Since the first report of laryngoplasty by Marks et al1 in 1970, studies2–5 have been performed on suture material, placement, and the biomechanical strength of the constructs. Laryngoplasty with ipsilateral ventriculocordectomy remains the surgery of choice for RLN in horses, yet important obstacles to success remain. The incidence of coughing, postoperative loss of abduction, and the precision required to place the suture for optimal abduction still present challenges.6,7 A 2013 study6 shows that of 41 horses with long-term follow-up after laryngoplasty, 3 (7%) had chondritis and 32 (78%) had abnormalities noted during dynamic endoscopy, most notably continued soft tissue collapse. Although many improvements have been made, laryngoplasty continues to be a difficult procedure with frequent suboptimal outcomes.
The partial and subtotal arytenoidectomy procedures were proposed as treatments of last resort when laryngoplasty fails.8,9 Removal of cartilage has less risk of infection, compared with the placement of a prosthesis, but may cause irreversible destabilization of the airway. Additionally, reasonable success with removal of cartilage can be achieved with less precision than placement of a prosthesis. Subtotal arytenoidectomy, a less extensive procedure, did not improve the airway impedance of exercising horses with RLN.10 Partial arytenoidectomy allowed 14 out of 18 (78%) racehorses to return to performance in 1 study8; however, coughing, nasal discharge, and dysphagia were observed in some cases. Partial arytenoidectomy has also been shown to have comparable rates of return to performance as laryngoplasty in racehorses.11
The arytenoid cartilage consists of both hyaline and elastic cartilage, with the elastic portion comprising most of the corniculate process.12 This process collapses into the airway during dynamic endoscopy in horses with RLN. Removal of this portion of the arytenoid cartilage, here termed COR, is expected to improve airway impedance without destabilizing the airway as occurs with PA. Further, COR does not require prosthesis placement, obviates the need for precise abduction and possible subsequent failure, and reduces the amount of soft tissue available to collapse into the airway.6
The primary objectives of the study reported here were to compare laryngeal impedance, in terms of air flow and pressure, following COR versus 3 other airway interventions (LLP, LLPCOR, and PA) performed on cadaveric equine larynges with simulated left RLN and to determine whether COR performed to remove the corniculate process of the left arytenoid cartilage (left COR) would provide similar improvement in laryngeal impedance as would LLP. The secondary objective was to determine whether relative laryngeal collapse correlated with the intervention procedures. The primary hypothesis of this study was that the improvement in laryngeal impedance after left COR in this model of left RLN in horses would be equivalent to the improvements following either LLP or LLPCOR, but greater than the improvement with PA and that the relative collapse of the cross-sectional area of the rima glottis measured from digital videos during simulated airflow would not differ significantly between the COR and LLP intervention states.
Materials and Methods
Samples
Larynges were collected from cadavers of horses euthanized for reasons unrelated to upper respiratory illness. The intrinsic laryngeal musculature was left intact on each larynx, and the larynges were dissected from the thyrohyoid bones to the fifth or sixth tracheal ring caudal to the cricoid cartilage. The isolated larynges were preserved in gauze saturated in saline (0.9% NaCl) solution to prevent desiccation during frozen storage. Larynges were frozen within 2 hours after euthanasia of donor horses, stored at −20°C, then thawed at 23°C for 20 hours prior to use. After thawing, the remaining extrinsic musculature, esophagus, and tracheal rings beyond the third ring were trimmed. The cricoarytenoideus dorsalis muscles were left intact bilaterally to simulate intact, diseased muscle.
