Assessment of menace response in neurologically and ophthalmologically healthy cats

Abstract

Objectives

Assessment and interpretation of menace response (MeR) in cats can be challenging. The prevalence of abnormal MeR in healthy cats is unknown. The aim of this study was to prospectively evaluate MeR in visually healthy cats.

Methods

Fifty cats without history or clinical evidence of neurological or ophthalmological disease were assessed by two examiners: standing behind the cat (mode A), in front of the cat (mode B), and in front of the cat, covering the contralateral eye (mode C). MeR was scored from 1–5 (absent, weak, moderate, strong, complete). Examination modes were compared concerning presence and score (descriptive statistic, 95% confidence interval, χ² test). This was compared to a three-level scoring system (negative, reduced, positive). Score reproducibility between the two examiners was assessed (Cohen’s kappa [κ] test). Video footage allowed self-re-evaluation and evaluation of the second examiner (κ analysis). Learning/tiring effect (McNemar test), influence of age, body weight (Spearman’s rho test), skull type (χ² test) and being an indoor or outdoor cat (Mann–Whitney U-test) were evaluated.

Results

MeR was always elicited with at least one technique. Comparable results were obtained with the five- and three-level scoring systems. Mode A achieved strong/complete (positive) MeR in 84.5%, mode B in 82% and mode C in 60%. Exact score reproducibility between the two examiners was slight to fair (κ = 0.208–0.281). Intrarater agreement for video self-assessment (κ = 0.544–0.639), as well as inter-rater agreement (extrinsic video assessment), was moderate to substantial (κ = 0.584–0.645). No learning/tiring effect (P = 0.530) or association with body weight (P = 0.897), age (P = 0.724), skull type (P >0.05) and being an indoor/outdoor cat (P = 0.511) were evident.

Conclusions and relevance

The majority of visually healthy cats revealed a strong/complete MeR when the contralateral eye remained uncovered, but 40% failed when the contralateral eye was covered. The most reliable examination mode was achieved standing behind the cat.

Introduction

Assessment and interpretation of the menace response (MeR) is an important tool in evaluating the visual system in dogs and cats.^1–3 However, in cats, this part of the neurological examination can be very challenging, especially when they are exposed to a stressful environment such as a veterinary practice. In our experience, cats often do not perform the MeR as reliably as dogs. To our knowledge, no investigation has been reported that has systematically assessed the reliability of the feline MeR. The aim of this prospective study was to evaluate MeR in neurologically and ophthalmologically normal cats and to define the most reliable and reproducible mode of examination.

Materials and methods

Inclusion criteria

The cohort was selected from cases that presented to the Centre for Clinical Veterinary Medicine, Faculty of Veterinary Medicine, LMU Munich, between June and October 2014.

Inclusion criteria for this prospective study were no history or clinical evidence of neurological or ophthalmological abnormalities; no currently affected demeanour by profound disease; and age over 3 months, to ensure that the cat had developed a MeR.^1–4 If the owner reported abnormal vision, behaviour, gait or jumping ability, the cats were excluded. All examinations were performed in a quiet room with no other animals present. Vision was tested via observation of the cat’s behaviour and orientation when moving in a room with obstacles, via laser pointer and/or the cotton-ball test. If any of these tests were questionable or negative, the cat was excluded. The remaining full neurological examination and ophthalmological assessment of adnexa (eyelids, third eyelid, conjunctiva), ⁴ anterior segment (cornea, anterior chamber, iris, lens) and fundus by direct ophthalmoscopy had to be unremarkable. Cats with documented hypertension were excluded, even if the neurological and ophthalmological examinations were normal, in order to rule out the possibility of early hypertensive encephalopathy.^5,6

All cats were assigned to a stress level according to cat stress score by Van Haaften et al. ⁷

Follow-up via telephone conversation and clinical notes were obtained where available.

Assessment mode and scoring of MeR

Clinical and hands-off neurological examinations were done first. Then, the palpebral reflex was tested. Next, the MeR assessment was performed, followed by the remaining parts the neurological and ophthalmological examination.

The MeR of each cat was assessed by two examiners in random order. For this purpose, lots were drawn. Examiner 1 was always the same person (PQ, Neurology Resident [ECVN]). Examiner 2 varied between two persons (CT, SB [both second-year postgraduate students]). The examiners were supervised by a board-certified (ECVN) neurologist (LM).

Three modes of performing the MeR were evaluated: (1) standing behind the cat (mode A; Figure 1a); (2) standing in front of the cat (mode B; Figure 1b); and (3) standing in front of the cat, covering the contralateral eye (mode C; Figure 1c).

