Retrospective study investigating causes of abnormal respiratory noise in horses following prosthetic laryngoplasty
Reasons for performing the study: To investigate causes of respiratory noises in horses following prosthetic laryngoplasty (with or without a ventriculocordectomy) and to examine potential associations between degree of arytenoid abduction and the presence of other upper respiratory tract (URT) abnormalities, including right-sided collapse.
Methods: Clinical records and dynamic videoendoscopic recordings were examined from horses presented between 1995 and 2010 for investigation of respiratory noise during exercise, following a prosthetic laryngoplasty (+/− a ventriculocordectomy). Relationships between the degree of left arytenoid cartilage abduction and the presence of other URT disorders were investigated.
Results: Thirty horses matched the inclusion criteria. All horses had previously undergone a prosthetic laryngoplasty in different hospitals and in 63% (19/30) of these horses a left-sided ventriculocordectomy had also been performed. The majority of cases (87%) had multiple respiratory abnormalities and only 13% had a single URT disorder. Palatal dysfunction was the most common diagnosis (83%), followed by axial deviation of the aryepiglottic folds (60%) and vocal cord collapse (43%). The right arytenoid cartilage was fully abducted in all cases and no statistically significant correlation between the degree of left arytenoid abduction and any other URT disorders was detected.
Conclusions: Multiple forms of dynamic URT disorders were diagnosed in horses that presented with abnormal respiratory noise following laryngoplasty. There was no association between degree of left arytenoid abduction and other URT abnormalities. Furthermore, horses with suboptimal left arytenoid cartilage abduction were not predisposed to right-sided laryngeal collapse.
Clinical relevance: Our results demonstrate the fundamental role of dynamic endoscopy in correctly diagnosing dynamic airway collapse in horses that have undergone surgical treatment of the upper respiratory tract.
Recurrent laryngeal neuropathy (RLN) is a common cause of abnormal respiratory noise and poor performance in athletic horses [1–3]. The condition arises due to a progressive degenerative distal axonopathy  affecting both recurrent laryngeal nerves . However, the left side is more severely affected. Clinically, the condition results in collapse of the left arytenoid cartilage and associated vocal fold as a result of atrophy of the left cricoarytenoideus dorsalis muscle.
The prevalence of the disease varies according to the diagnostic criterion and type of horse. In Thoroughbreds it is estimated to be 0.3–3% [6–9]. However, it is higher among other breeds such as draught horses, where it is described to affect 24–46% [10,11]. Furthermore, subclinical RLN is reported to affect 40–100% of horses [12–15].
A number of surgical treatments for RLN have been described, including prosthetic laryngoplasty, ventriculectomy, ventriculocordectomy, laser ventriculectomy, partial, total and subtotal arytenoidectomy and laryngeal reinnervation . Prosthetic laryngoplasty (first described by Marks et al. in 1970)  remains the treatment of choice for athletic horses in which airway obstruction and exercise intolerance are the primary concern .
Success rates for prosthetic laryngoplasty vary widely among studies. Higher success rates (approximately 90%) are reported for horses that perform at submaximal exercise level than for racehorses (48–72%) [18–21]. There are also differences among studies with respect to the definition of ‘success’ and whether this relates to performance or reduction of the respiratory noise . For show horses, the persistence of a respiratory noise is more important as this may result in disqualification from the ring [22,23]. In the case of an eventer or racehorse, however, a respiratory noise is of no significance unless it is associated with impaired performance (e.g. decreased speed) .
Laryngoplasty alone does not entirely eliminate the abnormal respiratory noise, as much of the sound arises from movement of air over the ventricle . A left-sided ventriculocordectomy is most effective at reducing respiratory noise [3,25]. However, Hawkins et al.  reported that respiratory noise was present both in horses that had laryngoplasty alone and those that had laryngoplasty with ventriculectomy, suggesting that continued respiratory noise might also occur for other reasons. Robinson et al.  reported that laryngoplasty together with left-sided ventriculocordectomy may not completely resolve respiratory noise in some horses owing to billowing of the contralateral ventricle and vocal fold. Hence, it has since been suggested that bilateral ventriculocordectomy may be necessary for optimal noise reduction [25,26].