Box model construct
A box with outer dimensions of 21.5 × 21.5 × 47.0 cm was built of 0.64-cm-thick clear acrylic sheets and had 2 circular portals (an inlet and an outlet) of 5.08 cm in diameter as previously described13–15 (Figure 1). For testing each larynx, the specimen was mounted in the box with either a 2.5-cm or 3.81-cm-diameter polyvinyl chloride adapter for best fit inserted into the tracheal lumen and secured with releasable nylon cable ties. The other end of the adapter connected to the outlet portal of the box. The epiglottis was maintained in extension with a nail fixed in the foam base support in the box. The air pressure within the lumen of the trachea (tracheal air pressure) and that in the box (pharyngeal air pressure) were each measured with a separate differential pressure transducera,b connected to polyurethane tubing placed behind the third tracheal ring in the adapter and another length of tubing placed in the box. The measurements were taken relative to the air pressure in the room, then the difference between the tracheal and pharyngeal air pressures (each relative to the pressure in the room) was used to calculate the change in pressure across the larynx. Upstream of the inlet portal, an airflow regulation valve was used to adjust airflow into the box. Downstream of the outlet portal, a butterfly valve within the pipe was attached to a direct current motor with manually adjustable speed that was used to cycle the air flow at 2 Hz. An orifice plate conforming to ISO standard 516716 was used to measure airflow 71 cm upstream from the vacuum and approximately 115 cm downstream from the rima glottis. The air pressure difference across the orifice plate was measured with a differential pressure transducer,a and from this pressure difference, the airflow rate was calculated per ISO 5167–1.16 The minimum and maximum airflow rates were tested in steady-state conditions to confirm accurate measurement during airflow oscillation. The airflow through the construct was adjusted to target a maximum negative pressure in the tracheal lumen of −4.3 kPa and a maximum airflow rate < 70 L/s for each cycle, consistent with parameters in horses at maximal exercise.17 Once the airflow was adjusted to maintain these parameters for each trachea in each intervention state, data were collected for ≥ 20 seconds.
Image of the box construct used in the present study to evaluate 28 cadaveric equine larynges for laryngeal impedance following simulation of left RLN alone and then with subsequent airway interventions (LLP, LLPCOR, COR, and PA). Airflow (blue arrow), measured by the orifice plate, started at the airflow regulation valve and traveled through the inlet portal, into the test box, through the cadaveric larynx and seated adapter (not shown), through the outlet portal, and through the pipe connected to the orifice plate, oscillating valve, and vacuum. The pharyngeal air pressure (obtained at the site of the larger red star) and tracheal air pressure (obtained at the site of the smaller red star) relative to air pressure in the room were obtained, and the difference (Δ Pressure) was calculated. These values were recorded during testing cycles for each larynx, evaluated with a targeted maximum negative tracheal pressure of −4.3 kPa and a maximum airflow rate of < 70 L/s.
Citation: American Journal of Veterinary Research 80, 12; 10.2460/ajvr.80.12.1136
Tracheal air pressure and pharyngeal air pressure and flow were recorded and monitored for each test (simulated RLN alone and with each intervention state [LLP, LLPCOR, COR, and PA]) performed on each larynx. The flow oscillation frequency was monitored by performing a Fourier transform on the flow measurement in real time, and the rotational speed of the butterfly valve was adjusted to ensure the flow oscillation frequency was maintained at 2 Hz. Data acquisition was performed at 100 Hz with a USB analog-to-digital converter.c These were simultaneously monitored and recorded with an individualized algorithm created in commercial software.d The laryngeal impedance was calculated from the difference in pressure (pharyngeal air pressure minus tracheal air pressure) divided by the airflow measured for the given duration.
Each test was recorded with a digital camera,e and video analysis of the resulting digital images was conducted by one of the authors (MLT). The point of maximal collapse was determined by scrolling through the frames for 3 cycles and determining when relaxation began, then selecting the frame prior to initiation of relaxation. This image was saved and imported into the imaging softwaref (Figure 2). The length of the right corniculate process was measured manually on the physical specimen and the digital image to create the scale for each image. Similarly, measurements between the right and left thyrohyoid joints on the physical specimen and the digital image were obtained to verify the scale of each image. The scale was used to measure the open area of the rima glottis by tracing along the internal border of the arytenoid cartilages, the laryngeal saccules, and the epiglottis in digital images. This was the area through which active airflow was expected to occur. For each larynx, 3 measurements of area were taken, and the mean was calculated for each tested simulated state. In addition, still images were taken when the larynx was not subjected to a vacuum so that the LRQ could be calculated in the LLP state; this still image was also used to measure the open area of the rima glottis, and this measurement was used as the baseline area. Relative laryngeal collapse was defined as the mean luminal area of each tested state divided by the baseline open rima glottis area from the still images (reported as a percentage). Each larynx was in turn secured in the box for testing first in the simulated state of left RLN alone, then in intervention states in the order of LLP, LLPCOR, COR, then PA before moving on to the next specimen.
