In horses, RLN is a common upper airway disorder resulting in decreased athletic performance and altered upper airway dynamics.1–3 Disturbance of upper airway dynamics is precipitated by the characteristic progressive degenerative axonopathy of the recurrent laryngeal nerve, leading to atrophy of the cri-coarytenoideus dorsalis muscle and subsequent collapse of the arytenoid cartilage and associated vocal fold.4,5 Loss of arytenoid abduction in affected horses can lead to airway noise and exercise intolerance because of decreased airway diameter at the level of the rima glottis and functional obstruction during respiration.6,7 Targeted surgical treatments of RLN include laryngeal ventriculectomy or ventriculocordectomy, prosthetic laryngoplasty, laryngeal reinnervation, and arytenoidectomy.1,2,8 Laryngoplasty, frequently combined with ventriculectomy or ventriculocordectomy, is the most common surgical technique selected to ameliorate clinical signs in horses with RLN.2,6,7
Surgical treatment of RLN has been evaluated in Thoroughbred and Standardbred racehorses1,2; however, little information is available regarding performance after prosthetic laryngoplasty in racing Quarter Horses with RLN. The purposes of the study reported here were to determine effects of prosthetic laryngoplasty on return to racing, PI, and career longevity in racing Quarter Horses with RLN and to evaluate performance variables for racing Quarter Horses with RLN undergoing prosthetic laryngoplasty (case horses), compared with racing Quarter Horses without RLN (control horses).
Materials and Methods
Racing Quarter Horses that underwent prosthetic laryngoplasty for treatment of RLN between January 2000 and December 2015 were identified from the medical records of 5 referral centers, including the Colorado State University Veterinary Teaching Hospital, Elgin Veterinary Hospital, Equine Medical Center, Oakridge Equine Hospital, and Equine Sports Medicine and Surgery. Horses were excluded as case horses if they were not racehorses, had incomplete medical records, were > 8 years of age at the time of surgery, or failed to have ≥ 1 racing start in the preoperative or postoperative period. Information collected from medical records included age at the time of surgery (surgery reference point); sex; grade of RLN; whether ventriculectomy, ventriculocordectomy, or concurrent arthroscopy was performed; and postoperative complications identified and treated by the attending veterinarian. Ventriculectomy or ventriculocordectomy was performed as previously described, either with laryngotomy and sharp transection2 or with transendoscopic diode laser excision,9 at the discretion of the attending surgeon. Laryngeal appearance was evaluated during endoscopic examination with the horse standing without chemical restraint and was assigned a laryngeal function grade10 on a scale from grade I (synchronous and symmetrical full arytenoid abduction achieved and maintained) to grade IV (complete immobility of the arytenoid cartilage and vocal fold) by the attending clinician or surgeon. Surgical intervention was planned on the basis of results of the resting laryngeal function examination and whether the horse had abnormal respiratory noise and signs of exercise intolerance. Prosthetic laryngoplasty was performed as previously described11 with 2 separate sutures of size-5 braided polyester.
For each case horse included in the study, 2 racing Quarter Horse control horses were identified to allow comparison of performance variables with peer athletic horses. Control horses were of the same age and sex that placed closest in finish time in the race that each case horse ran immediately before or after the surgery reference point. The surgery date for each case horse was used as the surgery reference point, separating pre- and postoperative periods for each case horse and its 2 matched control horses. Racing starts, earnings, distance, and career longevity were obtained from an online database.a The PIs for pre- and postoperative periods were calculated as the sum of points (first place finish = 3 points, second place = 2 points, third place = 1 point, and fourth place or lower = 0 points) earned in the 3 races immediately before and after the surgery reference point. Career longevity for each horse was measured from the date of a horse's first race to the date of its last race or retirement.
