Pectinate ligament dysplasia (PLD) is significantly associated with primary closed angle glaucoma (PCAG) in several dog breeds. Gonioscopy screening for PLD is advised in breeds in which PLD and PCAG are particularly prevalent in order that affected dogs may be eliminated from the breeding population. The Border Collie (BC), Hungarian Vizsla (HV) and Golden Retriever (GR) breeds are currently under investigation for PLD by the British Veterinary Association/Kennel Club/International Sheep Dog Society (BVA/KC/ISDS) Eye Scheme. The authors aimed to determine the prevalence of PLD in UK populations of BC, HV and GR and to investigate possible associations between the degree of PLD and age and sex. Gonioscopy was performed in 102 BCs, 112 HVs and 230 GRs and the percentage of iridocorneal angle affected by PLD was estimated and classified as unaffected (0 per cent), mildly affected (<20 per cent), moderately affected (20–90 per cent) or severely affected (>90 per cent). Eleven of 102 (13.8 per cent) BCs, 16/112 (14.3 per cent) HVs and 60/230 (26.1 per cent) GRs were moderately or severely affected by PLD. The prevalence of PLD was significantly higher in GR than both BC and HV. There was a significant positive correlation between PLD and age in the HV and GR but not in the BC. There was no association between PLD and sex in any breed.
- Pectinate ligament dysplasia
- Hungarian Vizsla
- Golden Retriever
- Border Collie
Statistics from Altmetric.com
Pectinate ligament dysplasia (PLD) is an abnormality of the iridocorneal angle (ICA) and is associated with primary closed angle glaucoma (PCAG) in several dog breeds (Cottrell and Barnett 1988; Ekesten and Narfström 1991; Wood and others 1998; Ahonen and others 2013). PLD is highly heritable and both PLD and PCAG are prevalent in several breeds (Ekesten and Narfström 1991, 1992; Wood and others 1998; Bjerkås and others 2002; Gelatt and MacKay 2004; Fricker and others 2016; Oliver and others 2016a, b). Gonioscopy describes the examination of the ICA and is the method employed to detect PLD (Bedford 1977).
The British Veterinary Association/Kennel Club/International Sheep Dog Society (BVA/KC/ISDS) Eye Scheme is based on eye examination and is a means of identifying inherited ocular conditions in dogs, such as PLD, with the aim being to reduce the incidence of inherited eye disease over time through selective breeding. The Eye Scheme currently operates two lists, or schedules, of breed-related eye diseases. Schedule A lists the known inherited eye diseases in the breeds where there is enough scientific information to show that the condition is inherited and Schedule B lists those breeds in which the conditions are, at this stage, only suspected of being inherited and therefore are listed as ‘under investigation’. The Eye Scheme currently lists nine breeds as certifiable for PLD on Schedule A and recommends gonioscopy in these breeds before breeding. A further seven breeds, including the Border Collie (BC), Hungarian Vizsla (HV) and Golden Retriever (GR), are included on Schedule B as under investigation for PLD.
The authors have recently reported PLD prevalence in UK populations of Welsh Springer Spaniel, Basset Hound, Dandie Dinmont Terrier and Flatcoated Retriever dogs (Oliver and others 2016a, b). Seventy-six of 198 (38.4 per cent) Basset Hounds, 36/170 (21.2 per cent) Flatcoated Retrievers and 21/95 (22.1 per cent) Dandie Dinmont Terriers were moderately or severely affected by PLD and would be considered as ‘affected’ for PLD under the Eye Scheme. In all breeds, there was a significant positive correlation of the degree of PLD with age. This finding was suggestive of PLD progression and this is supported by a separate cross-sectional study in the Leonberger and also by longitudinal evidence of PLD progression over time in individual Flatcoated Retrievers and Welsh Springer Spaniels (Pearl and others 2015; Fricker and others 2016; Oliver and others 2016a). This recent demonstration of PLD progression has caused the Eye Scheme to reconsider its advice on the frequency of gonioscopy screening in predisposed breeds. Gonioscopy was originally advised as ‘once in a lifetime test’ for predisposed breeds but now, following the initial study of PLD progression in the Flatcoated Retriever, the examination is advised every three years in this breed and this advice may be extended to other breeds as more information comes to light (Oliver and others 2016a).