Representative images of a cadaveric equine larynx with simulated left RLN evaluated in the box construct described in Figure 1 in simulated states of LLP at rest (A), maximally collapsed RLN alone (B), and RLN with either LLP (C), LLPCOR (D), COR (E), or PA (F). The basic lines and angles of measurements used to determine LQR overlay the image of the LLP intervention state at rest (A).
Citation: American Journal of Veterinary Research 80, 12; 10.2460/ajvr.80.12.1136
Simulation of left RLN
To imitate right arytenoid abduction and left RLN in each isolated larynx, a standard right-sided laryngoplasty was performed and the left arytenoid cartilage was unaltered. For the laryngoplasty, a single suture of size 5 polyester suture materialg was placed approximately 10 mm from the caudal edge of the cricoid cartilage medial to the most prominent palpable notch, when present, and 8 mm from the caudal edge of the muscular process of the arytenoid cartilage, engaging the spinous process. Maximal abduction of the right arytenoid cartilage was achieved by pulling the suture as tightly as possible before securing it with a knot.
Simulation of airway interventions
Once left RLN had been simulated in an isolated larynx and tested in the construct, an LLP was performed on that same larynx. A continuous suture pattern of 4–0 poliglecaprone 25 suture materialh was used to suture the mucosal aperture, and cyanoacrylate glue was placed along the rostral border of the saccule to seal it. The laryngotomy was left open to replicate the clinical situation immediately after surgery.
To use the LRQ method to compare the left and right sides of the larynx with LLP, a picture of each larynx was taken with a camera mounted above the testing box, 32 cm from the larynx opening and angled downward at 67.5° as previously described.18–20 The picture was subsequently imported into imaging softwaref with which the distance from the junction of the ventral aspect of the laryngeal saccules to the junction of the tips of the corniculate processes of the arytenoid cartilages was measured and one-third this length was then added to the dorsal aspect of this line (Figure 2). The angles between this line and a line drawn tangential to the dorsal edge of each corniculate process was measured, and the LRQ calculated. An LRQ of 0.88 was attempted; however, an LRQ between 0.85 and 0.95 was considered acceptable.
After testing was completed for the larynx with simulated left RLN and LLP intervention, an LLPCOR was performed by incising the arytenoid cartilage on the left side at the junction of the elastic and hyaline cartilage of the corniculate process in the transverse plane. The mucosa was fixed to the remaining cartilage with cyanoacrylate glue, and the larynx was tested again.
Next, the suture in the left side of the larynx was cut and removed, thus resulting in a replication of the laryngeal anatomic configuration following a simple left COR. The larynx was then tested in this configuration. Finally, a PA was performed by removing the remaining portion of the left arytenoid cartilage, leaving the muscular process as previously described,9 and the larynx was tested.
Statistical analysis
Twenty peaks of airflow and each pressure were included in statistical analysis for each test, but only if the corresponding peak tracheal pressure along the curve fell within the range of −4.0 to −4.6 kPa. When oscillation of the tissue resulted in an inability of the orifice plate to give a meaningful flow value, commercially available softwarei was used for interpolation with a curve of best fit derived from the method of least squares to determine the data at the peak of the curve.21 The apex of each peak was determined, and the mean peak value was calculated for tracheal pressure, pharyngeal pressure, and airflow for each tested simulated state in each larynx. A multivariable mixed-effects linear regression model was used to compare results for each of the simulated states (RLN alone and with each of LLP, LLPCOR, COR, and PA), controlling for multiple tests on each larynx. Measurements for laryngeal impedance were logarithmically transformed to normalize the data during analyses. Horse breed and age, adapter size (2.54 cm vs 3.81 cm), LRQ, surgical procedure (LLP, LLPCOR, COR, and PA), whether the set required interpolation, airflow adjustment, use of nonconsecutive peak air pressures, and percentage of visibly open area of the rima glottis were included in the model, with nonsignificant factors removed. The mixed-effects model accounting for repeated measures for each larynx was used to determine the predicted mean impedance of each simulated state, correcting for individual larynx and relative laryngeal collapse. Statistical analyses were performed with commercial software,j and values of P < 0.05 were considered significant.
