Ventral bulla osteotomy is a commonly performed surgical procedure in cats for the management of middle ear diseases, including inflammatory polyps, otitis media, and benign and malignant neoplasia.1 The procedure involves exposing the osseous bulla through a ventral cervical paramedian approach, entering the ventral aspect of the osseous bulla with an intramedullary pin and rongeurs or a pneumatic drill, removing the bony septum separating the larger ventromedial and smaller dorsolateral compartments of the middle ear, curetting the epithelial lining of the osseous bulla, and excising the inflammatory polyp or tumor if indicated.2–6 Bilateral VBO is performed in cats with bilateral middle ear disease and can be staged or performed simultaneously as a single-stage procedure.
The most common complications following VBO in cats include persistent otic infection, Horner syndrome, vestibular signs, facial nerve paralysis, and local recurrence of the inflammatory polyp or tumor.1–9 Although complications and outcomes have been documented for cats undergoing TECA-LBO for the treatment of diseases of the middle ear and external ear canals,10 few reports3–6,8 exist of outcomes in cats following VBO, and the largest study5 included only 26 cats. Anecdotally, high incidences of respiratory-related complications and death following single-stage bilateral VBO in cats have been noted by several surgeons participating in the Veterinary Society of Surgical Oncology listserv.
The purpose of the retrospective study reported here was to evaluate cats with middle ear disease treated with unilateral, staged bilateral, or single-stage bilateral VBO to determine whether single-stage bilateral VBO was associated with a higher risk of respiratory complications and perioperative death; to determine rates of transient and permanent postoperative complications following VBO; and to describe the outcomes following VBO in cats for the treatment of nonneoplastic, benign, and malignant disease. We hypothesized that single-stage bilateral VBO would be associated with higher likelihood of postoperative respiratory complications and disease-related death than the other 2 VBO procedures.
Materials and Methods
Animals
Medical records of cats that underwent VBO at 25 veterinary referral and academic hospitals from January 1, 2005, to December 31, 2016, were obtained and reviewed. Cats with incomplete follow-up information or that were lost to follow-up within 100 days after surgery were excluded from analysis.
Data collection
Data were retrieved from medical records, including patient signalment (age, sex, and breed), clinical signs, diagnostic investigations (CBC, serum biochemical analysis, oropharyngeal examination, radiography, otoscopy, endoscopy, CT, and MRI) and noted abnormalities, anesthetic details, surgical procedure details, intraoperative diagnostic test (histologic evaluation and bacterial culture) results, postoperative management details, complications, and outcome. Anesthetic details included premedication, induction, and inhalant agents; intraoperative analgesics and antimicrobials; and anesthesia time (defined as the interval from induction of anesthesia to extubation). Surgical details included whether the VBO was performed as a unilateral, staged bilateral, or single-stage bilateral procedure; surgical time (defined as the interval from initiation of the incision to completion of closure of the incision); and surgical technique (intramedullary pin and rongeurs, osteotome and rongeurs, or pneumatic burr). For cats treated with staged bilateral VBO, surgical time was defined as the accumulative time to perform both unilateral VBOs. Postoperative management details included drain use and duration as well as analgesic and antimicrobial administration.
Complications were classified as anesthetic, intraoperative (surgical), and postoperative, with postoperative complications further classified as immediate (< 24 hours), short term (1 to 14 days), long term (> 14 days), or permanent. Permanent complications were defined as complications present at the time of patient death or loss to follow-up or at the end of the study period. Anesthetic complications were classified as hypotension, hypovolemia, cardiac arrest, bradycardia, arrhythmia, or hypothermia. For this purpose, hypotension was defined as a mean arterial blood pressure < 60 mm Hg, systolic arterial pressure < 90 mm Hg, or both. Postoperative complications were classified as respiratory related, neurologic (Horner syndrome, head tilt, or facial nerve paralysis), incisional (including surgical site infections), and otic. Respiratory complications were classified as mild (eg, stertor or nasal congestion without respiratory distress) or severe (eg, respiratory distress with or without cyanosis).
Measured outcomes included local recurrence, time to local recurrence, and survival time. Local recurrence was defined as a recurrence or progression of clinical signs following VBO ipsilateral to the side of the original disease. Death was considered surgery related if it occurred during surgery or postoperative hospitalization and disease related if death occurred as a result of the underlying disease.
Statistical analysis
Statistical analyses were performed with the aid of statistical software.a Continuous variables were evaluated for normality of distribution by creation of histograms and use of skewness, kurtosis, and Shapiro-Wilk tests. Normally distributed data are reported as mean and SD and nonnormally distributed data as median (range). Categorical data are reported as frequencies and percentages.
Associations between the presence of specific preoperative abnormalities (ie, middle ear mass, inner ear abnormalities, middle ear abnormalities, external ear abnormalities, bulla changes, CNS extension, rhinitis, and nasopharyngeal mass) and histopathologic diagnoses were evaluated with the Fisher exact test. Differences in median anesthesia and surgery times between unilateral and bilateral VBO procedures were assessed by use of the Mood median test. Associations with disease recurrence following surgery for the histopathologic diagnoses of inflammatory polyps and malignant neoplasia were assessed with the Fisher exact test.
