DISCUSSION
This manuscript describes a novel exercise-related (dynamic) disorder of the epiglottis induced during periods of poll flexion in harness racehorses. Compression of the base of the epiglottis appeared visually during the first phase of induced poll flexion (phase 2) in all 18 horses. In all cases, whether unilateral or bilateral, it appeared videoendoscopically as if there was extraluminal compression of the base of the epiglottis from the lateral nasopharyngeal wall (green arrows in figures) locally on the affected side(s), potentially caused by a change in position of the larynx/hyoid apparatus within the intermandibular space during poll flexion. In most cases the compression of the epiglottis was of a magnitude that it visually obscured the abducted vocal fold(s) on the affected side(s). In certain cases, the lateral margins of the epiglottis displayed progressive inward deviation with prolonged duration of exercise, potentially due to greater inspiratory luminal pressures in the region of the larynx.
Poll flexion during exercise, defined as a decreased angle between the mandible and ventral neck, affects upper airway function in horses competing in multiple athletic disciplines (Petsche et al. 1995; Strand et al. 2009; Davidson et al. 2011). Harness racehorses perform in a position of “high” poll flexion, due to the overcheck which prevents lowering of the head and neck and aids steering the horse tactically in races. These horses are gathered onto the bit with long reins and are restrained from premature exhaustion or galloping. Poll flexion during maximal exercise in clinically normal Standardbreds causes an increase in inspiratory impedance and greater negative inspiratory pressures (7-8 cm H2O difference), relative to when these horses are exercised in free head carriage (head and neck extended), creating a mild inspiratory obstruction (Petsche et al. 1995). Similar findings (6 cm H2O difference) have been documented in clinically normal elite Norwegian-Swedish coldblooded trotters during maximal exercise in poll flexion relative to maximal exercise with free head carriage (Strand et al. 2009). This physiologic event is explained by the Bernoulli principle and Venturi effect, occurring as a consequence of narrowing of the URT during poll flexion (Cehak et al. 2010). Narrowing of a region of the URT will lead to increased velocity of air through the constricted region and therefore lower intraluminal pressures according to Poiseuille’s Law (fluid dynamics notation). If a form of URT obstruction is present, local narrowing (collapse) of the region can become self-perpetuating with progressive lowering of inspiratory pressures over time and worsening of the obstruction. This has been demonstrated to occur in harness racehorses diagnosed with DLC (Strand et al. 2009).
Poll flexion can also induce or exacerbate both simple and complex forms of exercise-induced URT disorders reported in harness racehorses (Strand et al. 2004; Fjordbakk et al. 2008; Strand et al. 2012). Exacerbation of already existing URT disorders during poll flexion may be due to the increase in airway turbulence, impedance and intraluminal pressures occurring during poll flexion, while the pathophysiology behind induction of new URT disorders solely evident during poll flexion is likely more complex.
In addition to altering airway dynamics, poll flexion also causes anatomical conformational changes of the laryngohyoid apparatus which can be readily measured from lateral radiographs of the region (McCluskie et al. 2008; Fjordbakk et al. 2013). The larynx moves rostrally in relation to the hyoid apparatus during poll flexion; McCluskie et al. (McCluskie et al. 2008) demonstrated a rostral advancement of the ossified portion of the thyroid cartilage in relation to the thyrohyoid bone. Achieving this local conformation change is the basis of the “laryngeal tie-forward” procedure for treating iDDSP (Woodie et al. 2005).
Fjordbakk et al. (2013) demonstrated that horses affected by DLC showed a statistically significantly greater rostral advancement of the larynx in relation to the thyrohyoid bone during poll flexion compared to unaffected control horses. This rostral advancement within the intermandibular space resulted in external compression of the larynx as shown by the authors in form of a reduced laryngeal lumen width in affected horses measured ultrasonographically (Fjordbakk et al. 2013). Though DLC to this date has been diagnosed in a number of different breeds and disciplines, the aforementioned study and more recent research into genetics illustrate that URT disorders occurring solely during poll flexion often result from a combination of phenotypic predisposition along with the consequences of a rostral laryngohyoid displacement (Fjordbakk et al. 2013; Velie et al. 2020).