Results
Twenty-eight larynges were collected from cadavers of horses euthanized for reasons unrelated to upper respiratory illness. During testing of each larynx in each simulated state, the air pressures, airflow, and airflow oscillation frequency were continuously monitored to allow adjustment of peak air pressure, flow, and butterfly valve frequency. Airflow ranged from 10.4 to 85 L/s, and peak negative tracheal air pressure ranged from −4.07 to −4.54 kPa (Table 1). Air pressure in the construct was adjusted when testing each larynx until the peak negative tracheal air pressure was approximately −4.30 kPa. Nonconsecutive peaks were occasionally used. Two tests (the LLP and LLPCOR states on larynx 9) resulted in insufficient peaks falling within the acceptable range
Results for laryngeal airflow and pressures and relative laryngeal collapse in 28 cadaveric equine larynges with simulated states of left RLN alone and with subsequent airway interventions (LLP, LLPCOR, COR, and PA).
| Simulated states | Airflow (L/s) | TP (kPa) | PP (kPa) | Relative laryngeal collapse (%) |
|---|---|---|---|---|
| RLN | 25.0 (10.4 to 52.8) | −4.28 (−4.49 to −4.13) | −0.157 (−0.47 to −0.05) | 0.27 (0.12 to 0.50) |
| LLP | 53.3 (12.3 to 85.0) | −4.22 (−4.45 to −4.08) | −0.853 (−2.97 to −0.04) | 0.71 (0.21 to 0.94) |
| LLPCOR | 50.3 (20.7 to 80.8) | −4.24 (−4.46 to −4.07) | −0.782 (−2.96 to −0.10) | 0.67 (0.27 to 1.03) |
| COR | 26.8 (10.9 to 48.5) | −4.30 (−4.54 to −4.11) | −0.252 (−2.20 to −0.07) | 0.32 (0.12 to 0.54) |
| PA | 32.5 (14.2 to 53.3) | −4.27 (−4.47 to −4.12) | −0.270 (−2.36 to −0.16) | 0.51 (0.20 to 0.86) |
Data are reported as the mean (range).
The mixed effects model accounting for repeated measures for each larynx was used to determine the predicted mean impedance of each simulated state, correcting for individual larynx and relative laryngeal collapse.
PP = Pharyngeal air pressure. TP = Tracheal air pressure.
The mixed effects model accounting for repeated measures for each larynx was used to determine the predicted mean impedance of each simulated state, correcting for individual larynx and relative laryngeal collapse. PP = Pharyngeal air pressure. TP = Tracheal air pressure of −4.0 to −4.5 kPa, and the missing 20 peaks for the tests were determined by interpolation with a curve of best fit derived from the method of least squares to determine the data at the peak of the curve.21
When corrected for individual larynx and relative laryngeal collapse, the predicted mean laryngeal impedance was significantly (P < 0.01) greater for larynges fit with the 2.54-cm adapter versus those fit with the 3.81-cm adapter when tested in the simulated states of RLN (0.170 kPas/L [95% CI, 0.142 to 0.203 kPas/L] and 0.122 kPas/L [95% CI, 0.092 to 0.161 kPas/L], respectively), LLP (0.094 kPas/L [95% CI, 0.078 to 0.112 kPas/L] and 0.046 kPas/L [95% CI, 0.035 to 0.059 kPas/L], respectively), and LLPCOR (0.094 kPas/L [95% CI, 0.079 to 0.112 kPas/L] and 0.052 kPas/L [95% CI, 0.040 to 0.068 kPas/L], respectively). However, the predicted mean laryngeal impedance did not differ significantly (P = 0.875 and P = 0.115, respectively) between larynges fit with the 2.54-cm adapter versus those fit with the 3.81-cm adapter when tested in the simulated states of COR (0.151 kPas/L [95% CI, 0.128 to 0.179 kPas/L] and 0.148 kPas/L [95% CI, 0.113 to 0.192 kPas/L], respectively) or PA (0.139 kPas/L [95% CI, 0.119 to 0.164 kPas/L] and 0.109 kPas/L [95% CI, 0.084 to 0.141 kPas/L], respectively; Figure 3; Table 2). In addition, the predicted mean impedance did not substantially differ among the RLN alone, COR, and PA simulated states for either adapter size, but was significantly (P < 0.01) higher for those simulated states than for the simulated states of LLP and LLPCOR, regardless of adapter size. The predicted mean impedance did not differ substantially between the simulated states of LLP and LLPCOR for either adapter size. Residual intraclass correlation of the model was 27.6%, which suggested that individual laryngeal variation affected airway impedance because this remained once all the other factors had been accounted for.