The Kaplan-Meier method was used to calculate the median duration of neurologic conditions as well as the median value and 95% CI for the time to local recurrence and overall survival time for all cats and cats with benign and malignant tumors. The log-rank test was used to test the difference between survival distributions. In addition, the overall 1- and 2-year survival rates were calculated.
Univariate logistic regression analysis was performed to test for associations with respiratory complications, severe respiratory complications, surgery-related death, and disease-related death for the following variables: preoperative upper respiratory tract signs, anesthesia time, premedication with buprenorphine or butorphanol, premedication with an opioid, preoxygenation, propofol induction, isoflurane anesthesia, surgical time, type of VBO procedure, and histopathologic diagnosis of inflammatory polyps, infectious disease, benign tumor, or malignant tumor. A multivariate logistic regression model was constructed by use of a stepwise procedure to determine whether associations existed between type of VBO procedure and development of respiratory complications, severe respiratory complications, and disease-related death. Variables were included in the multivariate analysis if they had a value of P < 0.20 on univariate analysis and retained if they had a value of P ≤ 0.05 after controlling for other variables. The Hosmer-Lemeshow goodness-of-fit test was performed to evaluate the final selected model, and ORs were calculated with 95% Wald CIs for each included variable. Values of P < 0.05 were considered significant for all analyses.
Results
Signalment
A total of 282 cats (141 castrated males, 93 spayed females, 24 sexually intact males, and 24 sexually intact females) met the criteria for inclusion in the study. Most cats were classified as mixed breed, including domestic shorthair (n = 180), domestic longhair (20), and domestic medium hair (9). The remaining 73 cats were classified as Maine Coon (n = 29), Russian Blue (5), Ragdoll (4), Siamese (4), British Shorthair (3), Tonkinese (3), and Abyssinian, Bengal, Birman, Burmese, Chartreux, Cornish Rex, Devon Rex, Egyptian Mau, Himalayan, Japanese Bobtail, Norwegian Forest Cat, Ocicat, Oriental Shorthair, Persian, Snowshoe, Somali, Sphinx, and Turkish Angora (≤ 2 each). Age ranged from 0.2 to 17.0 years (median, 3.4 years), and body weight ranged from 2.3 to 8.0 kg (5.1 to 17.6 lb; median, 4.1 kg [9.0 lb]).
Clinical signs
The most common primary clinical signs at initial evaluation were head tilt (n = 101 [35.8%]) and upper respiratory signs (92 [32.6%]), the latter of which included stertor (64 [22.7%]), dyspnea (26 [9.2%]), cough (18 [6.4%]), stridor (10 [3.5%]), and snoring (3 [1.1%]). Other signs included sneezing (41 [14.5%]) and serous, purulent, or mucoid nasal discharge (39 [13.8%]). Ninety-one (32.3%) cats had otitis media, which was unilateral in 63 cats (right-sided in 34 cats and left-sided in 29 cats) and bilateral in 28 cats. Neurologic signs included Horner syndrome (n = 40 [14.2%]); horizontal, positional, or rotary nystagmus (31 [11.0%]); dysphagia (12 [4.3%]); and dysphonia (8 [2.8%]).
Preoperative findings
A preoperative CBC and serum biochemical analysis were performed for 199 cats, of which 123 (61.8%) had abnormalities identified. Such abnormalities included hyperglobulinemia (n = 18), leukocytosis (18), anemia (16), neutrophilia (15), hypoalbuminemia (8), hypoproteinemia (4), high liver enzyme activities (17), and azotemia (7).
An oropharyngeal examination was performed for all cats. Various imaging techniques were also used, including CT (n = 180) or MRI (45) of the head, otoscopy (158), radiography of the skull (35), and endoscopy of the oropharynx (12). Eighteen cats were examined via 2 and 1 cat via all 3 of the following imaging modalities: radiography, CT, and MRI. Otitis externa, otitis media, and otitis interna were diagnosed on the basis of the imaging reports from the medical records in 89 of 278 (32.0%), 195 of 233 (83.7%), and 11 of 234 (4.7%) cats, respectively. A middle ear mass was detected in 184 of 281 (65.5%) cats and a nasopharyngeal mass in 46 of 282 (16.3%) cats. Changes in the osseous bulla were identified via radiography, CT, or MRI in 55 of 235 (23.4%) cats. No changes were recorded for the remaining 54 cats that were examined by these modalities. Cats with changes in the osseous bulla identified via CT were more likely to have a malignant tumor than to not have a malignant tumor (P = 0.0497).
Intraoperative details
Before surgery, antimicrobials were orally administered to 92 cats and locally administered to 24 cats. Antimicrobials were administered during surgery to 264 cats. Overall, 289 separate VBO procedures were performed. Unilateral VBO was performed in 211 of 282 (74.8%) cats (right-sided in 115 cats and left-sided in 96 cats), staged bilateral VBO was performed in 7 (2.5%) cats, and single-stage bilateral VBO was performed in 64 (22.7%) cats. Median anesthesia times were 130 minutes (range, 45 to 360 minutes) for unilateral VBO, 100 minutes (range, 60 to 180 minutes) for staged bilateral VBO, and 135 minutes (range, 45 to 360 minutes) for single-stage bilateral VBO (n = 239 cats in total). No significant difference in anesthesia times was identified between unilateral or staged bilateral VBO and single-stage VBO.