Epiglottic compression appears soon after inducing poll flexion by gathering the reins and driving the horse onto the bit, similar to how DLC visually appears. The authors have seen compression of the epiglottis manifesting most commonly as part of a DLC diagnosis, and epiglottic compression may be a variation of a complex array of exercise-induced laryngeal disorders seen only during phases of poll flexion. The occurrence of the disorder primarily, without concurrent DLC, has however not been described before.
Dynamic laryngeal collapse is a form of bilateral arytenoid cartilage and vocal fold collapse occurring during exercise in poll flexion in predisposed horses, not associated with recurrent laryngeal neuropathy (Fjordbakk et al. 2015; Ducharme 2016). It is most commonly diagnosed in NSCTs but is also reported to be common in the Colombian Criollo Paso breed (Joó et al. 2021) and other gaited horses that exercise in “high” poll flexion (Hanche-Olsen et al. 2010; Allen et al. 2011; McCarrel & Woodie 2015; Hackett & Leise 2018). In DLC, a phenotypic predisposition combined with rostral laryngohyoid advancement in the intermandibular space occurring during poll flexion leads to inward displacement of the arytenoid cartilages with secondary collapse of the vocal folds (Strand et al. 2004; Fjordbakk et al. 2008). The phenotypic, or conformational traits include: a rostrally positioned larynx relative to the vertical ramus of the mandible, a high head and neck position during exercise, a narrow intermandibular space and an energetic attitude when driven “onto the bit” (Strand et al. 2004; Velie et al. 2020). A rostrally positioned larynx, relative to the vertical ramus of the mandible, is a trait of many draft and pony breeds of horses (Hawe et al. 2001). In the cases presented here, the rostral laryngohyoid advancement caused by periods of poll flexion leads to compression of the base of the epiglottis while the arytenoid cartilage’s ability to abduct is unaffected. As opposed to horses diagnosed with DLC, horses showing epiglottic compression likely have more compression of a rostroventrally positioned articulation between the thyrohyoid bone and thyroid cartilage which is not interfering with the ability of the cricoarytenoideus dorsalis muscles to abduct the arytenoid cartilages, as seemingly occurs with DLC (Figure 5) (Strand et al. 2004; Fjordbakk et al. 2008).
The majority of horses in this study presented for investigation of poor performance and/or abnormal respiratory noise (Table 1), however one horse (nr. 3) presented after physical collapse during a pari-mutuel race. This horse had visually the most severe degree of URT obstruction of all 18 horses in this review during HSTE with both compression of the base of the epiglottis and dorsomedial deviation of the epiglottic margins. This highlights the importance of examining harness racehorses during exercise using bit, bridle, long reins and overcheck to induce different head and neck angles similar to what these horses experience under competition if one is to fully understand the clinical condition. This can be accomplished using high-speed treadmill videoendoscopy, as described here, or utilising a vigorous exercise protocol on the racetrack with overground endoscopy.
Unilateral compression of the base of the epiglottis was seen in 57.1% of STBs and in 9.1% of NSCTs, whilst bilateral compression was seen in 42.9% of STBs and in 90.9% of NSCTs. This uneven breed distribution may be related to the low sample size, or to a potential anatomical conformational difference of the hyoid apparatus and/or thyrohyoid articulation between these two genetically distinct breeds. Another plausible cause of unilateral compression of the base of the epiglottis is uneven head carriage while being driven, but this variable was not investigated in this study. However, the experienced drivers of the horses in this study always attempted to drive them evenly onto the bit during the examination. Additionally, any horses showing signs of lameness during the examination, a common cause of uneven head carriage in harness racehorses, were excluded from the study. The position of the endoscope should also not affect the unilateral appearance of the disorder as the endoscope was passed via the right ventral meatus in all cases, and there was an even distribution between left and right unilateral epiglottic compression. Finally, the compression of the epiglottis occurred only during phases of poll flexion and was never evident in phase 1 (free head carriage) prior to first induction of poll flexion.