Histogram of the predicted mean laryngeal impedance stratified by adapter size (2.54 cm in diameter [gray bars] vs 3.81 cm in diameter [white bars]) for the 28 cadaveric equine larynges tested in simulated states of left RLN alone and with subsequent airway interventions (LLP, LLPCOR, COR, and PA) when controlling for individual larynx and relative laryngeal collapse. The bars represent the predicted mean laryngeal impedance, and the whiskers represent the 95% CI of the predicted mean. *Predicted mean laryngeal impedance differed significantly (P < 0.01) between larynges fit with the 2.54-cm versus 3.81-cm adapter in the procedure group. a,bPredicted mean laryngeal impedance for procedures with different superscripts differed significantly (P < 0.05) for larynges tested with the same size adapter.
Citation: American Journal of Veterinary Research 80, 12; 10.2460/ajvr.80.12.1136
Summary of the P values determined with the mixed-effects model controlling for repeated measures for each larynx to assess potential differences in results for predicted mean laryngeal impedance stratified by simulated state (RLN state alone and with each of LLP, LLPCOR, COR, and PA) and adapter size (2.54 cm vs 3.81 cm in diameter) for the 28 cadaveric equine larynges tested. Values of P < 0.05 were considered significant.
| Adapter size | RLN | LLP | LLPCOR | COR |
|---|---|---|---|---|
| 2.54 cm in diameter | ||||
| LLP | < 0.01 | |||
| LLPCOR | < 0.01 | 0.934 | ||
| COR | 0.24 | < 0.01 | < 0.01 | |
| PA | 0.93 | < 0.01 | < 0.01 | 0.44 |
| 3.81 cm in diameter | ||||
| LLP | < 0.01 | |||
| LLPCOR | < 0.01 | 0.378 | ||
| COR | 0.213 | < 0.01 | < 0.01 | |
| PA | 0.482 | < 0.01 | < 0.01 | 0.05 |
The mean LRQ was 0.88 (range, 0.81 to 0.94), and the overall mean video-derived measurement of the open area of the rima glottis was 6.43 cm2 (range, 1.06 cm2 [a larynx with RLN simulated state alone] to 16.84 cm2 [a larynx with RLN and LLP simulated states]). The mean relative collapse of the rima glottis area differed significantly among airway intervention states (P < 0.001) and after controlling for adapter size (P = 0.004). Specifically, the mean relative collapse of the rima glottis area was 35.0% (95% CI, 31.7% to 38.3%) for the COR, 51.2% (95% CI, 45.7% to 56.7%) for the PA, 66.6% (95% CI, 61.0% to 72.2%) for the LLPCOR, and 71.4% (95% CI, 65.1% to 77.7%) for the LLP.
Discussion
The box model used in the present study has been validated to yield comparable values for pharyngeal and tracheal air pressures and airflow as measured in exercising horses.13 Results of the present study indicated similar laryngeal impedance values for the simulated RLN alone and with LLP intervention as previously reported by other ex vivo models,13,14 which further supported the replicability of this model. Hawkins et al15 reported different values for impedance, compared with other ex vivo studies,13,14 but included surgeries and postoperative healing in live horses before the animals were euthanized and the larynges tested. The validity of the box model construct used in the present study was further supported with our results for the simulated RLN alone and with LLP intervention, which were similar to pressure and flow measurements obtained from different portions of the upper airway during exercise in horses that underwent airway surgeries.10,22–25
The box model yielded a wide range of results for airflows for the simulated states tested; however, this was unavoidable given the variation in the open area of the rima glottis across the different states. The adapters that were used to accommodate different laryngeal diameters in the present study were not used in previous studies but best replicated the tracheal diameter with the associated larynx. Incising the tracheal rings to force fit a 5-cm-diameter pipe end is one potential solution to avoid airflow losses associated with the change in diameter from the larynx to the downstream components of the testing system. However, doing so may stretch the soft tissues and interfere with the abduction of the arytenoid cartilages. The placement of the adapters in the present study was observed to avoid interference with the soft tissues of each larynx but allowed the tracheal mucosa to form a seal with the pipe connected to the vacuum. Thus, our model resulted in a controlled change in diameter, replication of the patient geometry, and inclusion of the adapter as a factor within the statistical model.