The VBO (n = 289) was performed with an intramedullary pin and rongeurs (240 [83.0%]), pneumatic burr (33 [11.4%]), or osteotome and rongeurs (17 [5.9%]). One VBO was performed with both an osteotome and pneumatic burr. All 46 cats with a nasopharyngeal polyp were treated with traction avulsion of the polyp through the oral cavity or ear canal followed by VBO, including 1 cat treated with staged bilateral VBO and 9 cats treated with single-stage bilateral VBO. Median surgical times were 62 minutes (range, 20 to 240 minutes) for unilateral VBO, 41 minutes (range, 20 to 120 minutes) for staged bilateral VBO (representing a median time for each unilateral surgery), and 82 minutes (range, 30 to 240 minutes) for single-stage bilateral VBO (n = 239 cats in total). Surgical time differed significantly (P < 0.001) between unilateral or staged bilateral VBO and single-stage bilateral VBO. The median time between surgeries for cats treated with staged bilateral VBO was 36 days (range, 28 to 1,477 days).
Intraoperative diagnostic findings
Histologic evaluation of biopsy specimens and bacterial culture of middle ear contents were performed for 214 and 230 cats, respectively. Histopathologic diagnoses for the 214 tested cats included inflammatory polyp (n = 150 [70.1%]), inflammation or otitis media (46 [21.5%]), malignant tumor (13 [6.1%]; including squamous cell carcinoma [5], ceruminous gland adenocarcinoma [3], lymphoma [3], osteosarcoma [1], and anaplastic carcinoma [1]), and benign tumor (5 [42.8%]; including ceruminous gland adenoma [2], cholesteatoma [2], and papillary adenoma [1]). The median ages of cats with an inflammatory polyp, otitis media, malignant neoplasia, and benign neoplasia were 2.0, 5.5, 10.0, and 12.0 years, respectively. Cats with an inflammatory polyp were significantly (P < 0.001) younger than cats with other middle ear diseases.
No bacterial growth was reported for 150 of the 230 (65.2%) tested cats, and growth of the following bacteria was obtained for the remaining 80 cats: Staphylococcus spp (20 [25.0%]), Pasteurella spp (18 [22.5%]), Streptococcus spp (15 [18.8%]), multidrug-resistant bacteria (11 [13.8%]; including methicillin-resistant Staphylococcus aureus [3], Pseudomonas spp [2], Bordetella spp [2], Corynebacterium spp [1], Streptococcus spp [1], Escherichia coli [1], and Corynebacterium spp and Pasteurella spp [1]), Pseudomonas spp (7 [8.8%]), Escherichia coli (6 [7.5%]), Bordetella spp (5 [6.2%]), Corynebacterium spp (3 [3.8%]), Enterococcus spp (3 [3.8%]), nonfermentative gram-negative bacilli (2 [2.5%]), and Mycoplasma spp, Enterobacteriaceae spp, Proteus mirabilis, and gram-positive anaerobe (1 [1.2%] each).
Postoperative management
A drain from the surgical site was placed following 83 of the 289 (28.7%) VBO procedures. Median duration of drain placement was 2.0 days (range, 0.1 to 10.0 days). Antimicrobials were administered to 242 cats in the postoperative period. Postoperative analgesia was recorded for 271 of the 289 (93.8%) procedures and included opioids (n = 218 cats; including buprenorphine [161], fentanyl patch [40], and tramadol [17]). Nonsteroidal anti-inflammatory drugs were administered to 98 cats, and corticosteroid drugs were administered to 31 cats. Gabapentin was used for 3 cats.
Complications
Anesthetic complications were recorded for 56 of the 289 (19.4%) VBO procedures and included hypotension (n = 39 [13.5%] procedures), bradycardia (19 [6.6%]), hypovolemia (5 [1.7%]), arrhythmia (3 [1.0%]), hypothermia (3 [1.0%]), and cardiac arrest (1 [0.3%]). No deaths occurred intraoperatively or as a result of anesthetic-related complications. Intraoperative surgical complications were recorded for 12 (4.2%) procedures and included hemorrhage (6 [2.1%]), anatomic problems (5 [1.7%]), and technical problems (1 [0.3%]). Anatomic complications included observation of a mass in the external ear canal following VBO (n = 3), observation of a mass in the middle ear attached to the malleus (1), and rupture of the dorsal wall of the tympanic bulla. No procedures were converted from a VBO to a TECA-LBO. The technical complication involved loss of normal anatomic landmarks owing to thickening of the bulla. The VBO was abandoned in this situation, and a biopsy specimen was collected from the middle ear mass following myringotomy. Histopathologic findings were consistent with an inflammatory polyp.
Immediate (< 24 hours), short-term (1 to 14 days), and long-term (> 14 days) complications following surgery were summarized (Table 1). Overall, respiratory complications were reported for 18 (9%), 2 (29%), and 30 (47%) cats undergoing unilateral, staged bilateral, and single-stage bilateral VBO, respectively (Table 2). Respiratory signs were classified as mild (n = 5) or severe (34) and included respiratory distress (16), airway occlusion or swelling (11), apnea (9), and aspiration pneumonia (3). Eleven of the 34 (32%) cats with severe respiratory complications developed cardiac arrest, died, or were euthanized within 5 days after surgery: 2 cats within the first 12 hours after surgery, 5 between 12 and 24 hours after surgery, 2 between 24 and 48 hours after surgery, 1 at 3 days after surgery, and 1 at 5 days after surgery. Of these 11 cats, 9 had been treated with single-stage bilateral VBO and 2 with unilateral VBO. Tracheostomy was performed in 2 of these cats, both of which failed to improve and were euthanized owing to poor mentation and anticipated poor prognosis.
