Airway injury associated with cervical bite wounds in dogs and cats: 56 cases

Summary Objective: To investigate the frequency of airway injury and damage to other vital structures associated with cervical bite wounds in dogs and cats and the implications for management and outcome. Methods: A retrospective search of electronic patient records was used to identify dogs and cats suffering cervical bite wounds that were presented to a large multidisciplinary veterinary hospital over a four year period. Results: Complete records were available for 55 animals, with one animal suffering two separate injuries. Fourteen animals (25%) had injuries to vital structures, including airway injury in nine (17%) which was surgically confirmed and treated in six (11%). Airway injuries were associated with either subcutaneous or mediastinal emphysema in all affected animals. Other structures injured included the jugular vein, pharynx, oesophagus and spine. Airway injuries were treated with primary repair in five animals and a fasciomuscular patch in one. Temporary tracheostomy was performed in three animals. Median duration of hospitalisation was one day (0-19) with 53 animals (54 cases) (96%) surviving to discharge. Long-term follow-up (16-114 months) revealed that 43 of 49 animals were alive with six that died due to unrelated reasons. Clinical significance: Cervical bite wounds are associated with significant injury to vital structures. Up to 17% of animals may have injury to their airway. Identification and treatment of airway injury is vital and was associated with an excellent outcome in six animals.

cutaneous or mediastinal emphysema in all affected animals. Other structures injured included the jugular vein, pharynx, oesophagus and spine. Airway injuries were treated with primary repair in five animals and a fasciomuscular patch in one. Temporary tracheostomy was performed in three animals. Median duration of hospitalisation was one day (0-19) with 53 animals (54 cases) (96%) surviving to discharge. Long-term follow-up (16-114 months) revealed that 43 of 49 animals were alive with six that died due to unrelated reasons. Clinical significance: Cervical bite wounds are associated with significant injury to vital structures. Up to 17% of animals may have injury to their airway. Identification and treatment of airway injury is vital and was associated with an excellent outcome in six animals.

Introduction
Bite wounds are common injuries in dogs and cats, representing 11% of 1000 cases admitted to a 24 hour emergency service in one study (1). Bite wounds are widely recognised as being a management challenge as they produce a combination of shearing, tensile and compressive forces manifesting as tissue lacerations, tears and punctures (2). Indeed, animals with multiple bite wounds are at risk of developing systemic inflammatory response syndrome (SIRS) due to extensive tissue trauma, infection and devitalised tissue (2)(3)(4)(5)(6)(7). In many instances the bites are inflicted by a larger attacker, often referred to as a 'big dog -little dog' injury, in which the victim is grasped and shaken violently, potentially leading to additional trauma. Mortality rates from bite wounds are approximately seven percent, with between two and five percent of animals being euthanatized (8,9).
Between 22% and 27% of bite wound cases involve the cervical region with between 25% and 38% involving multiple regions (1,8,9). The cervical region was found to be the third most commonly injured area in two studies (8; Parent C. Unpublished bite wound survey, University of Pennsylvania, School of Veterinary Medicine, 1993). Bite wounds were the most common cause of cervical injury in one study of 1000 dogs and cats admitted to a 24 hour emergency service (1). Cervical bite wounds are associated with higher rates of mortality and complications (8). Twenty percent of animals with wounds in the cervical region alone or as part of multiple wounds died compared to a seven percent overall mortality rate (8). Two animals (2%) in this study were euthanatized with both having suffered bite wounds to their cervical region as part of multiple bite wounds (8). Non-fatal consequences including infection, dehiscence, excessive swelling or drainage occurred in 25% of cases with wounds to their cervical region compared to a 17% overall non-fatal complication rate (8).
The various subcutaneous manifestations of bite wounds are not always immediately obvious due to skin wounds often appearing to be small and innocuous, yet the inherent mobility of small animals skin allows extensive movement of teeth in the subdermal tissues (10). This may be particularly relevant to bite wounds to the cervical region where a number of vital structures such as the carotid arteries, jugular veins, vagosympathetic trunk, larynx, trachea, oesophagus, and cervical spine lay relatively superficially. This may explain the greater mortality and complication rate seen in animals with cervical bite wounds.
It is clear that cervical bite wounds and in particular those associated with airway injury, represent a unique challenge in small animal veterinary practice. There are limited data on cervical bite wounds, particularly those pertaining to airway injury (10). The aim of the current study was to report a large series of cervical bite wounds in dogs and cats, and in particular to assess the frequency of airway injury, management, frequency of complications, and outcome.