Statistical analysis
Return to racing and performance variables were evaluated separately. Return to racing was evaluated in all horses, whereas performance variables were evaluated in horses that raced at least once after the surgery reference point. Associations between return to racing and RLN grade for case and control horses were evaluated with χ2 analysis. Influence of prosthetic laryngoplasty and RLN grade were further evaluated by logistic regression, controlling for age and sex. The odds of returning to racing were reported as OR (95% CI). Influence of ventriculectomy or ventriculocordectomy, concurrent arthroscopy, and development of postoperative complications on return to racing for case horses, compared with control horses, were evaluated with the Fisher exact test. Continuous variables were reported as median (IQR). Earnings, number of racing starts, and PI were evaluated for case and control horses for pre- and postoperative periods with a Wilcoxon signed rank sum test and bootstrap regression analyses. For bootstrap regression analyses, 1,000 resamples were drawn from the subject-level population, and age was included as a covariate. The influence of RLN grade on earnings, number of racing starts, and PI was evaluated for pre- and postoperative periods with a Kruskal-Wallis test, followed by pairwise evaluation of RLN grade of case horses, compared with control horses, with Dunn tests, if significant differences were identified. Analysis of postoperative performance variables only included case and control horses that raced after the surgery reference point. Racing starts, earnings, and PI during the pre- and postoperative periods were evaluated within groups with a Wilcoxon signed rank sum test, and only included horses that returned to racing. Effect of prosthetic laryngoplasty and RLN grade on career longevity was investigated with a log-rank test. All statistical tests were performed with available statistical software,b and significance was defined as P ≤ 0.05.
Results
Animals
The medical records review of the 5 referral centers identified 162 racing Quarter Horses with RLN treated with prosthetic laryngoplasty (case horses), which were matched with 324 racing Quarter Horse control horses. At the surgery reference point, the median age of all 486 horses was 3 years (IQR, 2 to 5 years), and the median race distance was 350 yards (IQR, 330 to 400 yards). Database reports of race distance selected by trainers for individual horses did not change throughout the study period.
Case horses—The 162 case horses consisted of 92 (57%) geldings, 38 (23%) mares, and 32 (20%) stallions. One hundred fifty-nine (98%) had left-sided RLN, and 3 (2%) had right-sided RLN. Sixty-eight of the 162 (42%) horses had grade IV RLN, 59 (36%) had grade III RLN, and 35 (22%) had grade II RLN. At the surgery reference point, the median age of case horses was 3 years (IQR, 2 to 5 years), and median race distance was 350 yards (IQR, 330 to 400 yards). One hundred thirty-six of the 162 (84%) case horses had raced before surgery.
Control horses—The 324 control horses consisted of 184 (57%) geldings, 76 (23%) mares, and 64 (20%) stallions. At the surgery reference point, the median age of the control horses was 3 years (IQR, 2 to 5 years), and median race distance was 350 yards (IQR, 330 to 400 yards).
Treatments and postoperative complications
Of the 162 case horses, 109 (67%) underwent prosthetic laryngoplasty combined with ventriculectomy or ventriculocordectomy for treatment of RLN, whereas 53 (33%) underwent prosthetic laryngoplasty alone. In addition, 29 of the 162 (18%) case horses underwent concurrent arthroscopy in ≥ 1 joints when treated for RLN. Postoperative complications occurred in 29 of the 162 (18%) case horses after surgical treatment of RLN and included seroma formation (n = 10), cough (5), pneumonia (3), colic (2), incisional drainage (2), nasal discharge (2), incisional dehiscence (2), postoperative fever (2), and gastric ulceration (1).