Preoperative, immediate (0 to 24 hours after surgery), short-term (1 to 14 days after surgery), and long-term (> 14 days after surgery) complications of 289 VBO procedures performed in 282 cats, by histopathologic diagnosis of the underlying otic disease.
| Respiratory signs | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| Complication, by histopathologic diagnosis | Horner syndrome | Head tilt | Facial nerve paralysis | Mild | Severe | Overall | Dysphagia | Otitis externa or media | Local recurrence | Incisional complication | Overall complications |
| Inflammatory polyp | |||||||||||
| Preoperative (n = 150) | 14 (9) | 38 (25) | 4 (3) | — | — | 61 (41) | 3 (2) | 40 (27) | — | — | — |
| Immediate (n = 150) | 104 (69) | 39 (26) | 14 (9) | 1 (1) | 10 (7) | 11 (7) | 0 | 0 | 0 | 3 (2) | 120 (80) |
| Short-term (n = 148) | 96 (65) | 37 (25) | 12 (8) | 6 (4) | 3 (2) | 9 (6) | 9 (6) | 8 (5) | 0 | 7 (5) | 116 (78) |
| Long-term (n = 144) | 65 (45) | 28 (19) | 11 (8) | 3 (2) | 2 (1) | 5 (3) | 4 (3) | 9 (6) | 3 (2) | 2 (1) | 89 (62) |
| Otitis media | |||||||||||
| Preoperative (n = 46) | 8 (17) | 22 (48) | 7 (15) | — | — | 11 (24) | 5 (11) | 20 (43) | — | — | — |
| Immediate (n = 46) | 30 (65) | 17 (37) | 11 (24) | 0 | 11 (24) | 11 (24) | 0 | 0 | 0 | 0 | 37 (80) |
| Short-term (n = 45) | 26 (58) | 15 (33) | 11 (24) | 1 (2) | 3 (7) | 4 (9) | 4 (9) | 2 (4) | 0 | 1 (2) | 35 (78) |
| Long-term (n = 43) | 19 (44) | 14 (33) | 8 (19) | 0 | 2 (5) | 2 (5) | 2 (5) | 3 (7) | 6 (14) | 1 (2) | 33 (77) |
| Benign tumor | |||||||||||
| Preoperative (n = 5) | 1 (20) | 2 (40) | 1 (20) | — | — | 0 | 0 | 2 (40) | — | — | — |
| Immediate (n = 5) | 4 (80) | 1 (20) | 1 (20) | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 5 (100) |
| Short-term (n = 5) | 4 (80) | 1 (20) | 1 (20) | 0 | 0 | 0 | 1 (20) | 1 (20) | 0 | 1 (20) | 5 (100) |
| Long-term (n = 5) | 5 (100) | 1 (20) | 1 (20) | 0 | 0 | 0 | 0 | 1 (20) | 1 (20) | 0 | 5 (100) |
| Malignant tumor | |||||||||||
| Preoperative (n = 13) | 5 (38) | 7 (54) | 4 (31) | — | — | 3 (23) | 2 (15) | 3 (23) | — | — | — |
| Immediate (n = 12) | 7 (58) | 4 (33) | 6 (50) | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 10 (83) |
| Short-term (n = 12) | 7 (58) | 4 (33) | 5 (42) | 0 | 0 | 0 | 6 (50) | 1 (8) | 0 | 1 (8) | 8 (67) |
| Long-term (n = 11) | 4 (36) | 2 (18) | 3 (27) | 1 (9) | 0 | 1 (9) | 2 (18) | 0 | 4 (36) | 0 | 9 (82) |
| No histologic evaluation performed | |||||||||||
| Preoperative (n = 75) | 12 (16) | 32 (43) | 2 (3) | — | — | 17 (23) | 2 (3) | 26 (35) | — | — | — |
| Immediate (n = 75) | 48 (81) | 24 (32) | 5 (7) | 5 (7) | 5 (7) | 10 (13) | 0 | 0 | 0 | 2 (3) | 59 (79) |
| Short-term (n = 72) | 45 (62) | 25 (35) | 4 (6) | 5 (7) | 2 (3) | 7 (10) | 6 (8) | 2 (3) | 0 | 6 (8) | 55 (76) |
| Long-term (n = 71) | 32 (45) | 23 (32) | 3 (4) | 5 (7) | 1 (1) | 6 (8) | 1 (1) | 9 (13) | 3 (4) | 1 (1) | 47 (66) |
Values represent the number (%) of procedures performed in cats with the indicated histopathologic diagnosis.
— = Variable not relevant for the preoperative category.