The association of PLD with sex is less clear. A study into breed-related glaucomas in North America reported a predominance of affected females in some breeds and a predominance of affected males in others (Gelatt and MacKay 2004). The reason for this is unclear but may relate to differences in anterior chamber depth between males and females, because, in the majority of breeds investigated thus far, PLD prevalence has been found to not be significantly different between male and female dogs (Ekesten and Narfström 1991; Wood and others 1998; Bjerkås and others 2002; Tsai and others 2012; Oliver and others 2016a, b).
In this study, the authors examined large numbers of BCs, HVs and GRs. The aims of the study were to estimate PLD prevalence and to investigate possible associations between PLD and age and sex in these breeds.
Study design and population
This study was designed as a prospective cross-sectional study and was performed as the authors have previously described (Oliver and others 2016a, b). Three different breeds of dogs were enrolled during gonioscopy screening sessions at different locations across the UK between January and December 2015 between the hours of 10:00 and 16:00 GMT. To the best of the authors’ ability, the authors recruited BCs, HVs and GRs onto the present study that were representative of the UK populations of these breeds using the same strategies as the authors have previously reported (Oliver and others 2016a, b). Gonioscopy screening sessions were undertaken in different locations around the UK and at different types of events, including dog shows, ‘fun days’, kennel visits and breed information days. The gonioscopy screening was promoted by a variety of different mechanisms, including correspondence from the KC to the owners of KC registered dogs of each breed, via breed club websites and via social media. All dogs that were volunteered for screening were accepted, regardless of their age, ancestry or KC registration status. Dogs enrolled in the study included: 102 BCs, 112 HVs and 230 GRs. For each dog, data were collected on sex and age at the time of examination. All dogs were pets and ophthalmological examination was only performed after informed and written owner consent.
Gonioscopy was performed bilaterally in conscious dogs following application of 0.5 per cent proxymetacaine to the ocular surface (Bausch & Lomb, Chauvin Pharmaceuticals, Aubenas, France) using a 19 mm Koeppe goniolens (Ocular Instruments, Redmond, Washington, USA) filled with 2 mg/g carbomer gel (Viscotears; Alcon, Hemel Hempstead, UK) before placing onto the cornea (Fig 1). The entire 360° of the ICA was then examined using a handheld slit-lamp biomicroscope at 10× magnification (Keeler PSL Classic, Berkshire, UK) for the presence of PLD, which was quantified according to the percentage of the ICA circumference affected, estimating this to the nearest 5 per cent. Regions of ICA were judged to be affected by PLD where they exhibited abnormally broad pectinate ligament fibres or solid sheets of tissue with or without flow holes as previously described (Read and others 1998; Pearl and others 2015). No attempt to measure ICA width was made. The same examiner, a diplomate of the European College of Veterinary Ophthalmologists (ECVOs) (JACO) performed all the examinations to reduce the influence of subjectivity.
Initially, the mean percentages of ICA affected by PLD for the left and right eyes were averaged for each dog, and the averaged data were converted into ordinal grades (grades 0, 1, 2, 3) as previously described (Table 1) (Pearl and others 2015). Eyes were classified as ‘unaffected’ if 0 per cent of the ICA was affected by PLD (ordinal grade 0), ‘mildly affected’ if less than 20 per cent was affected (grade 1), ‘moderately affected’ if 20–90 per cent was affected (grade 2) and ‘severely affected’ if more than 90 per cent was affected (grade 3). Full details of how the scoring system was implemented can be found elsewhere (Pearl and others 2015; Oliver and others 2016a, b). PLD prevalence was determined by calculating proportions of dogs graded as 0, 1, 2 and 3 for each breed. The linear relationship between the ordinal variable PLD (grades 0, 1, 2, 3) and age was evaluated using Spearman's correlation coefficient (rho). Before that analysis, the continuous variable age was also converted into ordinal variables with four ranks based on quartile values. The proportions of male and female dogs affected with PLD (grades 1–3) within each breed and among the three breeds were compared using the χ2 test of independence. Mean ages of dogs in the three breeds were compared using one-way analysis of variance test. A P value of <0.05 (two-sided) was considered statistically significant. All analyses were performed using IBM SPSS statistical software (version 24).