Figure 1

Three modes of assessing the menace response (MeR): (a) assessing the MeR standing behind the cat (the examiner is the person seen behind the cat; mode A); (b) assessing the MeR standing in front of the cat (mode B); (c) assessing the MeR standing in front of the cat, covering the contralateral eye (mode C)

The order of the examination mode, as well as starting the examination with the left or right eye, were randomly chosen (via drawing lots). Examiner 1 was left-handed and the other two examiners were right-handed. MeR was performed five consecutive times on each eye by both examiners, with the fingers spread to cause minimal air puff and touching lashes and whiskers was avoided. The whole procedure was recorded on video for subsequent self-assessment (intrinsic video assessment) and assessment by the other examiner (extrinsic video assessment). Video assessment was always blinded to the previous results.

MeR score

MeR was scored on a scale of 1–5, where 1 = absent (complete absence of any [upper and lower] eyelid movement); 2 = weak (subtle upper eyelid movement); 3 = moderate (subtle movement of both [upper and lower] eyelids); 4 = strong (strong movement of both eyelids but incomplete closure of palpebral fissure); and 5 = complete (strong movement of both eyelids with complete closure of the palpebral fissure).

The maximum score achieved by each examiner was used for further comparisons.

Data analysis and statistics

A sample size of 50 cats was calculated by power analysis based on a minimum power of 80%, a maximum alpha error of 5% and the following minimum effective sizes: Cohen’s kappa 0.4, Spearman’s rho 0.4.

Modes of examination were compared relative to their presence and MeR score. The results of examiners 1 and 2 were added for each examination mode on each eye. A descriptive statistic was performed by determining the relative frequency for each score value. Ninety-five per cent confidence interval (CI) was calculated. The χ²test was used to compare scores between examination modes. To assess a more practical setting for daily clinics, a three-level scoring system was applied comparing modes of examination relative to their presence and score. Therefore, scores were assigned as follows: 1 = negative (according to score 1 [absent] of the five-level scoring system); 2 = reduced (including score 2 [weak] and 3 [moderate] of the 5-level scoring system); and 3 = positive (including score 4 [strong] and 5 [complete] of the 5-level scoring system). Statistics was performed as described above for the five-level scoring system.

Score reproducibility between the two examiners was calculated for all examination modes by Cohen’s κ test and the absolute difference in score stated as delta score.

Intra- and inter-rater agreement (intrinsic and extrinsic video assessment) was calculated using Cohen’s κ test.

Kappa agreement between the left and right eye was compared for each examination mode and the score difference was evaluated. Results of the two examiners were added together and the 95% CI. The difference between the left and right eye scores was calculated (Wilcoxon test) as a methodical control to exclude a systemic variation caused by different examiners and handedness.

McNemar’s test was used to evaluate possible learning or tiring effects of the MeR. Therefore, the first examination mode of the first examiner was compared to the last examination mode of the second examiner for each eye in all 50 cats assessing 200 MeR examinations altogether.

In addition, we assessed whether there was a difference in MeR presence and score between brachycephalic and mesaticephalic cats (χ² test).The influence of age and body weight on the MeR score was evaluated with Spearman’s rho test. Mann–Whitney U-test was used to compare the MeR scores of indoor and outdoor cats.

The complete statistical analysis was performed by a professional biostatistician (SR) using commercial software (SPSS Statistics for Windows, version 21.0; IBM). P-value was considered significant if it was <0.05. For Cohen’s κ tests, the strength of agreement was interpreted as follows: almost perfect agreement (κ = 0.81–1.00), substantial agreement (κ = 0.61–0.80), moderate agreement (κ = 0.41–0.60), fair agreement (κ = 0.21–0.40), slight agreement (κ = 0.00–0.20) and poor agreement (κ <0.00).^8,9

Results

Signalment and cause of presentation

Fifty cats were included in the study. Age of the cats ranged from 0.5 to 15.4 years (mean 5.7 years). Sex distribution was as follows: female entire (n = 2), female neutered (n = 18), male entire (n = 4) and male neutered (n = 26). Cat breeds included domestic shorthair (n = 27), Bengal (n = 5), Birman (n = 2), British Short Hair (n = 2), Chartreux Blue (n = 2), Maine Coon (n = 1), Norwegian Forest Cat (n =1), Persian (n = 1), Ragdoll (n = 1) and Siamese (n =1), as well as Persian mix (n = 4) and domestic shorthair mix (n = 3). Body weight ranged from 2.1 to 7.9 kg (mean 4.6 kg). Eighteen of the 50 cats were outdoor cats and 32 were indoor cats.