Classification of respiratory complications associated with unilateral (n = 211 procedures), staged bilateral (14 total procedures), or single-stage bilateral (64 procedures) VBO for the cats of Table 1.
| Diagnosis category | Respiratory complications | Unilateral | Staged bilateral | Single-stage bilateral |
|---|---|---|---|---|
| Preoperative (n = 289) | None | 144 (68) | 10 (71) | 43 (67) |
| Any severity | 67 (32) | 4 (29) | 21 (33) | |
| Inflammatory polyp (n = 150) | None | 113 (93) | 6 (100) | 12 (52) |
| Mild | 4 (3) | 0 (0) | 2 (9) | |
| Severe | 4 (3) | 0 (0) | 9 (39) | |
| Otitis media (n = 46) | None | 25 (83) | 2 (50) | 4 (33) |
| Mild | 0 | 0 | 1 (8) | |
| Severe | 5 (17) | 2 (50) | 7 (58) | |
| Benign tumor (n = 5) | None | 4 (100) | 0 | 1 (100) |
| Mild | 0 (0) | 0 | 0 | |
| Severe | 0 (0) | 0 | 0 | |
| Malignant tumor (n = 13) | None | 10 (100) | 0 | 1 (33) |
| Mild | 0 | 0 | 1 (33) | |
| Severe | 0 | 0 | 1 (33) | |
| No histologic evaluation performed (n = 75) | None | 41 (89) | 4 (100) | 16 (64) |
| Mild | 2 (4) | 0 | 6 (24) | |
| Severe | 3 (7) | 0 | 3 (12) |
Values represent the number (%) of VBO procedures with the indicated severity of respiratory complications.
Single-stage bilateral VBO was associated with a significantly (P < 0.001) higher odds of any respiratory complication (OR, 16.4; 95% CI, 5.9 to 45.2) and severe respiratory complications (OR, 12.9; 95% CI, 4.5 to 36.4), compared with unilateral or staged bilateral VBO, even after controlling for the presence of preoperative respiratory signs and a histopathologic diagnosis of inflammatory polyp on multivariate analysis (Tables 3 and 4). No significant difference was detected in the odds of respiratory complications for unilateral or single-stage bilateral VBO versus bilateral staged VBO (OR, 0.8; 95% CI, 0.2 to 3.6). No significant (P = 0.57) difference was found between the type of surgical technique used (intramedullary pin, osteotome, or pneumatic burr) and the development of respiratory complications (OR, 1.3; 95% CI, 0.5 to 3.8).
Factors associated with respiratory complications of any type (both mild and severe) on univariate (unadjusted) and multivariate (adjusted) analyses for the cats of Table 1.
| Unadjusted | Adjusted | |||
|---|---|---|---|---|
| Factor | OR (95% CI) | P value | OR (95% CI) | P value |
| Preoperative upper respiratory signs | 0.3 (0.2–0.5) | < 0.001 | 8.5 (3.0–24.0) | < 0.001 |
| Anesthetic induction with propofol | 2.5 (1.1–5.8) | 0.03 | — | — |
| Single-stage bilateral VBO | 9.0 (4.6–17.7) | < 0.001 | 16.4 (5.9–45.2) | < 0.001 |
| Histopathologic diagnosis of inflammatory polyp | 2.2 (1.1–4.6) | 0.03 | — | — |
| Positive bacterial culture result | 0.3 (0.1–0.6) | < 0.001 | 5.1 (1.8–14.6) | 0.001 |
A significant OR > 1 suggests an increased risk of the outcome relative to the risk in the referent group (here, cats without the indicated factor), and a significant OR < 1 indicates a decreased risk.
— = No adjusted OR was calculated for this factor because it was not included in the final multivariate model.
Factors associated with severe respiratory complications on univariate (unadjusted) and multivariate (adjusted) analyses for the cats of Table 1.
| Unadjusted | Adjusted | |||
|---|---|---|---|---|
| Factor | OR (95% CI) | P value | OR (95% CI) | P value |
| Preoperative upper respiratory signs | 2.4 (1.2–5.0) | 0.02 | 6.1 (2.0–18.3) | < 0.001 |
| Anesthetic induction with propofol | 3.0 (1.0–8.7) | 0.05 | — | — |
| Single-stage bilateral VBO | 6.8 (3.2–14.6) | < 0.001 | 12.9 (4.5–36.4) | < 0.001 |
| Histopathologic diagnosis of inflammatory polyp | 0.3 (0.2–0.7) | 0.007 | — | — |
| Positive bacterial culture result | 6.4 (2.8–14.9) | < 0.001 | 7.2 (2.4–22.0) | 0.002 |
See Table 3 for key.
Numbers of cats with transient and permanent Horner syndrome, head tilt, and facial nerve paralysis were summarized (Table 5). Cats with Horner syndrome, head tilt, or facial nerve paralysis had 2.6 times (95% CI of the OR, 1.3 to 5.5; P = 0.01), 3.3 times (95% CI of the OR, 1.2 to 9.1; P = 0.02), and 5.6 times (95% CI of the OR, 1.3 to 25.3; P = 0.02) the odds of having these respective neurologic conditions permanently, compared with cats that had no such preexisting signs. The overall median duration of these signs and the median time to their resolution were summarized (Table 6). The duration of these signs was longer for cats that had them before surgery versus those that developed them after surgery.