Of the 102 BCs examined, 49 (48 per cent) were males and 53 (52 per cent) were females. The mean age (±sd) of BC was 58.98 months (±46.36, min=2.0 months, max=172.8 months). Eighteen of 102 (17.6 per cent) of BCs were PLD affected (ordinal grades 1–3); seven (6.9 per cent) mildly affected (grade 1), seven (6.9 per cent) moderately affected (grade 2) and four (3.9 per cent) severely affected (grade 3) (Table 2). No significant correlation was observed between PLD and age (rho=−0.015, P=0.879). The proportion of male and female dogs affected by PLD (grades 1–3) was not significantly different (P=0.169).
Of the 112 HVs examined, 35 (31.3 per cent) were males and 77 (68.8 per cent) were females. The mean age (±sd) of HV dogs examined was 55.58 months (±37.68, min=3.1 months, max=145.2 months). Thirty-eight of 112 (33.9 per cent) HVs were affected by PLD (ordinal grades 1–3); 22 (19.6 per cent) being mildly affected (grade 1), 14 (12.5 per cent) moderately affected (grade 2) and 2 (1.8 per cent) severely affected (grade 3) (Table 2). A weak but significant positive correlation was observed between PLD and age (rho=0.366, P<0.01). The proportion of dogs affected by PLD (grades 1–3) was not significantly different between male and female dogs (P=0.957).
Of the 230 GRs examined, 93 (40.4 per cent) were males and 137 (59.6 per cent) were females. The mean age (±sd) of GR examined was 53.31 months (±38.57, min=4.8 months, max=173.8 months). One hundred and eleven of 230 (48.3 per cent) GRs were affected by PLD; 51 (22.2 per cent) mildly affected (grade 1), 51 (22.2 per cent) moderately affected (grade 2) and 9 (3.9 per cent) severely affected (grade 3) (Table 2). A weak but significant positive linear relationship was observed between PLD and age (rho=0.173, P=0.009). The proportion of dogs affected by PLD (grades 1–3) was not significantly different between male and female dogs (P=0.114).
The prevalence of PLD (grades 1–3) was significantly higher in HV (38/112) compared with BC (18/102, P=0.007), higher in GR (111/230) compared with BC (18/102, P<0.001) and higher in GR (111/230) compared with HV (38/112, P=0.011). Age distribution among the three breeds was not significantly different (P=0.494, Table 3). Distribution of male and female dogs among the three breeds was significantly different (P=0.042, Table 3). The proportion of males and females was significantly different between BC and HV dogs (P=0.012); number of males was higher in BC (49/102, 48 per cent) compared with HV (35/112, 31 per cent) but also number of females was higher in HV (77/112, 68.8 per cent) compared with BC (53/102, 52 per cent). The proportion of males and females was not significantly different between BC and GR dogs (P=0.196) or HV and GR dogs (P=0.100).