Cats were presented to various clinical services: general health check and vaccination (n = 24), cardiology (n = 8; seven for purpose of breeding certification, one for a recheck of restrictive cardiomyopathy), dermatology (n = 6), oncology (n = 2; one for assessment of suspected interscapular fibrosarcoma, the other with cutaneous haemangiosarcoma) and internal medicine (n = 10; four for urinary tract infection recheck, one for urolithiasis, two for diarrhoea, two for vomiting and one for recheck of coughing). All cats were fairly relaxed upon examination, consistent with either score 0 (no resistance to handling) or score 1 (minimally resistant to handling). ⁷

In 38/50 cats, follow-up via clinical notes or owner telephone conversation was available and revealed that no evidence of visual impairment, ophthalmological or neurological disease had developed in a 1 year period after assessment of MeR. In 10 cats no follow-up data could be obtained, one cat was euthanased 1 month after MeR assessment because of cardiac failure (until then no evidence of neurological or ophthalmological disease had occurred) and another cat was euthanased 2 months after MeR assessment owing to feline infectious peritonitis (there was no histopathological evidence of eye or brain involvement and no visual impairment was reported).

MeR presence and score

MeR was present in each cat with at least one of the three examination modes. It was absent/negative in 0.5% with examination mode A and in 1.0% with examination mode C. It was always present with examination mode B.

Using the five-level scoring system, the degree of elicitability had a wide range, from 2 (weak MeR) to 5 (strong movement of both eyelids with complete closure of the palpebral fissure). Comparing the different examination modes, a score of 5 was achieved in 58.0% and 57.5% of cats with mode A and B, respectively, whereas examination mode C achieved a score of 5 in 35.5% of cats. MeR score for each examination mode, including CIs, are given in Table 1. The same analysis was repeated for the 38 cats where a 1 year follow-up could be obtained, and reflects similar results (see supplementary material Table 1).

Table 1

Percentages of presence and strength of menace response (MeR) with three examination modes in 200 MeR examinations using a five-level scoring system

Examination mode
Score	Mode A (%)	95% CI	Mode B (%)	95% CI	Mode C (%)	95% CI
1	0.5	0.0–1.5	0.0	0.0–1.4	1.0	0.0–2.5
2	2.0	0.5–10.6	6.0	1.3–16.5	16.0	7.2–29.1
3	13.0	5.8–26.7	12.0	4.5–24.3	23.0	11.5–36.0
4	26.5	14.6–40.3	24.5	13.1–38.1	24.5	13.1–28.2
5	58.0	43.2–71.8	57.5	43.2–71.8	35.5	22.9–50.8

MeR score: 1 = absent; 2 = weak; 3 = moderate; 4 = strong; 5 = complete

Mode A = standing behind the cat; mode B = standing in front of the cat; mode C = standing in front of the cat, covering the contralateral eye; CI = confidence interval

Using the 3-level scoring system on the 50 cats examined, a positive MeR was achieved in 84.5% with mode A, in 82.0% with mode B and in 60.0% with mode C. MeR score for each examination mode, including CIs, are given in Table 2.

Table 2

Percentages of presence and strength of menace response (MeR) with three examination modes in 200 MeR examinations using a three-level scoring system

Examination mode
Score	Mode A (%)	95% CI	Mode B (%)	95% CI	Mode C (%)	95% CI
1	0.5	0.0–1.5	0.0	0.0–1.4	1.0	0.0–2.5
2	15	10.5–20.0	18.0	12.5–23.5	39.0	32.5–45
3	84.5	79.0–89.5	82.0	76.5–87.5	60.0	53.5–66.5

MeR scores: 1 = negative; 2 = reduced; 3 = positive

Mode A = standing behind the cat; mode B = standing in front of the cat; mode C = standing in front of the cat, covering the contralateral eye; CI = confidence interval

With both, the 5- and 3-level scoring system, the relative frequency of the score presence (χ2 test) showed a significant difference between mode A and C (P <0.001) and between mode B and C (P <0.001). There was no significant difference between mode A and B using the 5-level (P = 0.919) and 3-level (P = 0.445) scoring system.

Score reproducibility between the two examiners

The exact score reproducibility between the two examiners was fair for examination modes A (κ = 0.219; P = 0.002) and C (κ = 0.281; P <0.001), and slight for examination mode B (κ = 0.208; P = 0.002). A complete agreement was achieved in 55% for examination mode A, in 53% for mode B and in 47% for mode C. The exact reproducibility of each examination mode is displayed in detail in Table 3.