Number (%) of VBO procedures (n = 289) associated with various neurologic conditions in the cats of Table 1.
| Transient | Permanent | Overall | |||||||
|---|---|---|---|---|---|---|---|---|---|
| Condition | Preexisting | Not preexisting | Total | Preexisting | Not preexisting | Total | Preexisting | Not preexisting | Total |
| Horner syndrome | 25 (8.7) | 116 (40.1) | 141 (48.8) | 15 (5.2) | 41 (14.2) | 56 (19.4) | 40 (13.8) | 157 (54.3) | 197 (68.2) |
| Head tilt | 12 (4.2) | 11 (3.8) | 23 (8.0) | 47 (16.3) | 17 (5.9) | 64 (22.1) | 59 (20.4) | 28 (9.7) | 87 (30.1) |
| Facial nerve paralysis | 3 (1.0) | 13 (4.5) | 16 (5.5) | 13 (4.5) | 10 (3.5) | 23 (8.0) | 16 (5.5) | 23 (8.0) | 39 (13.5) |
Preexisting signs were considered those reported prior to the VBO procedure, and nonpreexisting signs were considered those reported after the procedure.
Median (range) duration (days) of various neurologic conditions in the cats of Table 5 in which the signs were transient (ie, not permanent).
| Condition | Overall | Preexisting | Not preexisting |
|---|---|---|---|
| Horner syndrome | 42 (1–3,124) | 42 (1–3,124) | 24 (0.1*–3,102) |
| Head tilt | 327 (1–2,928) | 454 (2–2,928) | 55 (1–2,432) |
| Facial nerve paralysis | 37 (1–3,124) | 42 (1–2,548) | 37 (1–3,124) |
In 1 cat, Horner syndrome was reported in the first hour after surgery and then resolved.
See Table 5 for remainder of key.
Outcome
By the end of the study follow-up period, 201 cats remained alive, 28 were lost to follow up, and 37 had died or been euthanized for reasons related to the underlying otic disease. Median duration of follow-up for the cats lost to follow-up was 3,274 days (range, 2,684 to 3,883 days). Local disease recurrence was noted following 18 of the 289 (6.2%) VBO procedures. Histopathologic diagnosis was significantly associated with recurrence. The local recurrence rate was 5% (4/88) for inflammatory polyps and 21% (10/48) for benign and malignant tumors (P = 0.006). Malignant tumors were more likely to recur (44% [4/9]) than benign tumors (15% [6/39]; P = 0.007). Median time to local recurrence for malignant tumors was 638 days and could not be estimated (ie, was not reached) for inflammatory polyps.
Median survival times and overall uncensored 1-and 2-year survival rates for cats with inflammatory polyps, benign tumors, and malignant tumors were summarized (Table 7). The median survival time for cats with malignant tumors was significantly (P < 0.001) shorter than for cats with nonmalignant conditions. Disease-related deaths were reported for 37 of the 282 (13.1%) cats, including 21 (10.0%) cats treated with unilateral VBO and 16 (25.0%) cats treated with single-stage bilateral VBO. Surgery-related deaths were reported for 12 (4.3%) cats overall, including 2 (0.9%) cats treated with unilateral VBO and 10 (15.6%) cats treated with single-stage bilateral VBO. Unilateral and staged VBO had no association with disease-related deaths. Although single-stage bilateral VBO was significantly associated with surgery-related death (OR, 11.4; 95% CI, 2.0 to 66.4; P = 0.007) and disease-related death (OR, 2.7; 95% CI, 1.1 to 6.9; P = 0.04) on univariate analyses, the association with disease-related death was not significant on multivariate analysis.
Median survival times and overall survival rates (proportion [%]) on the basis of histopathologic diagnosis for the cats of Table 1.
| Histopathologic diagnosis | Median survival time (d) | 1-year survival rate | 2-year survival rate |
|---|---|---|---|
| Inflammatory polyp | 1,322 | 117/132 (89) | 96/130 (74) |
| Benign tumor | 1,244 | 3/5 (60) | 1/3 (33) |
| Malignant tumor | 855 | 3/4 (75) | 1/2 (50) |
Discussion
The primary aim of the retrospective study reported here was to evaluate whether cats treated with single-stage bilateral VBO had a worse outcome than cats treated with either unilateral or staged bilateral VBO. Our hypothesis was supported by the finding that cats treated with single-stage bilateral VBO were significantly more likely to have postoperative respiratory complications, surgery-related deaths, and disease-related deaths than were cats treated with the other approaches. Severe respiratory complications were diagnosed in 31% of the 64 cats treated with single-stage bilateral VBO, compared with 14% and 5.7% of cats treated with staged bilateral and unilateral VBO, respectively; and the odds of developing severe respiratory complications for cats treated with single-stage bilateral VBO were 12.9 times those for the other cats. Cats treated with single-stage bilateral VBO were also 11.4 times as likely to have a surgery-related death and 2.7 times as likely to have a disease-related death as were cats treated with unilateral or staged bilateral VBO.
Factors significantly associated with the development of respiratory complications on univariate analysis included preoperative upper respiratory signs, anesthetic induction with propofol, single-stage bilateral VBO, histopathologic diagnosis of inflammatory polyp, and positive bacterial culture results. A reported adverse effect of propofol is apnea, which usually occurs at the time of anesthetic induction.11 If apnea is improperly addressed, this can result in respiratory-related death; however, propofol appeared to be an unlikely cause of respiratory-related complications in the present study given that such complications were reported in the postoperative period for all patients. Another possibility was that the finding regarding propofol was a type I statistical error, particularly given that it did not retain a significant association with outcome in the multivariate model. It is more likely that the presence of preoperative upper respiratory signs, presence of an inflammatory polyp, and positive bacterial culture results contributed to the respiratory complications of cats in this study, although cause-effect relationships cannot be established from retrospective studies.
