

The most common malignant tumors of the canine and feline ear canal reported in the literature are CGAs, squamous cell carcinomas, and carcinomas of undetermined origin.^1,15,16 Those neoplasms tend to be locally invasive, with a low metastatic rate, making TECAs a successful treatment modality for local disease control of neoplasms contained within the cartilage of the ear canal.⁷ In one study, dogs undergoing surgical excision of tumors confined to either the horizontal or vertical canal had a median survival time of > 30 mo, whereas dogs with tumors that had extensive soft tissue involvement had a median survival time of only 5.3 mo.¹⁵ Radiation therapy is also a safe and effective adjunct to surgery when incomplete resection of carcinomas results, as an alternative to surgery when complete resection of carcinomas cannot be achieved, or when recurrence is observed.^15,17

The current study was consistent with previous findings in that CGAs and squamous cell carcinomas were the most common malignant neoplasms of the canine and feline ear canal. To the authors’ knowledge, only two studies in dogs and one study in cats have examined the incidence of neoplasms in a population of patients undergoing TECA-LBO.^1,4,6 In the aforementioned studies of canine TECA-LBOs, 4.2% and 9.6% of patients had CGAs, which was comparable to 8% in the canine population included herein.^4,6 In the feline study, CGAs were seen in 27.3% of ears, which is higher than reported in the current study (8.3%).¹ The overall percentage of malignant neoplasia seen in biopsied feline ears was 25% versus 10% seen in biopsied canine ears. The relatively higher percentage of cats with malignant neoplasia is consistent with trends in the previous literature on TECA-LBO, but was not significant.^1,15

The most common bacterial pathogens reported from cultures of the tympanic bulla of dogs were Staphylococcus spp. (43.5%) and Pseudomonas aeruginosa (23.5%), which are also the most common bacterial pathogens isolated from dogs with otitis media and otitis externa in previous studies.^18,19 Although cultures of otic exudate from the vertical ear canal and adjacent ear canal tissue have shown similar bacterial species and sensitivity spectrum, there are often discrepancies between the bacterial isolates and antimicrobial sensitivity patterns in the ear canal compared with those in the middle ear.¹⁹ For example, in only 4 of 10 cases was there agreement between preoperative and intraoperative bacterial cultures in the current study.

More than 70% of ears with otitis media may have an intact tympanic membrane; thus, an intact tympanum should not preclude suspicion of middle ear disease.¹⁸ This emphasizes the importance of obtaining bacterial cultures from the tympanic bulla at the time of surgery regardless of the gross appearance of the middle ear. Based on susceptibility patterns for Staphylococcus spp. and Pseudomonas spp., either enrofloxacin or a combination of amoxicillin and clavulanic acid appear to be the most efficacious empirical postoperative oral antibiotics to initiate pending culture and sensitivity.^18,20

In feline ears, Staphylococcus spp. were also the most common bacterial isolates (25%). This is consistent with a previous study of bacterial pathogens in the feline ear canal.¹⁹ Of the bacteria isolated, both Moraxella oslenosis and Prevotella spp. were unique to feline ears in this study. Cats were more likely than dogs to have a negative bacterial culture. This is consistent with the finding that a relatively smaller percentage of cats had biopsies consistent with chronic otitis compared with dogs; however, seven cats with either inflammatory polyps or neoplasia in the ear canal also had a positive bacterial culture. In addition, only 10% of cats with nonneoplastic middle ear disease may have clinical signs.²¹ Thus, bacterial cultures should be obtained in cats, even when noninfectious causes of disease are suspected.

A subjective assessment of postoperative hearing was requested during telephone follow up. Only two owners that were contacted believed their pet’s hearing was normal postoperatively. Roughly half of the remaining owners thought hearing was diminished (but present), while the other half thought their pet was completely deaf following the procedure. No baseline preoperative evaluation of hearing was performed. It was also noted that 22 dogs and 5 cats undergoing unilateral procedures were reported by the owner to have no hearing following the TECA-LBO. This implicates the contralateral ear in the disease process. The majority of those patients (77.8%) had a biopsy of the ipsilateral ear canal that was consistent with chronic otitis. Chronic otitis externa is often a bilateral condition.²² Although the medical records were insufficient to conclude a diagnosis of contralateral otitis externa, it was suspected by the perceived hearing loss. An association has also been suggested, in humans and dogs, between chronic inflammation and the development of aural neoplasms, which could account for a pre-existing bilateral otitis externa in the remaining four canine ears.¹⁵

Specific studies aimed at quantifying hearing loss following TECA-LBO have previously been conducted to evaluate brainstem auditory evoked potentials.^23,24 It has been suggested that hearing is diminished preoperatively in the majority of dogs undergoing TECA-LBO for chronic otitis, and although some patients lose behavior response to voice, they often retain brainstem auditory evoked responses to bone-conducted and, to a lesser extent, air-conducted stimuli.^23,24 A study of normal dogs undergoing TECA-LBOs experimentally reported postoperative deafness to be as high as 84%, but this determination was made on air-conducted stimuli alone and in patients in which the tympanic membrane and ossicles were routinely removed.²³ Although it is important to discuss the potential for diminished hearing postoperatively with clients, the extent of further hearing loss is equivocal in patients with end-stage otitis. According to the findings of this study, postoperative hearing does not seem to have any effect on owner satisfaction with the surgical outcome, unless unstaged bilateral TECA-LBOs are performed. This outcome was measured only by client perception; thus, no objective data were used here to compare the pre- and postoperative hearing status of the patients undergoing unilateral versus unstaged, bilateral TECA-LBO procedures.