Return to racing
Of the 162 case horses, 137 (85%) returned to race at least once (52 of the 137 [38%] won ≥ 1 race) and 124 (77%) ran in ≥ 3 races after the surgery reference point. Case horses that returned to racing included 91 of the 109 (83%) case horses that underwent prosthetic laryngoplasty combined with ventriculectomy or ventriculocordectomy and 46 of the 53 (87%) case horses that underwent prosthetic laryngoplasty alone. Twenty-seven of the 29 (93%) case horses that underwent concurrent arthroscopy and 26 of the 29 (90%) case horses that had postoperative complications returned to racing. Of the 324 control horses, 270 (83%) returned to race at least once (67 of the 270 [25%] won ≥ 1 race) and 234 (72%) ran ≥ 3 races after the surgery reference point. All 26 case horses that had not raced before surgery raced after surgery. Median interval between the surgery reference point and return to racing was 137 days (IQR, 96 to 198 days) for case horses. There was no significant (Fisher exact test, P = 0.862) difference in the proportion that returned to racing between case horses and control horses, and the odds of returning to racing did not differ significantly (OR, 1.09; 95% CI, 0.66 to 1.86; χ2 test, P = 0.728) for case horses, compared with control horses. Further, the proportion that returned to racing did not differ significantly (Fisher exact test, P = 0.382 and P = 0.796, respectively) for case horses that underwent concurrent arthroscopy or that had postoperative complications, compared with control horses.
Thirty-one of 35 (89%) horses with grade II RLN returned to racing, as did 50 of 59 (85%) horses with grade III RLN and 56 of 68 (82%) horses with grade IV RLN. The proportion that returned to racing did not differ significantly (χ2 test, P = 0.854) among groups for the 3 RLN grade subgroups of case horses and the control horses. Also, the odds of case horses categorized by RLN grade returning to racing did not differ significantly (grade II [OR, 1.50; 95% CI, 0.56 to 5.32; P = 0.423], grade III [OR, 1.10; 95% CI, 0.53 to 2.52; P = 0.787], and grade IV [OR, 0.93; 95% CI, 0.48 to 1.93; P = 0.844]) from the odds for control horses.
When case horses were grouped by age at the surgery reference point, 47 of 49 (96%) horses ≤ 2 years of age, 42 of 46 (91%) horses 3 years of age, 40 of 49 (82%) horses between 4 and 5 years of age, and 6 of 18 (33%) horses between 6 and 8 years of age returned to racing. Similarly, when control horses were grouped by age at the surgery reference point, 95 of 98 (97%) horses ≤ 2 years of age, 83 of 92 (90%) horses 3 years of age, 80 of 98 (82%) horses between 4 and 5 years of age, and 12 of 36 (33%) horses between 6 and 8 years of age returned to racing. After controlling for age and sex with logistic regression analysis, no significant differences in the odds of returning to racing were detected for case horses overall, compared with control horses (OR, 1.25; 95% CI, 0.70 to 2.32; P = 0.456), or for case horses categorized by RLN grade (grade II [OR, 1.40; 95% CI, 0.46 to 5.41], grade III [OR, 1.55; 95% CI, 0.67 to 4.03], and grade IV [OR, 0.99; 95% CI, 0.46 to 2.31]; P = 0.748), compared with control horses. However, older age was significantly (OR, 0.44; 95% CI, 0.37 to 0.54, P < 0.001) associated with lower odds of returning to racing.
Racing performance
Performance variables of case and control horses that returned to racing were calculated as medians and IQRs and further reported by RLN grade. The median number of racing starts during the preoperative period was 6 (IQR, 2 to 13) and 7 (IQR, 2 to 14) for case and control horses, respectively, and during the postoperative period was 8 (IQR, 5 to 13) and 8 (IQR, 4 to 14) for case and control horses, respectively. The median number of racing starts during the preoperative period did not differ significantly (Wilcoxon signed rank sum test, P = 0.298) between case and control horses, nor was there a significant (Wilcoxon signed rank sum test, P = 0.942) difference between groups for racing starts during the postoperative period. Further, median number of racing starts during the postoperative period did not differ significantly (bootstrap regression analysis, P = 0.878) between case and control horses when adjusted for age at the surgery reference point. Similarly, number of racing starts during the preoperative period was not significantly (Wilcoxon signed rank sum test, P = 0.099 and P = 0.123, respectively) different from number of racing starts during the postoperative period for case horses or for control horses. When case horses were grouped according to RLN grade, number of racing starts did not differ significantly between case and control horses during either the pre- or postoperative period (Kruskal-Wallis test, P = 0.695 and P = 0.963, respectively).