Compared with benign and malignant tumors, inflammatory polyps may be associated with a higher degree of inflammation given their histologic characteristics, may be more likely to have bilateral middle ear involvement, and may be more likely to be associated with some degree of nasopharyngeal involvement, whether it be microscopic or macroscopic in the form of nasopharyngeal polyps. In addition, cats with upper respiratory signs or bacterial otitis media may have substantially more inflammation of the middle ear, Eustachian tube, and pharynx than other cats. Although these factors would also contribute to pharyngeal swelling following unilateral VBO procedures and were likely responsible for the 5.7% to 14% severe respiratory complication rate noted following unilateral and staged bilateral VBO in the present study, the degree of pharyngeal swelling following single-stage bilateral VBO is more likely to be clinically relevant, compared with that following unilateral procedures. Importantly, single-stage bilateral VBO in the present study remained a significant risk factor for severe postoperative respiratory complications when adjusted for the presence of preoperative respiratory signs and histopathologic diagnosis of inflammatory polyp on multivariate analysis. Other factors that may contribute to postoperative respiratory complications that were not investigated in the present study include the proximity of the osseous bulla to the pharynx, drainage of blood or lavage fluid into the pharynx via the Eustachian tube, and bleeding and inflammation associated with traction avulsion of aural and nasopharyngeal polyps through the oral cavity or ear canal.
Surgical techniques used to perform VBO in the study reported here included use of a pneumatic burr, intramedullary pin and rongeurs, or osteotome and rongeurs. In the authors' experience, the use of a pneumatic burr has been postulated by some surgeons to cause micromotion and subsequent inflammation of the middle ear, ultimately predisposing patients to respiratory complications; however, no significant difference was found among the 3 aforementioned surgical techniques with respect to rates of postoperative respiratory complications and disease-related deaths. Regardless of the cause of pharyngeal swelling and respiratory complications, we would not recommend single-stage bilateral VBO for cats because of the increased risk of respiratory complications and of surgery- and disease-related deaths as a result of these complications.
The secondary aims of the present study were to determine rates of postoperative complications associated with VBO in cats and outcomes for cats with inflammatory polyps, otitis media, and benign and malignant middle ear tumors. The rate of anesthetic and intraoperative complications was low, with hemorrhage noted as the most common intraoperative complication (2.1%). The overall rates of postoperative and non-resolving complications were higher than previously reported. The overall postoperative complication rate was 81.7%, with 80.0%, 77.7%, and 66.8% of cats having immediate, short-term, and long-term postoperative complications, respectively. The most common postoperative complications were Horner syndrome, head tilt, and facial nerve paralysis.
Horner syndrome is a combination of clinical signs (miosis, ptosis, enophthalmos, or prolapsed third eyelid) caused by damage to the sympathetic fibers as they course over the promontory.12 The high incidence of this complication in cats versus dogs is attributed to the fragility of the feline tympanic plexus during curettage and lavage of the diseased bulla.10 In the present study, Horner syndrome was present in 13.8% of cats before surgery and developed in 54.3% of cats after surgery. The overall rate of postoperative Horner syndrome (19.4%) was similar to that in previous studies3–6,8; however, in 1 study,3 only 10% of cats had permanent Horner syndrome. The most likely explanation for this difference is the higher percentage of cats with Horner syndrome prior to surgery in the present study, given the finding that cats with preoperative Horner syndrome were significantly more likely to have permanent Horner syndrome after surgery. In those situations, damage to the sympathetic fibers of the middle ear was likely caused by the underlying disease process prior to any surgical intervention and therefore VBO may not have reversed this damage.12,13 This is supported by the difference in median time to resolution of signs for cats with Horner syndrome before surgery (42 days) versus those without the condition before surgery (24 days). The longer duration of postoperative Horner syndrome in cats with this condition before surgery suggested that the underlying disease process causes chronic changes that take longer to resolve or never resolve.
A head tilt originates from disruption or trauma to the vestibulocochlear nerve as it courses through the inner ear, which can result from inflammation secondary to the underlying disease process or from trauma at the time of surgery.12 To the authors' knowledge, the postoperative development of head tilt has not been investigated previously. During VBO, all abnormal and inflamed tissue is removed and the bulla is lavaged with sterile saline (0.9%) solution. Removal of this tissue may relieve compression of the vestibular nerve, resulting in postoperative clinical improvement in some cases. In 74% of cats in the present study, however, the head tilt was permanent. Cats with (vs without) a preoperative head tilt had 3.3 times the odds of having a permanent head tilt after surgery, suggesting that irreversible damage to this nerve and associated vestibular structures was present prior to surgical intervention. The median time to resolution of a head tilt for cats with transient head tilt was longer for those that had this condition before surgery (454 days) versus after surgery (55 days). This further supported the idea that chronic changes from the underlying disease were affecting the vestibulocochlear nerve and that these changes took longer to resolve or in some cases never resolved.