Methods
A retrospective search of electronic patient records was used to identify dogs and cats suffering cervical bite wounds which were presented at The Queen Mother Hospital for Animals at The Royal Veterinary College (Hatfield, England) between June 2006 and August 2010. Animals were included if their records were complete and contained the following information: species, age, sex, weight, whether seen as a referral or first opinion emergency, area injured, type of wound, structures damaged, and management. The region of injury was described as ventral, dorsal, or left or right lateral cervical region as recorded in the clinical notes. Types of injury were defined as lacerations and punctures as recorded in the clinical notes. Management was defined as 'non-surgical' if no surgical intervention was performed, 'local surgical' if local surgical exploration and wound management was performed, and 'full surgical exploration' if ventral midline cervical exploration was performed. Follow-up information was obtained by telephone interview with the owner and the primary veterinary surgeon.
A control population of all dogs and cats was formed from emergency admis-sions to the hospital during the same period without cervical bite injury.
Analysis was performed using a statistical software package a . Continuous data were assessed graphically for normality. Median and range were reported for skewed data and compared with the Mann-Whitney U test. Mean and standard deviation were reported for normally distributed data and compared with an independent samples t-test. Significance was set at p <0.05.

Results
Complete records were available for 55 animals with 56 cervical bite injuries; 38 dogs and 17 cats with one cat having suffered two separate injuries on two separate occasions. Sixteen were referred to the emergency and critical care service (ECCS) and 40 were seen by the first opinion out of hours service (OOHS). Dogs had a median age of 72 months (range: 4 -156) and a median weight of 14.9 kg (range: 2.3 -40.0). Eighteen dogs were female and 20 were male. Cats had a median age of 48 months (range: 15 -168) and a median weight of 4.41 kg (range: 3.5 -6.0). Four cats were female and 13 were male. All animals were neutered except for one male cat, six male dogs and four female dogs.
During the study period 15,608 animals were admitted by the OOHS or referred to the ECCS. Of this total amount, 10,246 were dogs and 4,775 were cats. Dogs had a median age of 76 months (range: 1 -240) and cats had a median age of 84 months (range: 1 -324). Of the 8,797 dogs of known body weight, median body weight was 21.0 kg (range: 0.2 -115.0). Of the 1,511 cats of known body weight, mean body weight was 4.35 kg (SD = 1.27). There was no statistically significant difference between dog or cat cases and controls for age and weight.
Cervical bite wounds in dogs and cats represented four in every 1000 of combined emergency admissions to the OOHS and ECCS over the time period of our study.
According to the clinical histories, the altercations were witnessed in 29 of the dogs and two of the cats. All but one of the aggressors were dogs, with five Staffordshire Bull Terriers, two Rottweilers, and one each of an Akita, a Lurcher, a Labrador Retriever, a Golden Retriever, a Mastiff, a Dobermann, a Jack Russell Terrier and an American Bull Terrier. Five attacks were by multiple dogs. One dog was witnessed being attacked by a cat.
Thirty-one animals (56%) were managed with non-surgical wound management. Seventy-two percent of first opinion cases and 13% of referrals were managed non-surgically. Thirteen cases (24%) had local surgical management. Ten (18%) were treated with full surgical exploration. All of those cases treated with full surgical exploration were ECCS referrals. Of the cases with local surgical management, 11/13 had a drain placed (one active, 10 Penrose). Of the cases that had full surgical exploration, six out of 10 had a drain placed (two active, four Penrose).
Fourteen animals (25%) had injured vital structures including airway injury, which was documented at surgery in three out of 38 dogs and three out of 17 cats, three of which had injury to their trachea and three had injury to their larynx. There was no significant difference in the proportion of cats and dogs with cervical bite wounds involving the airway. Two dogs had tracheal injury and one dog had laryngeal injury. The dogs were all small breed with a median weight of 10.00 kg (range: 6.18 -12.24). Dogs with cervical bite wounds without injury to their airway had a median weight of 15.0 kg (range: 2.3 -40.0). There was a trend towards dogs with cervical bite wounds with airway injury weighing less than dogs without airway injury but this was not significant (p = 0.07). Two cats had laryngeal injury and one cat had tracheal injury. See ▶ Table 1 for further details of the animals with airway injury. All animals with airway injury confirmed