Yearling laryngeal function in thoroughbreds that underwent a laryngoplasty differs from controls.

Line
18 I would have the control group proportions first and the PL (treatment group) second
27 why is this a limitation? We are not looking at any outcome measure for this group we are only using them to see what distribution of grades are observed in yearlings in a random sample – it is presumed they have a spread of RLN based on the resting endoscopic grade and evaluation. Main limitation is the lack of performance data so the tieback group can be compared to the control horses.
59 Grade II.2 is not equivalent to grade 3 of 5 based on reference 24. The definitions in this reference do not have a grade whereby there is asymmetry at rest and full abduction is achieved and sustained. Therefore this should be amended to equivalent to grade III.1 on the 7- point scale
63 It could be argued that horses graded at yearling sales as III.1 or greater already had RLN – certainly III.2 or greater. Therefore it should be …YLF grades of horses that required a PL, to those of the control group
69 PL
70 YLF grade
87 Explain what dichotomous flow chart to assist in accurate categorisation means – also see line 279?
139 Figure 2 says 603 tiebacks performed
147 Should be….The percentage of males in the PL group was higher than in the unknown outcome group 78%vs56.5%
148 PL group
In table S1 is grade 0 VLAC no present, 1 left under right or right under left and 2 is bilateral? What is a Dixon et al (2003?) this should be reference 25?
167 I would document the the individual comparisions especially the III.1 vs 1
303 PL
Table 1 more tiebacks came from certain sales (if factors causing a horse to need a tieback are the same on any thoroughbred why is there a sale difference – must be due to type or genetics or age?)
Table 2 does not show percentages within the groups ie in the PL group and in the Unknown group
Not every III.1 needs a tieback
183 need better understanding of what this means. Is it important identical grades only 60.5% of the time disagreed on II.2 vs III.1 10% of the time – see discussion
190 ….almost half of the horses in the PL group had ‘normal” YLF (grades I and II.1) and a further 26% had an equivocal grade (grade II.2) at the time of sale.
192 In a previous study 93% of mixed breed mature racehorses with…..
Do we know what the laryngeal function grade was at the time of surgery ie is it in the surgical notes.
202 References needed for this. This concept is not new and reported elsewhere
203 The assumption is that the havemeyer grading system adequately categorises horses on resting laryngeal function into degree of underlying RLN
206 This would need to be validated by performance data – this study only looked at the risk of needing a PL. Therefore it might be better to relate the risk or needing a PL in this part of the discussion.
210 Because this study did not evaluate performance data a comparison between YLF grades and performance is not possible. It is only possible to say that there was no difference in the odds of requiring surgery.
211 Would it not be better to say…. There was no significant difference in the odds of requiring surgery between yearlings with grade I and grade II.1 laryngeal function. Coupled with the results of other studies that found no difference between these grades in terms of future performance[21,23], or arytenoid cartilage grade at exercise[32] it appears that categorising grades I and II.1 as low risk of developing clinically significant future laryngeal function is reasonable.
Comment: Is the risk of needing a tieback higher in horses with II.2 and III.1 when yearlings because they will have a higher level of subclinical pathology and are more prone to deterioration?. But if horses with grade III.1 stayed the same as racehorses the risk of arytenoid collapse on the treadmill to b or c is 44% so not all III.1 horses will be poor performers. Will need a prospective study to determine deterioration in grade, performance and the need for surgery
219 Again I think it needs to be pointed out these are mature horses with a history of poor performance and/or abnormal respiratory tract noise. Correct ….horses of horses
221 This is supported by reference 20
230 Please check these references carefully to make sure you can support the statements made. In reference 23 Significant differences in earnings were found between ‘normal’ and ‘abnormal’ (P = 0.02) and ‘inter- mediate-low’ and ‘abnormal’ grades (P = 0.03) while there were no significant differences between horses with ‘intermediate-high’ and ‘abnormal’ grades (P = 0.40). However, no significant differences were found between the low and high ‘intermediate’ grades (P = 0.60) or between horses with normal LF and either ‘intermedi- ate-low’ or ‘intermediate-high’ grades (P = 0.99).
Therefore, no differences in performance were found between horses graded II.2 and III.1
For reference 21 it should be noted the definition of the grading system used is as follows: Arytenoid function was initially graded by use of a modified scale of I to IV.7 Those grades were later retrofitted into a modified Havemeyer grading scale8 to standardize grading scores. Briefly, the Havemeyer scale is as follows: grade I, synchronous and symmetric arytenoid move- ment with full abduction achieved and maintained; grade II.1, transient asynchronous or asymmetric arytenoid movement with maximal abduction achieved and maintained; grade II.2, asynchronous or asymmetric arytenoid movement with maximal abduction achieved and maintained at times; grade III, asynchronous or asymmetric arytenoid movement with full abduction not achieved or not maintained; and grade IV, no arytenoid movement. The Havemeyer scale subdivides grade III into 3 subcategories, but the subcategories were not used in this study because there were too few yearlings that qualified for the subcategories.
The grade II.2 and III.1 do not match the system used in the current paper and therefore while there may be some similarities they are not the same. Care should be taken in extrapolating between this work and yours as I don’t think they are directly comparable.