In the current study, incisional complications such as minor dehiscence and discharge were only reported in 5.3% of ears, which is much lower than earlier reports of 15.7–41%.^2,3,5,6 This decrease in complication rates, as suggested by White et al. (1990) may be accounted for by thorough curettage of the bulla to remove the epithelial lining.² The intraoperative placement of drains has not produced superior results over meticulous primary closure.²⁵

CT is now a widely-used modality for selecting ears without gross disease extending beyond the auditory canal. Patients with extensive local disease may present a greater challenge in preserving normal anatomic structures. Preoperative CT scans were used in approximately half of the ears included in this study. CT has been shown to be more sensitive than radiography in detecting the presence and severity of otitis media.²⁶ Information from the CT scan, such as gross disease outside of the ear canal or bulla, was often used as either a preoperative surgical planning tool or to give prognostic information to the client if a neoplasm was confirmed/suspected. This insight preoperatively affected the case selection for surgery, providing one explanation why residual complication rates in the current study were lower than some of the previous published data.^1,3–6

The relatively small size of the study population limited meaningful subgroup comparisons by reducing statistical power to detect true differences. This was particularly true regarding cats. Feline TECA-LBOs occurred less commonly at the studied institutions, making those patients more difficult to accrue in adequate numbers. The statistical power of the study was also limited because some entries were excluded due to uncertainty.

Due to the retrospective nature of this study, there are inherent limitations, such as incomplete medical records, owner recall bias, variability in follow-up time between surgeries, and a small percentage of patients lost to follow-up. There was a subjective nature to assessing many parameters of interest such as deficits in palpebral reflex and hearing loss. More objective methods such as measuring brainstem auditory evoked potentials or measurement of palpebral fissure could have been implemented had this been a prospective study. It should also be noted that the overall percentages of chronic otitis in this group of patients may be slightly underrepresented due to the fact that some ear canals were not submitted for biopsy if there was a very high index of suspicion of chronic otitis versus a neoplastic process. Recall bias is an inherent issue when follow-up questionnaires are used in data collection. As much of the data as possible was obtained from the medical records because it was not subject to recall bias.

It should also be noted that follow-up information on complications of 21 TECA-LBO surgeries was determined < 1 yr postoperatively. The authors’ definition of residual neurologic deficits was those that persisted > 1 yr. This leaves open the theoretical possibility to underestimate the duration of postoperative complications due to the shorter follow-up period in those patients. For this cohort of patients, consisting of 4 cats and 17 dogs, it was noted that all facial nerve dysfunction and Horner’s syndrome had resolved by 6 wk postoperatively, except in 1 cat. That patient had received a TECA-LBO for squamous cell carcinoma of the ear canal and was euthanized approximately 2 mo postoperatively for progression of clinical signs, such as right-sided head tilt, absent pupillary light response, decreased gag response, and persistent Horner’s syndrome, which were all thought to be related to progression of local disease. Other late onset complications, such as draining tracts, may have been underestimated in other members of this subset of patients.

Conclusion

The goal of this study was to further quantify complication rates and determine underlying risk factors for postoperative complications for a relatively large clinical population of canine and feline TECA-LBO patients. Although no association was found between the duration of preoperative clinical signs and the incidence of postoperative facial nerve injury following TECA-LBO, several other useful conclusions were drawn. As demonstrated by Bacon et al. (2003), cats were significantly more likely to nerve damage than dogs.¹ In addition, the absence of a palpebral reflex in cats 12–24 hr postoperatively was a risk factor for residual facial nerve deficits. Feline patients also had a significantly higher risk of postoperative Horner’s syndrome. Cats with postoperative Horner’s syndrome had a significantly longer duration of preoperative clinical signs than those that did not have postoperative Horner’s syndrome. Cats, when compared with dogs, had a significantly shorter duration of preoperative clinical signs and were more likely to have a mass in the ear canal.

In cases of either chronic otitis or suspected ear canal neoplasia, progression of disease (such as ear canal stenosis, fibrosis, or invasion of normal structures) may be unpredictable. This creates too complex of a clinic scenario to allow establishment of an ideal window for surgical intervention. Even if a shorter duration of preoperative signs could be correlated to a lower incidence of neurologic complications, that would not negate the role of TECA-LBO as a salvage procedure. Clients and veterinarians generally prefer to manage otitis medically for as long as possible. The authors recommend early surgical intervention when neoplasia is suspected and when medical management of otitis externa is no longer effective.

Future studies involving larger numbers of feline TECA-LBOs may be useful for further evaluating trends of higher incidences of facial nerve injury and malignant neoplasia in cats undergoing this procedure. Regardless, important species differences were discovered in the current study that may help veterinarians to counsel owners about postoperative complications of the TECA-LBO.