The overall prevalence of facial nerve paralysis in the cats of the present study (13.5%) was lower than that identified in cats treated with VBO or TECA-LBO in another study (42.9%),3 which might be expected given the inclusion of cats treated with TECA-LBO in the other study and the anatomic course of the facial nerve in relation to the osseous bulla. Cats with (vs without) facial nerve paralysis before surgery in the study reported here had 5.6 times the odds of having permanent facial nerve paralysis following surgery. Similar to the situation for head tilt, this finding was likely due to irreversible damage to the facial nerve by the underlying disease process prior to any surgical intervention.
Most cats in the present study had a histopathologic diagnosis of inflammatory polyp or otitis media. The local recurrence rate for cats with inflammatory polyps was 5%, which was similar to the rate in another study3 but higher than the rate in other studies.4–6 This difference was most likely due to the duration of postoperative follow-up. The median follow-up time in the present study was 9.0 years, whereas follow-up times were ≤ 3 years in previous studies; consequently, late-occurring recurrent disease was more likely to be captured in the present study. The outcome for cats with benign tumors was good to excellent, with a local recurrence rate of 8% and a median survival time of 1,244 days. Although the uncensored 1- and 2-year survival rates for cats with benign tumors were 60% and 30%, respectively, most deaths accounted for in these calculations were unrelated to the tumor. Cats with malignant tumors had a significantly higher local recurrence rate (44%), similar to previous findings,6 and a significantly shorter median survival time of 855 days than cats with nonmalignant disease, but a 2-year survival rate of 75%. The high local recurrence rate suggested that adjunctive treatment options, particularly radiation therapy, should be considered following VBO for cats with malignant middle ear disease.
Limitations of the study reported here were related to its retrospective nature in that imaging interpretation, surgical technique, postoperative management, and follow-up were not uniform. No histologic evaluation was performed for a large proportion (30.0% [75/289]) of procedures, and results may have been different had all cats received a histopathologic diagnosis. A strength of the study was inclusion of data from multiple institutions to allow a large sample size; however, this also allowed greater variability in the methods of treatment and management of cases.
Regardless of any limitations, the present study revealed that cats treated with single-stage bilateral VBO had significantly higher risks of respiratory complications and surgery- and disease-related death than cats treated with unilateral VBO or staged bilateral VBO. On the basis of these findings, we recommend that cats with bilateral middle ear disease be treated with staged bilateral VBO rather than single-stage bilateral VBO. Cats that already had Horner syndrome, head tilt, and facial nerve paralysis were more likely to have these conditions permanently after VBO than were cats without these conditions before surgery. A thorough preoperative physical and cranial nerve examination should therefore be performed for cats with middle ear disease to identify preexisting neurologic conditions, if present, and inform owners of the risk of postoperative complications accordingly.
Acknowledgments
The authors declare that there were no conflicts of interest.
ABBREVIATIONS
| CI | Confidence interval |
| TECA-LBO | Total ear canal ablation and lateral bulla osteotomy |
| VBO | Ventral bulla osteotomy |
Footnotes
SAS, version 9.4, SAS Institute Inc, Cary, NC.
References
1. Fan TM, de Lorimier LP. Inflammatory polyps and aural neoplasia. Vet Clin North Am Small Anim Pract 2004;34:489–509.
2. Muilenburg RK, Fry TR. Feline nasopharyngeal polyps. Vet Clin North Am Small Anim Pract 2002;32:839–849.
3. Anderson DM, Robinson RK, White RAS. Management of inflammatory polyps in 37 cats. Vet Rec 2000;147:684–687.
4. Faulkner JE, Budsberg SC. Results of ventral bulla osteotomy for treatment of middle ear polyps in cats. J Am Anim Hosp Assoc 1990;26:496–499.
5. Kapatkin AS, Matthiesen DT, Noone KE. Results of surgery and long-term follow-up in 31 cats with nasopharyngeal polyps. J Am Anim Hosp Assoc 1990;26:387–392.
6. Trevor PB, Martin RA. Tympanic bulla osteotomy for treatment of middle ear disease in cats: 19 cases (1984–1991). J Am Vet Med Assoc 1993;202:123–128.
7. Smeak DD. Management of complications associated with total ear canal ablation and bulla osteotomy in dogs and cats. Vet Clin North Am Small Anim Pract 2011;41:981–994.
8. Anders BB, Hoelzler MG, Scavelli TD, et al. Analysis of auditory and neurological effects associated with ventral bulla osteotomy for removal of inflammatory polyps or nasopharyngeal masses in cats. J Am Vet Med Assoc 2008;233:580–585.
9. London CA, Dubilzeig RR, Vail DM, et al. Evaluation of dogs and cats with tumors of the ear canal: 145 cases (1978–1992). J Am Vet Med Assoc 1996;208:1413–1418.
10. Bacon NJ, Gilbert RL, Bostock DE, et al. Total ear canal ablation in the cat: indications, morbidity, and long term survival. J Small Anim Pract 2003;44:430–434.
11. Sano T, Nishmura R, Mochizuki M, et al. Clinical usefulness of propofol as an aesthetic agent in dogs and cats. J Vet Med Sci 2003;65:641–643.
12. Little CJL, Lane JG. The surgical anatomy of the feline bulla tympanica. J Small Anim Pract 1986;27:371–378.
13. Barlow CM, Root WS. The ocular sympathetic pathway between the superior cervical ganglion and the orbit of the cat. J Comp Neurol 1949;91:195–207.