Laryngoplasty improves laryngeal impedance to the pre-RLN state; however, studies22,23 of PA have yielded conflicting results. Results of the present study indicated that PA did not improve airway impedance as well as LLP, and that results for COR were not meaningfully different from those of PA. Additionally, the predicted mean impedance for the simulated RLN state alone was consistently an order of magnitude higher than that of the LLP intervention state, which was consistent with findings in another study.14 Further, the impedances for the PA and COR intervention states were not meaningfully different from the RLN state alone, which indicated that laryngeal collapse in this model was dissimilar to that observed in live horses. This finding was inconsistent with findings in an in vivo study23 that shows horses with induced RLN treated with a modified PA returned to pre-RLN impedance values. Additionally, Lumsden et al22 also demonstrated improved impedance with PA; however, impedance was measured across the entire upper airway.
The subtotal arytenoidectomy has also been evaluated in live horses; however, subtotal arytenoidectomy did not improve the laryngeal negative air pressure beyond that of the RLN state.10 In the present study, air pressure was controlled while the airflow was altered, and substantial drops in flow were observed for the RLN simulated state alone and for the COR and PA intervention states. In contrast, the subtotal arytenoidectomy has been largely abandoned and involves removal of the arytenoid cartilage excepting the corniculate and muscular processes. This outcome is not unexpected given that the corniculate process collapses into the airway during peak exercise in horses with RLN. In another study26 of PA, postoperative morbidity was reduced by reevaluation and removal of the remaining collapsing portions of the aryepiglottic fold. This finding suggests that targeted removal of the collapsing portions of the airway would be beneficial to improve airway mechanics, as with the COR procedure. The conformation of the laryngeal entrance to the airway could be adjusted during a standing endoscopic COR, and, given the differences between results observed in live horses versus the box model for these arytenoid procedures, further investigation into the usefulness of COR is warranted.
A study by Jansson et al14 measured cross-sectional area of the rima glottis in horses to compare cordopexy, cordopexy and laryngoplasty, and laryngoplasty. The areas ranged from 3.80 to 10.10 cm2; however, the measurements were calculated on the basis of airflow, translaryngeal pressure difference, and density, whereas the values in the present study were measured from still digital images. In the present study, some differences between larynges (eg, larynx length and angle) could not be controlled; therefore, the internal scale of the right arytenoid of each larynx was used. The percentage of reduction in area was calculated to avoid error resulting from laryngeal size variation. The force of the vacuum was observed to pull the larynx caudally along its long axis, despite the epiglottic anchor. Nonetheless, the relative laryngeal collapse and impedance for the COR intervention state were markedly greater than those in the LLP intervention state in the present study. Measuring cross-sectional areas within larynges is intricate, and, as has been previously demonstrated, flows within this region are complex and interaction between patient geometry and flow development warrants further investigation.27
Our results indicated that approximately 28% of the variation within our statistical model could have been attributed to the difference between individual larynges. This indicated that factors intrinsic to individual larynges and not captured in our model affected airway impedance. Similarly, other studies18,28,29 have also alluded to this unquantified aspect. For instance, Witte et al18 reported that some larynges required an unusually great force to achieve a similar amount of abduction, and other authors have commented that the stability of the cartilage is a contributing factor to the performance of the horse following laryngoplasty.28,29 In the present study, results for the LLPCOR intervention state indicated improved laryngeal impedance and open area of the rima glottis in some larynges and worsened impedance in others, compared with results for the respective larynges with the LLP intervention state. This finding suggested that for some larynges, corniculate process collapse may have resulted in suboptimal LLP performance and less improvement of impedance, compared with the simulated RLN state alone. Historically, failure of laryngoplasty has been blamed, in part, on the lack of cartilaginous stiffness of the arytenoid cartilages. Indeed, arytenoid corniculate process flexibility was proposed as one of the reasons for failure of improvement of a horse's performance in the face of adequate arytenoid abduction observed during resting examination.6,28 Additionally, induction of dorsal displacement of the soft palate, concurrent soft tissue collapse, and arytenoid chondritis have also been observed postoperatively in horses that underwent laryngoplasty, indicating that further intervention may be required in particular cases.6,28
Results of the present study indicated that neither COR nor PA improved impedance as well as LLP. These findings regarding PA conflicted with its reported clinical success and by extension also suggested that COR may have different clinical success in exercising horses than observed in the present study. The variation in results attributed to individual larynx properties in the present study were consistent with findings in other studies18,28,29 in which individual horses had various soft tissue conformations and collapse after surgery.