Median earnings for case and control horses during the preoperative period were $13,228 (IQR, $520 to $44,501) and $9,885 (IQR, $534 to $28,681), respectively, and for the postoperative period were $11,123 (IQR, $3,370 to $24,285) and $7,504 (IQR, $1,950 to $22,446), respectively. Racing earnings did not differ significantly (Wilcoxon signed rank sum test, P = 0.317 and P = 0.118, respectively) between case and control horses during the pre- or postoperative period, nor was there a significant (bootstrap regression analysis, P = 0.310) difference between the groups for earnings during the postoperative period after adjusting for age at the surgery reference point. However, median earnings decreased by 16% after the surgery reference point for case horses and by 25% for control horses. There was a significant difference between pre- and postoperative earnings for case horses (Wilcoxon signed rank sum test, P = 0.046) but not for control horses (Wilcoxon signed rank sum test, P = 0.062). Further, when case horses were grouped by RLN grade, racing earnings did not differ significantly (Kruskal-Wallis test, P = 0.334 and P = 0.318, respectively) between case and control horses during the pre- or postoperative period.
The PI for case horses during the preoperative period (PI, 2; IQR, 0 to 3) did not differ significantly (Wilcoxon signed rank sum test, P = 0.287) from the PI for control horses during the preoperative period (PI, 1.5; IQR, 0 to 3); however, the PI for case horses during the postoperative period (PI, 2; IQR, 0 to 4) was significantly (Wilcoxon signed rank sum test, P = 0.002; bootstrap regression analysis, P = 0.004) higher than the PI for control horses during the postoperative period (PI, 1; IQR, 0 to 3). From the preoperative period to the postoperative period, the PI for case horses and control horses, did not change significantly (Wilcoxon signed rank sum test, P = 0.082 and P = 0.315, respectively). Further, the PI during the preoperative period did not differ significantly (Kruskal-Wallis test, P = 0.702) between case horses grouped by RLN grade and control horses; however, the PI during the postoperative period differed significantly (Kruskal-Wallis test, P = 0.004) among groups, in that horses with the lowest RLN grade (grade II) had the highest median PI (3; IQR, 1 to 4), whereas horses with grades III and IV RLN had median PIs of 1.5 (IQR, 0 to 4) and 2.5 (IQR, 1 to 4), respectively, and median PI for control horses was 1 (IQR, 0 to 3). Subsequent pairwise comparisons also identified significantly (P = 0.010 and P = 0.035, respectively) higher PIs for case horses with grade II or grade IV RLN, compared with control horses.
Median career longevity was 942 days (IQR, 588 to 1,565 days) for case horses and 885 days (IQR, 525 to 1,329 days) for control horses. Median career longevity for horses with grades II, III, and IV RLN was 893 days (IQR, 512 to 1,655 days), 950 days (IQR, 602 to 1,719 days), and 935 days (IQR, 598 to 1,408 days), respectively. There were no significant (log-rank test, P = 0.11 and P = 0.41, respectively) differences in career longevity between case and control horses overall or between control horses and case horses grouped by RLN grade.
Discussion
To our knowledge, the present study was the first retrospective cohort study of performance in racing Quarter Horses undergoing prosthetic laryngoplasty for treatment of RLN, and results indicated that case horses returned to racing and performed at the level of control horses. Number of racing starts, racing earnings, and career longevity did not differ between case and control horses; however, case horses had a higher postoperative PI. Furthermore, neither concurrent surgical procedures nor postoperative complications affected race performance of case horses. Findings also indicated that age at the time of prosthetic laryngoplasty and RLN grade did affect race performance of case horses because younger horses were more likely than older horses to return to racing and because horses with the lowest RLN grade had the highest PI following surgery. Older horses had lower odds of returning to racing and lower race performance after the surgery reference point in the present study, indicating that full restoration of racing performance in older case horses was less likely. Compared with the preoperative period, earnings decreased in the postoperative period for case and control horses of the present study, highlighting the necessity of evaluating and interpreting performance results of case horses in conjunction with that of control horses.