Post operative complications are common following laryngoplasty [19,21] and may also lead to continued production of abnormal respiratory noise and/or exercise intolerance. Implant failure (by suture pull-out or prosthesis breakdown) has been reported in a small proportion (5%) of cases during the first week after surgery , and progressive loss of abduction may occur over time in many horses [21,27–30]. A recent experimental study hypothesised that suboptimal post operative abduction may predispose horses to other forms of collapse due to increases in negative airway pressure, particularly in the regions of the right aryepiglottal fold and both vocal folds . However, this has yet to be confirmed in clinical cases.
Respiratory noise is not specific to any single type of upper airway obstruction, and hence repeat endoscopic examination is indicated in horses that continue to make abnormal post operative respiratory noise, in order to determine the cause. To date, only one such study has been reported . The results showed that respiratory noise was not specific for arytenoid collapse and that most horses had complex upper respiratory tract (URT) obstructions. However, it was not possible to determine whether or not there was an association between the degree of abduction and other forms of URT collapse, as no horses with grades 1 or 2 abduction were included.
The aims of this study were to review cases referred to our clinic with respiratory noise following laryngoplasty with or without ventriculocordectomy and to investigate possible associations between the degree of left-sided arytenoid cartilage abduction and other forms or URT collapse, including right-sided URT collapse.
Materials and methods
Clinical records and corresponding videoendoscopic recordings were retrieved for horses referred for investigation of abnormal respiratory noise following laryngoplasty, during a 15 year period (1995–2010).
Examination of the clinical records provided details about the performance during the pre- and post operative periods and when, where and which surgical procedure(s) were performed. All patients underwent a full clinical examination at rest and details were recorded in the case card. A dynamic endoscopic examination (either on the treadmill or in the field, using an overground telemetric endoscope) was then performed as previously described [32,33]. Horses were exercised to fatigue or until a definitive diagnosis was reached. The video recordings were reviewed to confirm the diagnosis and the following criteria were assessed:
- 1) Grade of left and right arytenoid abduction. This was graded 1–5 according to the scale previously described by Dixon et al. .
- 2) Presence/absence of vocal cord collapse (VCC) and whether this was unilateral or bilateral.
- 3) Presence/absence of axial deviation of the aryepiglottic folds (ADAF) and whether this was unilateral or bilateral.
- 4) Presence/absence of other forms of URT collapse.
The data were stored in an electronic database (Excel 2003, Microsoft) and analysed using PASW statistic 17. Chi-square tests were used to examine associations between the presenting complaint and athletic discipline (racing vs. nonracing) and between the type of surgery performed (laryngoplasty alone vs. laryngoplasty plus ventriculocordectomy) and presenting complaint. Chi-square tests were also used to investigate associations between the type of surgery performed and the presence of VCC, and between the grade of left laryngeal abduction and the presence of other URT disorders and, in particular, associations with right-sided laryngeal collapse. Statistical significance was set at P<0.05.
Thirty horses (26 geldings and 4 mares) matched the inclusion criteria, presenting with a persistent respiratory noise following laryngoplasty and having a complete case record inclusive of a videoendoscopic recording of their exercise test. Mean age on presentation was 7.2 years (range 3–14) and athletic use included 8 eventers (one Irish Draught 2 pure English Thoroughbreds and 5 Thoroughbred crossbreds) and 22 Thoroughbred racehorses (5 flat and 17 National Hunt jump horses).
The median interval between surgery and post operative exercise testing was 15.5 months (range 3–84 months). Eleven (37%) cases had undergone a left-sided prosthetic laryngoplasty procedure alone, in the remaining 19 (63%) cases, a left ventriculocordectomy and right ventriculectomy (leaving the right vocal fold intact) had also been performed in the remaining 19 (63%) cases. The presenting complaint in all cases was of abnormal respiratory noise, and in 18 (60%), poor performance was also reported. There was no significant difference in horses presenting with poor performance (in addition to the persistent respiratory noise) between types of horse (racehorse vs. eventer) (P = 1.000) or between cases that had laryngoplasty alone or laryngoplasy in combination with ventriculocordectomy (P = 0.643).
Exercising endoscopy findings
During high intensity exercise, abduction of the left arytenoid cartilage was graded 1 in 3 (10%) cases, 2 in 7 (23%) cases, 3 in 8 (27%) cases, 4 in 10 (33%) cases and 5 in 2 2 (7%) cases. All horses maintained maximal abduction (grade 1) of the right arytenoid cartilage during exercise.
All horses had some form of dynamic airway collapse during exercise (Table 1). Twenty-six (87%) cases had a combination of multiple abnormalities. The most common abnormalities were palatal dysfunction (83%), ADAF (60%) and VCC (43%). A single form of URT collapse was diagnosed in only 4 (13%) cases (of which 2 had palatal dysfunction, one right-sided VCC and one right-sided ADAF).