240 This statement is not supported by the data in this study. The data confirms that grade III.1 horses have a greater risk for needing a tie-back. So yes future laryngeal dysfunction but we have no performance data.

Thank you for the thoroughness in answering my questions - see check list below

Review of responses to comments by reviewer 2 from first review.

Peer Reviewer: 2 Comments to the Author:
• 18: I would have the control group proportions first and the PL (treatment group) second.
Thank you for your comments. Group order has been altered.
THANK YOU

• 27 why is this a limitation? We are not looking at any outcome measure for this group we are only using them to see what distribution of grades are observed in yearlings in a random sample – it is presumed they have a spread of RLN based on the resting endoscopic grade and evaluation. Main limitation is the lack of performance data so the tieback group can be compared to the control horses.
*This is a limitation, because strictly speaking controls in a case:control study should be ‘disease-free’. As we cannot be sure that controls were not operated on elsewhere or retired, this could shift our findings towards the null (i.e. the reported differences in odds ratios between groups may underestimate the true difference). We have added the following:
27 Main limitations: Lack of performance data to compare the PL and control groups.

• 314: The main limitation of the study is the lack of racehorse performance data to compare the control group and PL group, as this would determine the true relevance of grading YLF.*

OK HAPPY WITH THIS

• 59 Grade II.2 is not equivalent to grade 3 of 5 based on reference 24. The definitions in this reference do not have a grade whereby there is asymmetry at rest and full abduction is achieved and sustained. Therefore this should be amended to equivalent to grade III.1 on the 7- point scale
*Respectfully disagree – this is a tricky area due to the sales companies continuing to use an outdated grading system in their conditions of sale, whereas the creator of the 5-point grading system (J.G. Lane) subsequently updated his definitions at the 2003 Havemeyer proceedings.
The definition of a grade 3 in the initial description of the grading system in 1993 only mentions the use of ‘partial asphyxiation using nostril closure’ to assess for full abduction, and yet clinically at sales most veterinarians use swallowing rather than nostril closure to grade laryngeal function. J.G. Lane’s updated definition of a grade 3 in 2003 Havemeyer proceedings includes assessment on swallowing AND there is no mention of sustained or maintained abduction. Grade II.2 and III.1 on the 7-point scale clearly fit into Lane’s 2003 updated 5-point grading system. Additionally, grade II.2 has been included in the grade 3 out of 5 category in the published literature (Ahern et al 2022).

Additionally, grade II.2 horses do not fit into the definition of grade 2 in both the 1993 and 2003 versions of the 5-point grading system: ie. “Grade 2: All major movements are symmetrical with a full range of abduction and adduction. Transient asynchrony, flutter or delayed or biphasic abduction may be seen especially by the left arytenoid. “ (Lane 1993); Or, “Grade 2: All major movements are symmetrical and a full range is achieved. Transient asynchrony, flutter or delayed opening may be seen (Baker 1983)” (Lane 2003).

In response to the reviewer’s comment, we have changed reference 24 to the 2003 version of the 5-point grading system.*

GOOD I AGREE
• 64 It could be argued that horses graded at yearling sales as III.1 or greater already had RLN – certainly III.2 or greater. Therefore it should be …YLF grades of horses that required a PL, to those of the control group
Change made.

GOOD

• 65 PL
Change made.

GOOD

• 66 YLF grade
Change made.

GOOD

• 89 Explain what dichotomous flow chart to assist in accurate categorisation means – also see line 279?
*See new text below.

THANKS

95: Due to previously identified disagreement between veterinarians when grading laryngeal function, a novel diagnostic decision tree, adapted from the 7-point grading system, was used when grading YLF.[10,32] This decision tree aims to assist with accurate categorisation of YLF and consists of a series of dichotomous (yes/no) decisions, that lead the observer down a pathway towards the most appropriate laryngeal functiongrade. The diagnostic decision tree was developed to clearly define certain parameters described in the 7-point grading system (i.e. the definition of maintenance of arytenoid cartilage abduction) and aims to encourage
observers to interpret the grading system similarly. The diagnostic decision tree is displayed in figure S1.*
THANKS

• 139 Figure 2 says 603 tiebacks performed.
*Correct. Figure 2 is showing the recruitment process for the PL group only. 603 horses underwent a PL, but only 150 had a post-sale scope available. There were 4 controls for each horse with a post-sale scope (600 controls). We have altered paragraph for clarity:

147: Out of the 603 horses that underwent a PL during the study period, 150 had a post-sale endoscopic exam available and were eligible for inclusion in the study. Recruitment data for the PL group are displayed in figure 2. Four controls were selected from the same sale as each PL case, therefore there were 600 horses in the unknown outcome group.*

THANKS
• 158 Should be….The percentage of males in the PL group was higher than in the unknown outcome group 78%vs56.5% Change made.