Fundamental differences exist in race length and speed between Quarter Horses and other racehorse breeds, and most Quarter Horses in the present study raced distances of ≤ 550 m, whereas Thoroughbreds frequently race distances of ≥ 1,600 m.1,12 Quarter Horses also race at speeds > 20% faster than Thoroughbreds.13 Exercise and airway dynamics at these specific speeds and distances have not been evaluated, and these faster speeds combined with shorter distances raced with Quarter Horses could alter how RLN impacts respiratory variables and performance in affected Quarter Horses, compared with other racehorse breeds. Nonetheless, findings regarding effects of prosthetic laryngoplasty on racing performance of horses in the present study were consistent with previous reports1–3,9 in Thoroughbred racehorses. For instance, results of the present study coincided with findings by Hawkins et al2 that show improvement in PI for Thoroughbred racehorses following prosthetic laryngoplasty and no identifiable difference in performance between horses treated with concurrent ventriculocordectomy. In addition, the finding in the present study that the highest PI after surgery was observed in case horses with grade II RLN was similar to findings reported by Witte et al1 that show higher RLN severity (ie, between grade III and IV) negatively impacts performance of Thoroughbred racehorses undergoing prosthetic laryngoplasty. Further, the proportion that returned to racing, interval from the surgery reference point to the first race, and number of racing starts in the postoperative period for case horses in the present study were comparable to findings for Thoroughbred racehorses following prosthetic laryngoplasty.1–3,9 Also similar to results of the present study, younger Thoroughbred racehorses or those that had not yet raced at the time of prosthetic laryngoplasty reportedly have a better prognosis for return to racing.11 Increased age of racing Quarter Horses negatively impacted return to racing in case and control horses in the present study, and the amount of attrition observed in the control horses suggested an industry-wide, age-related decline in racing performance for Quarter Horses. Similarly, decreased earnings for older Thoroughbred racehorses have been reported.9 Several variables (eg, musculoskeletal fatigue and injury from racing and training) could contribute to an overarching decline in racehorse performance with age.12
Limitations of the present study included reliance on medical records, in part, for data. Although data for all studied variables were available for each case horse, greater detail on laryngeal function and appearance was not available for case or control horses. We were unable to document the preoperative laryngeal function observed at exercise, the subgrade of preoperative resting laryngeal function, or the grade of postoperative laryngeal abduction for horses in the present study. Prospective studies, however, could allow purposeful recording of variables with predetermined interest, such as those used to estimate RLN severity. Further, a longer follow-up period would allow collection of more complete information on recovery, development of new pathological conditions, and relationships of airway function with racing career longevity.
Racing Quarter Horses of the present study performed similarly to racehorses of other breeds, indicating that surgeons should consider the possibility that a subset of racing Quarter Horses could be predisposed to decreased career performance following prosthetic laryngoplasty, as previously reported.3,14,15 There are a variety of reasons for decreased performance in horses following prosthetic laryngoplasty. In addition, failure of prosthetic laryngoplasty (eg, loosening of sutures or instability of the treated arytenoid cartilage) impacts breathing dynamics and patient outcomes after surgery.14 Inhalation of saliva, dust, and feed material are consequences of prosthetic laryngoplasty in racehorses because of an inability of the arytenoid cartilage to adequately seal and protect the upper airway, potentially resulting in inflammatory airway disease and negatively impacting performance.3,14,15 Even when outcomes of prosthetic laryngoplasty have been perceived to be beneficial, complex upper airway abnormalities after surgery have been documented with endoscopy at exercise.15 Therefore, endoscopic examinations in racing Quarter Horses during exercise, both before and after prosthetic laryngoplasty, is needed to fully understand the effects of RLN and surgery in these horses. In addition, we believe that extrapolating the findings of the present study to other racehorse populations would be difficult and that further study is needed to understand the impacts of RLN and surgical treatments for RLN in racing Quarter Horses.