Acknowledgments
Supported in part by the Townsend Equine Health Research Fund.
Equipment was provided by the Department of Mechanical Engineering, University of Saskatchewan.
Funding and equipment sources did not have any involvement in the study design, data analysis and interpretation, or writing of the manuscript.
The authors declare that there were no conflicts of interest.
ABBREVIATIONS
| CI | Confidence interval |
| COR | Corniculectomy |
| LLP | Left-sided laryngoplasty with ipsilateral ventriculocordectomy |
| LLPCOR | Left-sided laryngoplasty with ipsilateral ventriculocordectomy combined with corniculectomy |
| LRQ | Left-to-right angle quotient |
| PA | Partial arytenoidectomy |
| RLN | Recurrent laryngeal neuropathy |
Footnotes
Model P55D, Validyne Engineering, Northridge, Calif.
Model DP103–14, Validyne Engineering, Northridge, Calif.
USB-6221 National Instruments, Austin, Tex.
LabView, National Instruments, Austin, Tex.
Logitech C920 Pro HD Webcam, Logitech, Newark, Calif.
ImageJ, National Institutes of Health, Bethesda, Md.
Ethibond Excel, Ethicon US LLC, Bridgewater, NJ.
Monocryl, Ethicon US LLC, Bridgewater, NJ.
Microsoft Excel, version 1810, Microsoft Corp, Redmond, Wash.
Stata, version 14.0, StataCorp, College Station, Tex.
References
1. Marks D, Mackay-Smith MP, Cushing LS, et al. Use of a prosthetic device for surgical correction of laryngeal hemiplegia in horses. J Am Vet Med Assoc 1970;157:157–163.
2. Ahern BJ, Van Eps AW, Boston RC, et al. In vitro comparison of 3 techniques of prosthesis attachment to the muscular process of the equine arytenoid cartilage. Vet Surg 2017;46:700–704.
3. Brandenberger O, Rossingnol F, Perkins JD, et al. Ex vivo biomechanical stability of 5 cricoid-suture constructs for equine laryngoplasty. Vet Surg 2017;46:705–713.
4. Kelly JR, Carmalt J, Hendrick S, et al. Biomechanical comparison of six suture configurations using a large diameter polyester prosthesis in the muscular process of the equine arytenoid cartilage. Vet Surg 2008;37:580–587.
5. Willsallen H, Heller J, Kark L, et al. In vitro mechanical testing of braided polyurethane elastic fiber and braided polyester for equine laryngoplasty. Vet Surg 2015;44:223–230.
6. Barnett TP, O'Leary JM, Parkin TD, et al. Long-term exercising video-endoscopic examination of the upper airway following laryngoplasty surgery: a prospective cross-sectional study of 41 horses. Equine Vet J 2013;45:593–597.
7. Barnett TP, O'Leary JM, Parkin TDH, et al. Long-term maintenance of arytenoid cartilage abduction and stability during exercise after laryngoplasty in 33 horses. Vet Surg 2013;42:291–295.
8. Speirs VC. Partial arytenoidectomy in horses. Vet Surg 1986;15:316–320.
9. White NA, Blackwell RB. Partial arytenoidectomy in the horse. Vet Surg 1980;9:5–12.
10. Belknap JK, Derksen FJ, Nickels FA, et al. Failure of subtotal arytenoidectomy to improve upper airway flow mechanics in exercising Standardbreds with induced laryngeal hemiplegia. Am J Vet Res 1990;51:1481–1487.