THANKS

• 161 PL group
Change made.
THANKS

• In table S1 is grade 0 VLAC no present, 1 left under right or right under left and 2 is bilateral?
Yes. Table modified.
THANKS

• What is a Dixon et al (2003?) this should be reference 25?
EVJ author guidelines 3.3.4.iv Tables states: the table, legend, and footnotes must be understandable without reference to the text. Therefore, we interpreted this to apply to references also. We have amended references in all figures and table legends to also include reference number.
OK

• 167 I would document the individual comparisons especially the III.1 vs 1.
The individual comparisons are documented in Table 3. EVJ author guidelines 3.3.4.iv state: Tables should summarise data to complement, not duplicate, information contained in the text.

OK

• 301 PL
Change made.

THANKS

• Table 1 more tiebacks came from certain sales (if factors causing a horse to need a tieback are the same on any thoroughbred why is there a sale difference – must be due to type or genetics or age?).
More post-sale scopes were available at certain sales, mainly due to the size of the sale, but also dependant on whether we had access to multiple years of post-sale scope data from same sale. Unfortunately, several years of post-sale data were not available for certain large sales due to data storage issues. The location of the clinic that performed the PL’s also factors into this, so we cannot make assumptions about genetics playing a role. The time-period chosen ensured all horses were a minimum of 5 years old at time of data analysis, therefore year of sale/age of horse should not be a factor.
OK

• Table 2 does not show percentages within the groups ie in the PL group and in the Unknown group.
Please see figure 3 for the percentage of horses with each laryngealfunction grade in both the unknown outcome and PL groups. Table 2 displays the frequency (and percentage) of horses with each LF grade across both groups.

OK

• Not every III.1 needs a tieback
Is there a specific line number that you are referring to? Clinically we agree not all III.1’s need a PL, but we are not able to prove this with our results, because the full medical history is not available for horses in the control group.

OK I WAS MAKING NOTES I THINK…
• 183 need better understanding of what this means. Is it important identical grades only 60.5% of the time disagreed on II.2 vs III.1 10% of the time – see discussion.
*Re-worded for clarity:

186: Observer 1 and observer 2 assigned identical YLF grades to the same endoscopic video in 454 horses (60.5%), and they differed by 1 grade in 295 horses (39.3%) and by 2 grades for 1 horse (0.1%). The most frequent disagreements were between grades I and II.1 in 184 horses (62.2% of the differences). The remainder of the disagreements were mainly between grades II.2 and III.1 (85 horses, 28.7% of differences) and grades III.1 and III.2 (24 horses, 8.1% of differences). The observers disagreed once between grades III.2 and III.3, grades
III.3 and grade IV and grades III.2 and II.2.

The following has been added into the discussion, to address potential implications of interobserver disagreement:*

308: Whilst the majority of interobserver disagreements were between YLF grades I and II.1, these are of minimal concern. However, disagreements between higher YLF grades have greater implications given the difference in risk of a PL between grades II.2 and above. In high stakes scenarios such as yearling sales, it may be more appropriate for yearlings to be assigned a consensus YLF grade from a panel of experienced veterinarians, rather than relying on a single observer. Future investigations should focus on more objective methods of evaluating YLF, such computer assisted assessment.

AGREED
• 200 ….almost half of the horses in the PL group had ‘normal” YLF (grades I and II.1) and a further 26% had an equivocal grade (grade II.2) at the time of sale.
Change made.

THANKS
• 202 In a previous study 93% of mixed breed mature horses with…..
Change made.

THANKS

• Do we know what the laryngeal function grade was at the time of surgery ie is it in the surgical notes.
No, unfortunately not.

OK

• 212 References needed for this. This concept is not new and reported elsewhere.
Added references 2-5.

THANKS

• 203 The assumption is that the havemeyer grading system adequately categorises horses on resting laryngeal function into degree of underlying RLN.
*Edited the sentence as follows:

212: In other words, assuming resting laryngeal function grade accurately correlates horses with the appropriate degree of RLN,[39] yearlings should not be categorised into either an affected group or a disease-free group, but instead considered to exist at different points along the disease pathway with different odds of requiring a PL.*
YES THANKS IT IS REF 38 WHICH IS CORRECT IN THE MANUSCRIPT

• 206 This would need to be validated by performance data – this study only looked at the risk of needing a PL. Therefore it might be better to relate the risk or needing a PL in this part of the discussion.
*Altered to say:

AGREE

216:…with different odds of requiring a PL. A risk-based assessment may therefore be more appropriate when examining YLF at sales, but this needs to validated with performance data and is a limitation of this study.*

AGREE
• 210 Because this study did not evaluate performance data a comparison between YLF grades and performance is not possible. It is only possible to say that there was no difference in the odds of requiring surgery.
Ok. See below.