Table 1. Dynamic endoscopic findings after laryngoplasty
|LP (n = 11)||10||1||5||9*||1||2||5||1||0||7|
|LP + left VCC and right V (n = 19)||16||3||7||16**||0||2||3||0||2||8|
|Total (n = 30)||26||4||12||25||1||4||8||1||2||15|
Not surprisingly, there were significant differences with respect to VCC between horses that underwent laryngoplasty alone and those that also had ventriculocordectomy (P = 0.013). Of the 11 horses that underwent laryngoplasty alone, 8 (73%) showed some form of VCC (this was left-sided in one case, right-sided in 2 and bilateral in 5). Of the 19 horses that underwent laryngoplasty in combination with left ventriculocordectomy and right ventriculectomy, 5 (26%) had collapse of the right vocal cord and 3 of these also had some collapse of the remnants of the left cord.
In total there were 6 cases of right-sided laryngeal collapse. These included 2 cases of right-sided ADAF (which had undergone a left-sided ventriculocordectomy) and 4 of VCC (of which 2 had previously undergone a left ventriculocordectomy). The other 2 presented a concomitant marked bilateral ADAF). There was no statistically significant association between left laryngeal grade and the presence of other forms of URT collapse including palatal dysfunction (P = 0.461), bilateral ADAF (P = 0.858) or right-sided ADAF (P = 0.195) and bilateral VCC (P = 0.388) or right-sided VCC (P = 0.762).
The findings of our study confirm that the cause of respiratory noise following laryngoplasty may be attributed to the collapse of various structures within the URT and is frequently multifactorial, as previously reported by Davidson et al. .
We found that respiratory noise was not necessarily associated with implant failure and dynamic arytenoid cartilage collapse. In this study, 33% of horses had good arytenoid abduction (grades 1 or 2) and a further 27% had grade 3 abduction. Previously, Davidson et al.  hypothesised that the absence of grades 1 and 2 cases in their cohort might be due to the fact that these horses do not suffer from continued problems after surgery or that most horses do not maintain such levels of abduction after surgery. However, our findings refute this and indicate that other forms of dynamic airway collapse are responsible for respiratory noise in these horses.
The clinical implications of grade 3 abduction post surgery remain uncertain, particularly for racehorses. An experimental study by Rakesh et al.  suggested that horses with grade 3 abduction would experience a 6% decrease in minute volume. However, Barakzai et al.  reported no difference in performance in National Hunt racehorses with grades 1–3 abduction. Insufficient data exist from horses with grades 4 or 5 abduction to draw conclusions. Derksen  suggested that stabilisation of thearytenoid cartilage is more important than the degree of arytenoid abduction in preventing airflow limitation. However, more recent work suggests that although normal ventilation may be maintained during submaximal exercise, these horses will have further impaired airflow that may compromise performance when undertaking high intensity exercise .
All horses in this study with grades 4 or 5 arytenoid cartilage collapse also had other forms of dynamic airway collapse. Overall, complex dynamic collapse was identified in 87% of the horses. These findings are in accordance with previous reports that URT collapse may be multifactorial in many cases [32,36–38]. It remains unclear whether these horses in fact had multifactorial dynamic collapse prior to laryngoplasty or whether they developed additional URT collapse following surgery. Many horses in this study had surgery performed elsewhere and there was no history as to whether exercising endoscopy was performed beforehand. However, normal practice suggests that such horses would have had a confirmed diagnosis of laryngeal hemiplegia made on a resting endoscopy. The time between surgery and re-examination was extremely variable in this study (3–84 months). It is likely, therefore, that some horses were indeed afflicted with multiple forms of dynamic collapse prior to surgery, whereas others may have developed additional forms of URT collapse over time. From the results obtained in this study and that of Davidson et al. , we would suggest that there may be a benefit in performing dynamic endoscopy in all cases, regardless of the obvious presence of laryngeal hemiplegia at rest, in order to confirm or refute the presence of further concomitant URT disorders before performing surgery.