Table 3

Agreement for the exact score reproducibility of the examination modes between the two examiners

Agreement between the two examiners
Delta score	Mode A (%)	Mode B (%)	Mode C (%)
None	55	53	47
1	34	34	42
2	10	8	9
3	0	5	1
4	0	0	1

The absolute difference in score is stated as delta score

Mode A = standing behind the cat; mode B = standing in front of the cat; mode C = standing in front of the cat, covering the contralateral eye

Intra- and inter-rater agreement

Intra-rater agreement for video self-assessment was substantial for examination mode A (κ = 0.639; P <0.001) and C (κ = 0.620; P <0.001), and moderate for mode B (κ = 0.544; P <0.001) (see supplementary material Table 2).

Inter-rater agreement for extrinsic video assessment was moderate for examination mode A (κ = 0.597; P <0.001) and B (κ = 0.584; P <0.001), and substantial for mode C (κ = 0.645; P <0.001) (see supplementary material Table 3).

Comparison of the left and right eye

The MeR examination for the left and right eye had a fair agreement for all examination modes

(mode A: κ = 0.365 [P <0.001]; mode B: κ = 0.371 [P = 0.018] and mode C: κ = 0.400 [P <0.001]). An equal result for both eyes was obtained in 63% with mode A and B, and in 56% with mode C. The differences of score for the left and right eye for each examination mode is listed in Table 4 in the supplementary material.

There was no significant difference with the Wilcoxon test concerning the mean ranks between the examinations of the left and the right eye (P >0.05).

Learning or tiring effect of the MeR

There was no learning or tiring effect of the MeR noticed in the cats when comparing the first and the last examination (P = 0.530).

The influence of skull formation, being an indoor/outdoor cat, age and body weight on MeR

When differentiating between brachycephalic and mesaticephalic cats no significant difference was found for the three examination modes (P >0.05).

No significant difference existed between indoor and outdoor cats (P = 0.511).

There was no significant influence of body weight (rho = 0.019, P = 0.897) or age (rho = 0.051, P = 0.724) on MeR.

Discussion

The MeR is a learned response and is present at 10–12 weeks of age in kittens.^1,2 Its pathway is complex and requires a normal function of the eye and peripheral and central parts of the nervous system (Figure 2). An abnormal MeR can therefore occur as a result of dysfunctions in different brain areas, and a very thorough examination is necessary to locate a lesion correctly.

Figure 2

Dorsal view of the neuroanatomical pathway for the menace response. 1 = retina; 2 = optic nerve; 3 = optic chiasm; 4 = optic tract; 5 = lateral geniculate nucleus; 6 = optic radiation; 7 = occipital cortex; 8 = association fibres; 9 = primary motor cortex; 10 = projection fibres; 11 = pontine nucleus; 12 = cerebellar cortex; 13 = red nucleus; 14 = facial nucleus; 15 = facial nerve; 16 = orbicularis oculi muscle

The visual pathway takes its origin in the multi-layered sensory neuroepithelium of the retina. Axons of the retinal ganglion cells form the myelinated optic nerve and leave the eyeball at the optic disc.^1,3,10,11 In cats, 65% of optic nerve axons decussate in the optic chiasma to the opposite side.^1,12,13 Crossing axons arise from the medial (nasal) aspect of the retina. The remaining axons of the lateral (temporal) part of the retina stay ipsilateral.^14–18 Kondo et al ¹⁹ found that single retinal ganglion cells of the temporal retina simultaneously cross to both sides of the chiasm in cats. From the optic chiasm axons continue as optic tracts coursing caudo-dorsolateral to the diencephalon. In the cat, about 80% of the optic tract axons synapse in the lateral geniculate nucleus. ²⁰ From here projection via the optic radiation takes place into the visual cortex (occipital lobes).^21,22 Axons arising from the visual cortex reach the primary motor cortex via association fibres. From here, impulses are transmitted to the pontine nucleus. Pontine nuclei axons then decussate as transverse fibres to enter the cerebellum via the middle cerebellar peduncle projecting to the cerebellar cortex (ipsilateral to the tested eye). Cerebellar efferents (originating from Purkinje cells) then synapse (as some authors suggest, via the contralateral red nucleus) ²³ with the ipsilateral motor nucleus of the facial nerve. The latter innervates the orbicularis oculi muscle, which is responsible for the closure of the palpebral fissure during the MeR.^1,3,11

Considering this complexity of the MeR pathway, a normal neurological and ophthalmological examination (excluding MeR) was inclusion criteria for this study. Ability of orientation in an unfamiliar environment and appropriate response to a laser pointer or cotton-ball test was prerequisite. However, if cats are anxious and stressed, they might not perform adequately. In our study, all examinations were performed in a quiet room without any other animals present, and animals were all fairly relaxed. Results might therefore differ when performed under more stressful environmental conditions. Further studies evaluating MeR assessment on a cohort of cats where different levels of stress are applied to each individual cat would be necessary to elucidate this issue further.