at surgery had either subcutaneous or mediastinal emphysema, two of which had airway injury without skin penetration. Airway injury was suspected in a further three animals with pneumomediastinum, one of which had extensive bite wounds to multiple regions and died of a PASW Statistics 18.0.0: Education SPSS (UK), Limited IBM, Woking, UK septic shock before the source of pneumomediastinum could be identified. In the other two animals, exploratory surgery did not identify the source of the pneumomediastinum or was precluded by cost constraints. Both of these animals survived to discharge without clinical signs related to pneumomediastinum, but were lost to follow-up. Airway injuries were treated with primary repair with synthetic, absorbable, monofilament suture in five animals and a fasciomuscular patch in one. Primary repair of a thyroid cartilage fracture in one cat required revision surgery due to excessive granulation tissue formation and laryngeal distortion. Animals with airway injuries confirmed at surgery are summarised in ▶ Table 1.
Temporary tracheostomy was performed in three animals; two cats and one dog. Temporary tracheostomy was performed in these animals at the time of exploration surgery due to concern over airway patency in the postoperative period. One cat had a thyroid cartilage fracture which had a temporary tracheostomy tube placed at the initial full surgical exploration which was then removed 36 hours later. The cat then became dyspnoeic two days later as the tracheostomy site began to heal requiring the temporary tracheostomy tube to be replaced. Laryngeal examination revealed excessive intraluminal granulation tissue which was resected and the thyroid cartilage fracture revised. The temporary tracheostomy tube was removed 24 hours later. The second cat had an epiglottic tear and separation of the epiglottic and thyroid cartilages. Primary repair of the laryngeal cartilage separations was performed at full surgical exploration and a temporary tracheostomy tube was placed. This was removed the following day. The dog that required temporary tracheostomy did not suffer an airway injury but had a right jugular avulsion and extensive cervical inflammation with no other structures damaged. Full surgical exploration was performed and a temporary tracheostomy tube placed at the end of surgery. The temporary tracheostomy tube was removed three days postoperatively.
Other vital structures injured included the jugular vein (2/56), hyoid apparatus (2/56), left recurrent laryngeal nerve (2/56), mandibular salivary gland (1/56), pharynx (1/56), oesophagus (1/56), thyroid (1/56), parathyroid (1/56), vagosympathetic trunk (1/56), occipital condyle (1/56), tympanic bulla (1/56) and spine (1/56). Both jugular injuries were unilateral and were treated with ligation and resection of the damaged portion. The spinal injury was small, with fractures of the transverse processes of cervical vertebrae 3-5. These were managed conservatively, as were fractures to the tympanic bulla and fractures to the occipital condyle in two other animals. Injuries to the recurrent laryngeal nerve or vagosympathetic trunk were not repaired. Both animals with injuries to the recurrent laryngeal nerve had left sided laryngeal paralysis, which was diagnosed by observation of left sided laryngeal paralysis at intubation for general anaesthesia. Neither animal showed clinical signs related to laryngeal hemi-paralysis so no directed treatment was indicated. It  was not apparent from the medical records whether this was assessed in all cases. All other injuries to vital structures were treated by resection of devitalised tissue and primary repair.
Median duration of hospitalisation was one day (range: 0 -19) with 54/56 cases (96%) (or 53 animals) surviving to discharge. The two animals that died had multiple bite wound injuries to multiple regions of their bodies, with both animals suffering cardiopulmonary arrest as a consequence of sepsis and SIRS.
Follow-up was available for 49 out of the 55 animals (mean: 37; range: 16 -114 months); the remaining five animals that survived to discharge were later lost to follow-up (note as one cat suffered two injuries the total number of animals was 55 rather than 56). Six animals had died or were euthanatized for reasons unrelated to their cervical bite wound. Six animals had non-fatal complications of their cervical injury. One animal had a recurrent seroma that spontaneously resolved 17 days postdischarge and four animals required extended antibiosis prescribed for suspected ongoing infection. One eight-year-old Beagle developed signs of moderate cervical pain four weeks post-discharge which was documented as a calcified intervertebral disc on radiography seven months post-discharge. This dog had been witnessed being picked up by its cervical region and shaken. All animals which required temporary tracheostomy survived to discharge and were alive at follow-up.