Acknowledgments
Supported in part by the 2015 Young Investigator Grant Program, Center for Companion Animal Studies, Colorado State University. The results of the present study were independent of and not influenced by the funding source. The authors declare that there were no other conflicts of interest.
Presented in abstract form at the American College of Veterinary Surgeons Surgical Summit, Seattle, October 2016.
ABBREVIATIONS
| CI | Confidence interval |
| IQR | Interquartile (25th to 75th percentile) range |
| PI | Performance index |
| RLN | Recurrent laryngeal neuropathy |
Footnotes
Equibase, Equibase Company LLC, Lexington, Ky.
R: a language and environment for statistical computing, version 3.4.0, R Foundation for Statistical Computing, Vienna, Austria.
References
1. Witte TH, Mohammed HO, Radcliffe CH, et al. Racing performance after combined prosthetic laryngoplasty and ipsilateral ventriculocordectomy or partial arytenoidectomy: 135 Thoroughbred racehorses competing at less than 2400 m (1997–2007). Equine Vet J 2009;41:70–75.
2. Hawkins JF, Tulleners EP, Ross MW, et al. Laryngoplasty with or without ventriculectomy for treatment of left laryngeal hemiplegia in 230 racehorses. Vet Surg 1997;26:484–491.
3. Mason BJ, Riggs CM, Cogger N. Cohort study examining long-term respiratory health, career duration and racing performance in racehorses that undergo left-sided prosthetic laryngoplasty and ventriculocordectomy surgery for treatment of left-sided laryngeal hemiplegia. Equine Vet J 2013;45:229–234.
4. Hahn CN, Matiasek K, Dixon PM, et al. Histological and ultrastructural evidence that recurrent laryngeal neuropathy is a bilateral mononeuropathy limited to recurrent laryngeal nerves. Equine Vet J 2008;40:666–672.
5. Dixon PM, McGorum BC, Railton DI, et al. Clinical and endoscopic evidence of progression in 152 cases of equine recurrent laryngeal neuropathy (RLN). Equine Vet J 2002;34:29–34.
6. Compostella F, Tremaine WH, Franklin SH. Retrospective study investigating causes of abnormal respiratory noise in horses following prosthetic laryngoplasty. Equine Vet J Suppl 2012;43:27–30.
7. Leutton JL, Lumsden JM. Dynamic respiratory endoscopic findings pre- and post laryngoplasty in Thoroughbred racehorses. Equine Vet J 2015;47:531–536.
8. Fulton IC, Derksen FJ, Stick JA, et al. Treatment of left laryngeal hemiplegia in Standardbreds, using a nerve muscle pedicle graft. Am J Vet Res 1991;52:1461–1467.
9. Raffetto JA, Wearn JG, Fischer AT Jr. Racing performance following prosthetic laryngoplasty using a polyurethane prosthesis combined with a laser-assisted ventriculocordectomy for treatment of recurrent laryngeal neuropathy in 78 Thoroughbred racehorses. Equine Vet J 2015;47:60–64.
10. Robinson NE. Consensus statements on equine recurrent laryngeal neuropathy: conclusions of the Havemeyer Workshop. Equine Vet Educ 2004;16:333–336.
11. Carpenter RS, McIlwraith CW, Hill AE. Racing performance in 72 racehorses treated with prosthetic laryngoplasty for laryngeal hemiplegia. J Equine Vet Sci 2009;29:584–589.
12. Strand E, Martin GS, Haynes PF, et al. Career racing performance in Thoroughbreds treated with prosthetic laryngoplasty for laryngeal neuropathy: 52 cases (1981–1989). J Am Vet Med Assoc 2000;217:1689–1696.
13. Nielsen BD, Turner KK, Ventura BA, et al. Racing speeds of Quarter Horses, Thoroughbreds and Arabians. Equine Vet J Suppl 2006;36:128–132.
14. Davidson EJ, Martin BB, Rieger RH, et al. Exercising video-endoscopic evaluation of 45 horses with respiratory noise and/or poor performance after laryngoplasty. Vet Surg 2010;39:942–948.
15. 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.