• 220 Would it not be better to say…. There was no significant difference in the odds of requiring surgery between yearlings with grade I and grade II.1 laryngeal function. Coupled with the results of other studies that found no difference between these grades in terms of future performance[21,23], or arytenoid cartilage grade at exercise[32] it appears that categorising grades I and II.1 as low risk of developing clinically significant future laryngeal function is reasonable.
Yes, thank you. Changes made.

THANKS

• Comment: Is the risk of needing a tieback higher in horses with II.2 and III.1 when yearlings because they will have a higher level of subclinical pathology and are more prone to deterioration?. But if horses with grade III.1 stayed the same as racehorses the risk of arytenoid collapse on the treadmill to b or c is 44% so not all III.1 horses will be poor performers. Will need a prospective study to determine deterioration in grade, performance and the need for surgery.
Yes, we agree very few adult horses with grade III.1 LF will have grade B or C arytenoid cartilage collapse requiring a PL. However, yearlings with grade II.2 or III.1 YLF are considered to already be further along the disease pathway, than yearlings with grade I or II.1 laryngeal function. Therefore, given the degenerative nature of the condition this puts II.2 and III.1 at increased risk of future laryngeal dysfunction.

YES AGREED

230 Again I think it needs to be pointed out these are mature horses with a history of poor performance and/or abnormal respiratory tract noise.
*Changes made
THANKS

230: Similarly, Barakzai and Dixon (2011) found 88.5% of mature horses that presented with a history of poor performance and/or abnormal respiratory noise, with grade III.2 or above resting laryngeal function, had arytenoid cartilage collapse (grade B or C) during exercise.*

• Correct ….horses of horses
Corrected
THANKS

• 234 This is supported by reference 20
Changes made.
THANKS

• 230 Please check these references carefully to make sure you can support the statements made. In reference 23 Significant differences in earnings were found between ‘normal’ and ‘abnormal’ (P = 0.02) and ‘inter- mediate-low’ and ‘abnormal’ grades (P = 0.03) while there were no significant differences between horses with ‘intermediate-high’ and ‘abnormal’ grades (P = 0.40). However, no significant differences were found between the low and high ‘intermediate’ grades (P = 0.60) or between horses with normal LF and either ‘intermedi- ate- low’ or ‘intermediate-high’ grades (P = 0.99).
Fair point. Reference 23 removed.

THANKS

• Therefore, no differences in performance were found between horses graded II.2 and III.1
See above.

GOOD

• For reference 21 it should be noted the definition of the grading system used is as follows: Arytenoid function was initially graded by use of a modified scale of I to IV.7 Those grades were later retrofitted into a modified Havemeyer grading scale8 to standardize grading scores. Briefly, the Havemeyer scale is as follows: grade I, synchronous and symmetric arytenoid move- ment with full abduction achieved and maintained; grade II.1, transient asynchronous or asymmetric arytenoid movement with maximal abduction achieved and maintained; grade II.2, asynchronous or asymmetric arytenoid movement with maximal abduction achieved and maintained at times; grade III, asynchronous or asymmetric arytenoid movement with full abduction not achieved or not maintained; and grade IV, no arytenoid movement. The Havemeyer scale subdivides grade III into 3 subcategories, but the subcategories were not used in this study because there were too few yearlings that qualified for the subcategories.
See below.

• The grade II.2 and III.1 do not match the system used in the current paper and therefore while there may be some similarities they are not the same. Care should be taken in extrapolating between this work and yours as I don’t think they are directly comparable.
The following changes have been made:
248: Similarly, another study identified horses with grade II.2 YLF outperformed horses with grade III YLF as a 3- and 4-year-old, but they earnt significantly less as a 4-year old compared to horses with grade I and II.1 YLF.[21] The aforementioned study did not did not separate horses with grade III YLF into subcategories due to study numbers, therefore comparisons between their findings and ours requires caution.

GOOD

• 240 This statement is not supported by the data in this study. The data confirms that grade
III.1 horses have a greater risk for needing a tie-back. So yes future laryngeal dysfunction but we have no performance data.
*We have altered this sentence to:

261: Our findings indicate there is a significance difference in risk of future laryngeal dysfunction between horses with grade II.2 and III.1 YLF, however an appropriately powered study investigating race performance data is required to determine if this difference in risk translates into reduced earnings on the racetrack*

OK THANKS