In this study, the authors estimated the prevalence of PLD in populations of BC, HV and GR resident within the UK. The authors have recently provided prevalence data for four other breeds listed as either certifiable or under investigation for PLD under the Eye Scheme using the same gonioscopy grading scheme (Oliver and others 2016a, b). Under the Eye Scheme, dogs are considered to be ‘affected’ if 20 per cent or more of the entire ICA circumference is affected by PLD (Pearl and others 2015). Thus, those dogs classified as ordinal grades 2 or 3 (‘moderately’ or ‘severely’ affected, respectively) in authors’ current and previous studies would be classified as ‘affected’ under the Eye Scheme (Table 4). Such a classification is employed to allow those dogs with more severe PLD to be identified and eliminated from breeding populations while allowing those dogs with more minor abnormalities to continue to be bred from without restricting the gene pool. Dispute occurs, however, as to what constitutes a ‘normal’ level of PLD with different investigators suggesting figures between 6.25 and 25 per cent of the circumference of the ICA being affected by PLD may constitute normal, or at least acceptable, levels of variation (Ekesten and Narfström 1992; Read and others 1998; Pearl and others 2015; Fricker and others 2016). In the current study, the authors report 13.8 per cent BCs, 14.3 per cent HVs and 26.1 per cent GRs to be ‘affected’ by PLD under the guidelines of the Eye Scheme. Prevalences in the BC and HV populations studied were significantly lower than in the GR and lower than previously reported rates in the Basset Hound, Dandie Dinmont Terrier, Flatcoated Retriever and Welsh Springer Spaniel (Oliver and others 2016a, b) (Table 4). The prevalence in the GR, however, was more in line with the other breeds the authors previously investigated and very similar to that reported in the English Springer Spaniel in which PLD prevalence was found to be 25.5 per cent (Bjerkås and others 2002). A previous study reported a much higher PLD prevalence in GR in Switzerland with 56.5 per cent dogs affected, which would have to be excluded from a breeding programme according to the guidelines of the ECVOs (Spiess and others 2014). Although PLD prevalence may be genuinely higher in Swiss than in UK GR, it is important to note, however, that the ECVO scheme differs from the BVA/KC/ISDS Eye Scheme and so direct comparison of the results of the two studies is not possible. Under the ECVO scheme, an eye with any degree of pectinate ligament ‘sheeting’ (also known as laminae) would be classed as affected rather than only eyes with at least 20 per cent which might, in part, explain the greater apparent PLD prevalence in Swiss GR.
The three breeds the authors investigated are currently listed on Schedule B of the BVA/KC/ISDS Eye Scheme, being ‘under investigation’ for the condition. It is hoped that the information the authors publish here will assist the BVA/KC/ISDS Eye Scheme in deciding whether or not these breeds should be certifiable for PLD under the Scheme. Any decision to make PLD certifiable should be as evidence based as possible and should be guided by the knowledge of PLD prevalence in the wider canine population. In their study in Flatcoated Retrievers, Wood and others (1998) estimated the degree of PLD in 100 dogs of 30 different breeds as a control sample. In this study, 11 per cent of dogs were classified as ‘affected’ with at least 25 per cent of the ICA demonstrating PLD and this figure may serve as a useful benchmark in making such decisions.
In the current study, the authors found significant positive correlations between PLD and age in the HV and GR. This was expected and was in line with the authors’ previous findings in the Basset Hound, Dandie Dinmont Terrier, Flatcoated Retriever and Welsh Springer Spaniel and the findings of others in the Leonberger and provides further support of PLD progression (Pearl and others 2015; Fricker and others 2016; Oliver and others 2016a, b). The authors did not, however, find a correlation between age and PLD in the BC. This may be because either PLD is not progressive in this breed or, alternatively and more likely, the authors’ sample size was too small to detect a correlation in this breed, which appears to have a relatively low PLD prevalence. PLD was not associated with sex in any of the breeds investigated in this study. This was again expected and in line with the findings of similar studies in various breeds (Ekesten and Narfström 1991; Wood and others 1998; Bjerkås and others 2002; Oliver and others 2016a, b).
In conclusion, the authors have provided prevalence data for PLD in the BC, HV and GR. PLD prevalence in the BC and HV was relatively low and, as such, it is the opinion of the authors that PLD continues to be under investigation in these breeds. PLD prevalence in the GR was significantly higher, being similar to that reported in other breeds currently certified as affected for PLD under the BVA/KC/ISDS Eye Scheme. It is the authors’ opinion that for this breed, and after full discussion with all relevant stakeholders, that PLD should be considered to be certifiable under the Eye Scheme.
Provenance: Not commissioned; externally peer reviewed
Funding This study was funded by Dogs Trust as part of funding for a PhD project entitled ‘Pectinate ligament dysplasia and primary glaucoma in dogs: investigating prevalence and identifying genetic risk factors’.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.