In our study palatal dysfunction was the most common finding (in 83% cases) and ADAF was observed in 60% of cases. In other studies of dynamic URT collapse, palatal dysfunction and ADAF were the most common abnormalities detected [32,38,39]. Axial deviation of the aryepiglottal folds is commonly found to occur in combination with other forms of URT collapse [32,38–40]. An association between ADAF and palatal instability has been reported [33,40] and a relationship between ADAF and dynamic laryngeal collapse has also been identified . These authors concluded that secondary forms of URT collapse were likely to arise due to the Bernoulli principle (which states that as the speed of the moving fluid – liquid or gas – increases, the pressure within it decreases, creating a suction effect towards the lumen of the airways which will cause aberrant movements of the adjacent soft tissues), rather than due to a generalised neurological disorder.
Potential associations between RLN and palatal dysfunction have not been fully investigated, although early reports suggested that these conditions may occur concurrently . More recently, Strand and Skjerve  reported a trend for PI to be associated with dynamic laryngeal collapse. There is substantial evidence that many horses develop palatal dysfunction in the absence of laryngeal dysfunction [32,38,42]. However, whether or not horses with RLN might be predisposed to palatal dysfunction remains uncertain. It is possible that differences may be dependent on an individual's change in breathing strategy in response to dynamic airway collapse. Using a computational model, Rakesh et al.  found that when a decrease in laryngeal diameter was accompanied by driving pressure compensation, the nasopharynx was subjected to more negative pressures predisposing it to nasopharyngeal collapse. This did not occur when there was a decrease in laryngeal diameter without driving pressure compensation, although further laryngeal collapse was likely under both conditions.
Vocal cord collapse was identified in 43% horses, but unlike the study by Davidson et al. , it was not the most common finding. This may have been because many of the horses in our study had undergone a bilateral ventriculectomy in addition to a left cordectomy. Not surprisingly, VCC was observed more frequently in horses that had only laryngoplasty performed, and this is a well recognised cause of persistent respiratory noise following laryngoplasty [2,3,20,22,25,28,43–45]. Interestingly, however, 45% of cases did not show left VCC, despite previous suggestions that ACC is invariably accompanied by ipsilateral VCC .
In this study we found no association between the degree of post surgical abduction and the presence of other forms of dynamic collapse, specifically right-sided laryngeal collapse. Previously, Rakesh et al.  had suggested that structures of the right hemilarynx might be predisposed to collapse in horses with suboptimal arytenoid abduction, owing to increased negative pressure in these regions. However, neither our study nor that of Davidson et al.  found evidence to confirm this. In our study, bilateral VCC or ADAF was more frequently observed. We found evidence of unilateral right-sided collapse of the vocal cord or aryepiglottal fold in only 6 cases (of which 2 cases with right sided vocal cord collapse had previously undergone a left ventriculocordectomy). Other studies have suggested that right-sided ADAF occurs more frequently [30,47]. However, it is possible that this may be an artefact due to endoscope position, as when the endoscope is positioned in the right nostril, right-sided collapse will be more apparent.
Finally, we found no evidence that horses with RLN develop right-sided ACC over time, despite the fact that RLN has been confirmed to be a bilateral mono neuropathy . Left-sided progression of RLN has been previously reported in approximately 15% cases over a median of 12 months . Nevertheless, all horses in our study maintained grade 1 right arytenoid cartilage abduction throughout exercise (even in cases presented several years post operatively), when the disease was obviously established on the left side.
In conclusion, post operative noise in horses that had abnormal respiratory noise following prosthetic laryngoplasty was not associated with left arytenoid collapse in most patients, but was frequently associated with multiple forms of dynamic collapse within the URT. No association was found between degree of arytenoid collapse and the presence of additional forms of URT collapse. Furthermore, we found no evidence to suggest that horses with suboptimal left arytenoid abduction might be predisposed to right-sided collapse. These findings further support the use of dynamic endoscopy in the diagnosis of abnormal respiratory noise before surgery, even when laryngeal hemiplegia is detected at rest, in order to rule out additional forms of airway collapse.
Conflicts of interest
No competing interests have been declared.
Sources of funding
This work was supported by a research grant from The Horse Trust, Buckinghamshire, UK and the University of Bristol, Veterinary Department, Langford, Somerset, UK.
We aknowledge the precious support given by all the staff of the Equine Veterinary Clinic of the University of Bristol during the examination of the horse specimens, my father and all that helped in the writing and collecting of data.
Francesca Compostella advanced the study design, collected the data, performed some of the examinations and surgery and wrote the main manuscript. Samantha Franklin perfected the study design, performed all the examinations, helped analyse the data and corrected the statistical analysis and main manuscript. Henry Tremaine performed some of the surgical procedures, helped analyse the data and corrected the manuscript.