Discussion
This study reports a relatively large number of cervical bite wounds to dogs and cats. In our study, bite wounds to the cervical region represented four in every 1000 (0.4%) of combined admissions to the OOHS and referrals to the ECCS compared to 2.75% of emergency cases reported in a previous study (1). However, the previous study was performed in a 24-hour first opinion emergency hospital whereas although the ECCS Service accepts referrals 24 hours a day, the OOHS emergency service is only an out-of-hours service.
Cervical wounds can be challenging to manage due to the number of vital structures in relatively superficial locations. Previous data have shown that dog bite wounds commonly occur in the cervical region and increase the likelihood of death, euthanasia or complications (1,8). In the current study, despite a relatively large number of animals (14/56; 25%) with damage to vital structures only two (4%) died as a result of their injuries. This compares favourably with a previous study that described a 20% mortality rate from cervical bite wounds (8). The previous study described a population of first opinion and emergency referral cases and this may have resulted in the difference between the current study. The previous study was performed in 1989 and it is possible that advances in anaesthesia, surgery and postoperative care may have improved survival in these dogs. The two animals that died had multiple bite wound injuries to multiple regions of their bodies with both animals suffering cardiopulmonary arrest as a consequence of sepsis and SIRS. It is likely that the mortality in these two animals was related to the multiple wounds and extensive tissue trauma rather than the cervical injuries per se.
Airway injury was documented in six (11%) animals and suspected in a further three animals leading to a possible incidence of 17% of cervical bite wounds resulting in airway injury. It was interesting that two animals suffered airway injury without skin penetration. Injury to the trachea by blunt, non-penetrating trauma is described although it is uncommon (11). Thus an airway injury cannot be ruled out on the basis of the absence of a penetrating wound. It is important that an airway injury is promptly identified and treated. Subcutaneous emphysema of the neck detected on physical examination should raise the concern of airway injury. Cervical and thoracic radiographs should be considered in all animals suffering cervical bite wounds as air in the fascial planes of the neck or a pneumomediastinum were associated with airway injury in this study. All animals with these radiographic findings should have full surgical exploration via ventral cervical mid-line approach, in order to identify and repair the injury.
Temporary tracheostomy was performed in two cats and one dog. In all three cases, temporary tracheostomy was performed at the end of ventral midline exploration due to concerns that postoperative inflammation may lead to airway compromise and dyspneoa. In our opinion, temporary tracheostomy should be performed if there are any concerns about postoperative airway patency. The use of temporary tracheostomy was not associated with additional complications or mortality although the number of animals involved was small.
It has been suggested that tracheal or laryngeal trauma, particularly with disruption to large areas of the mucosa, can be associated with scarring and stenosis (11). Repair of all six airway injuries by primary repair or fasciomuscular patch was associated with an excellent outcome with one animal needing revision surgery, and all animals surviving to discharge with no complications reported at follow-up. Although the injuries described in this report were relatively small, they responded well to simple surgical repair techniques and there was no scarring or stenosis in five of the six animals. In one cat, excessive granulation tissue within the laryngeal lumen required revision surgery. However, with more accurate tissue apposition this problem did not recur.
Injuries to other vital structures in the neck occurred in eight animals. Two animals suffered documented damage to the recurrent laryngeal nerve resulting in unilateral laryngeal paralysis. Assessment of laryngeal function at induction of anaesthesia should be performed in animals with severe cervical injuries or evidence of inspiratory stridor preoperatively. Only one animal sustained an injury to the spinal column. Bite wounds as a cause of spinal trauma has a prevalence of 3.8 -14.3 % and are the second most common cause of spinal trauma following road traffic accidents (11)(12)(13)(14)(15). Spinal injury can result from direct penetration and crushing and indirectly from picking up and shaking of the victim (16,17).
Accepted recommendations on bite wound management are to explore them to their full extent, debride devitalised tissues and flush copiously with sterile saline (2,8,9,10,18,19). This allows control of potential infection and identification of occult trauma to deeper tissues. Although we documented a high rate of injury to vital structures in our study, the majority of cases (80%) were managed with either non-surgical wound management or local surgical wound management. These animals had an excellent outcome as did those with more complicated injuries requiring full surgical exploration via ventral cervical midline approach and that were found to have injuries to vital structures. Of the 48 animals available for follow-up, six (11%) suffered non-fatal consequences. This was considerably less than a previously reported non-fatal complication rate of 25% (8). Unfortunately, the authors of this previous study did not expand on their treatment protocol to a full enough extent to allow comparison with our study. The difficulty in managing all bite wounds, and in particular cervical bite wounds, is selecting the appropriate level of intervention. It is important to note that full surgical exploration is mandatory if damage to deeper structures is suspected.
The authors had expected to find that dogs with cervical bite wounds would be below the median weight of the control population in support of the commonly perceived nature of the protagonists, i.e. 'big dog -little dog' . However, there was no significant difference between the weights of dogs with cervical bite wounds and controls, indicating that larger breed dogs were also at risk from these injuries. Interestingly, dogs with injury to their airway were all small breeds. This suggests that the like-lihood of more serious injury is greater in smaller animals.

Conclusion
This study highlights the high rate of injury to vital structures, in particular the airway, associated with cervical bite wounds. Cervical and thoracic radiographs should be considered in all animals suffering cervical bite wounds. Primary repair of airway injuries was associated with an excellent outcome in five out of six animals with the sixth animal having successful revision surgery following intra-airway granulation tissue formation and laryngeal distortion.