11. Witte TH, Mohammed HO, Radcliffe CH, et al. Racing performance after combined prosthetic laryngoplasty and ipsilateral ventriculocordectomy or partial arytenoidectomy: 135 Thoroughbred racehorses competing less than 2400 m (1997–2007). Equine Vet J 2009;41:70–75.
12. Fulton IC, Anderson BA, Stick JA, et al. Larynx. In: Auer JA, Stick JA, eds. Equine surgery. 4th ed. St Louis: Elsevier Saunders, 2012;592–623.
13. Cheetham J, Witte TH, Soderholm LV, et al. In vitro model for testing novel implants for equine laryngoplasty. Vet Surg 2008;37:588–593.
14. Jansson N, Ducharme NG, Hackett RP, et al. An in vitro comparison of cordopexy, cordopexy and laryngoplasty, and laryngoplasty for treatment of equine laryngeal hemiplegia. Vet Surg 2000;29:326–334.
15. Hawkins JF, Couetil L, Miller MA. Maintenance of arytenoid abduction following carbon dioxide laser debridement of the articular cartilage and joint capsule of the cricoarytenoid joint combined with the prosthetic laryngoplasty in horses: an in vivo and in vitro study. Vet J 2014;199:275–280.
16. International Organization for Standardization. Measurement of fluid flow by means of pressure differential devices inserted in circular cross-section conduits running full—Part 1: general principles and requirements (ISO 5167). 2003. Available at: www.iso.org/standard/28064.html
17. Nielan GJ, Rehder RS, Ducharme NG, et al. Measurement of tracheal static pressure in exercising horses. Vet Surg 1992;21:423–428.
18. Witte TH, Cheetham J, Solderham LV, et al. Equine laryngoplasty sutures undergo increased loading during coughing and swallowing. Vet Surg 2010;39:949–956.
19. Perkins JD, Meighan H, Windley Z, et al. In vitro effect of ventriculocordectomy before laryngoplasty on abduction of the equine arytenoid cartilage. Vet Surg 2011;40:305–310.
20. Ahern BJ, Lim YW, van Eps A, et al. In vitro evaluation of the effect of a prototype dynamic laryngoplasty system on arytenoid abduction. Vet Surg 2018;47:837–842.
21. Johnson RA. Curve fitting. In: Johnson RA, ed. Miller & Freund's probability and statistics for engineers. 7th ed. Upper Saddle River, NJ: Pearson Prentice Hall Inc, 2005;337–357.
22. Lumsden JM, Derksen FJ, Stick JA, et al. Evaluation of partial arytenoidectomy as a treatment for equine laryngeal hemiplegia. Equine Vet J 1994;26:125–129.
23. Radcliffe CH, Woodie JB, Hackett RP, et al. A comparison of laryngoplasty and modified partial arytenoidectomy as treatments for laryngeal hemiplegia in exercising horses. Vet Surg 2006;35:643–652.
24. Williams JW, Pascoe JR, Meagher DM, et al. Effects of left recurrent laryngeal neurectomy, prosthetic laryngoplasty, and subtotal arytenoidectomy on upper airway pressure during maximal exertion. Vet Surg 1990;19:136–141.
25. Williams JW, Meagher DM, Pascoe JR, et al. Upper airway function during maximal exercise in horses with obstructive airway lesions: Effect of surgical treatment. Vet Surg 1990;19:142–147.
26. Davidson EJ, Parente EJ. Exercising videoendoscopic evaluation of 7 horses with abnormal respiratory noise and poor performance following partial arytenoidectomy. Equine Vet Educ 2011;23:626–629.
27. Rakesh V, Ducharme NG, Datta AK, et al. Development of equine upper airway fluid mechanics model for Thoroughbred racehorses. Equine Vet J 2008;40:272–279.
28. Davidson EJ, Martin BB, Rieger RH, et al. Exercising videoendoscopic evaluation of 45 horses with respiratory noise and/or poor performance after laryngoplasty. Vet Surg 2010;39:942–948.
29. Compostella F, Tremaine WH, Franklin SH. Retrospective study investigating causes of abnormal respiratory noise in horses following prosthetic laryngoplasty. Equine Vet J Suppl 2012;44:27–30.