Many textbooks suggest covering the contralateral eye when the MeR is assessed.^1,3,4 Results of our study show, that neurologically and ophthalmologically healthy cats are more likely to have a strong MeR with complete closure of the tested eye if both eyes remain uncovered. Best results, although not statistically significant, were achieved when standing behind the cat (examination mode A). Subjectively, with this technique, cats were the most relaxed. Results of intra-rater agreement also support this examination mode. Therefore, our recommendation for assessing the MeR in cats is examination mode A (Figure 1a). This might be the technique perceived by cats as the least threatening, as it requires less physical contact to the head than mode C, and the veterinarian is not standing directly in front of the cat, as with examination mode B. However, when assessing visually impaired cats, one also needs to consider the neuroanatomical complexity of the MeR, as different techniques may stimulate subtle different parts of the optic pathways and thus be more or less sensible to detect deficits.

In our study, a strong/complete MeR was achieved in 84.5% (mode A), 82.0% (mode B) and 60.0% (mode C). As this cohort only included visually healthy cats, MeR might need to be considered as ‘normal’, even if eyelid closure is incomplete. Another important observation is that every cat reacted to a certain degree to at least one of the examination modes. Therefore, if MeR is absent in all three modes, an underlying neurological or ophthalmological problem should be suspected. When comparing the MeR of the left and right side in our study, the probability of a one-score difference was fairly high (up to 38% depending on examination mode). A score difference of two or more, however, occurred in only 5–12%. Therefore, a score difference of two or more between the left and right eye is likely to be a relevant clinical finding, whereas subtle asymmetry should be considered clinically irrelevant. However, these conclusions warrant further validation by optimal cut-off points. Based on our results for reproducibility of MeR between two examiners and inter-rater agreement, we would advise that repeat examinations in view of disease progression should be performed by the same person.

Overall, in our healthy cohort there was no learning or tiring effects of MeR examination, which might have been expected. Moreover, one could expect that the physical condition and ‘lifestyle’ of a cat (eg, being young, athletic and active/outdoors vs being aged, overweight and indoors) could influence MeR performance. But, interestingly, neither age nor body weight or whether being an indoor/outdoor cat had significant influence on MeR presence and score. In addition, there was no significant difference between brachycephalic and mesaticephalic cats in any of the MeR examination techniques.

For our study design, a five-level scoring system was used for MeR evaluation and all comparisons. This offers accuracy in detecting smaller variations when assessing data such as score reproducibility, intra- and inter-rater agreement, or comparing results for the left and right eye as a methodical control. However, in order to assess a more practical setting for daily clinics, an additional three-level scoring system was retrospectively applied to evaluate MeR presence and score for the different examination techniques. Results were comparable to those of the five-level scoring system, and the power of conclusions has not changed. Therefore, applying a three-level scoring system (distinguishing between positive, reduced or negative MeR) in a clinical setting seems appropriate.

Conclusions

Every cat of our cohort, which were all visually healthy, reacted to at least one of the three examination modes. The majority revealed a strong/complete MeR when the contralateral eye remained uncovered upon assessment, but 40% failed when the contralateral eye was covered. The most reliable examination mode was achieved standing behind the cat, even though in 15% of cats closure of the eyelids was not complete. These results should be considered when assessing cats for ophthalmological or neurological disease. The MeR performance of our cohort was independent of age, body weight, being an indoor/outdoor cat and skull formation. For a clinical setting, a three-level scoring system for evaluation of feline MeR (positive, reduced, negative) is sufficient.

Supplemental Material

Supplementary Table 1

Percentages of presence and strength of menace response (MeR) with three examination modes in 152 MeR examinations (mode A = standing behind the cat; mode B = standing in front of the cat; mode C = standing in front of the cat, covering the contralateral eye). Only the 38 cats for which follow-up information 1 year after MeR assessment could be obtained, and with no neurological or ophthalmological problems, were included in this calculation.

Supplemental Material

Supplementary Table 2

Intra-rater agreement for video selfassessment with the three examination modes. Calculations are based on 200 cat eyes, resulting from 50 evaluated cats per examiner

Supplemental Material

Supplementary Table 3

Inter-rater agreement for extrinsic video assessment with three examination modes. Calculations are based on 200 cat eyes resulting from 50 evaluated cats per examiner

Supplemental Material

Supplementary Table 4

Comparison of menace response score differences (%) in the left and right eyes with the three examination modes