SAEVA Proceedings 2015 | Page 36

South African Equine Veterinary Association Congress 2015  Protea Hotel  Stellenbosch diagnosis at best (Papageorges et al. 1987). Oblique radiographs are helpful in certain circumstances (Withers et al. 2009). Objective assessment of vertebral canal diameter is more accurate than subjective evaluation of vertebral malformation for identifying young horses affected by CSM but may lead to false negative diagnoses in older horses (Levine et al. 2007). Both inter- and intra-vertebral measurements are used. The sensitivity and specificity of the intra-vertebral sagittal ratio method is approximately 90% for vertebral sites between the third and seventh cervical vertebrae (Moore et al. 1994). In most normal horses, the sagittal ratio exceeds 52% from the third to sixth cervical vertebrae and 56% at the seventh cervical vertebrae in horses greater than 320 kg. The positive predictive value of such measurements is probably higher, and the negative predictive value lower, in ataxic horses from countries where conflicting diagnoses (such as EPM) are not routinely encountered (i.e. false positives are less likely, but false negatives are more likely because the underlying prevalence of CSM in ataxic horses is higher). Similarly, the positive and negative predictive values of objective cervical radiography measurements in the absence of ataxia (for example during prepurchase radiography) have not been evaluated, but false positives are likely to be more, and false negatives, less common, since the prevalence of CSM in this population will be much lower. Some clinicians advocate use of ratiometric measurements that take into account the distance between adjacent vertebrae (inter-vertebral ratios) based on the rationale that most compressive lesions occur between, rather than within, the vertebrae (Hahn et al. 2008). Particularly high quality radiographs are usually required for such measurements, but analysis suggests that this approach may be helpful in differentiating CSM from other conditions (Van Biervliet 2007). Further comparison of both methods in a large group of horses is needed based on a gold standard diagnosis established at post mortem examination, since myelography is problematic (see discussion below), although available post mortem material may be skewed towards severely affected horses, since these animals may more often be euthanased. Plain radiography is often considered sufficient to make a presumptive diagnosis of cervical compression without the need for further tests. In countries where EPM or other conflicting differential diagnoses are possibilities, many clinicians favour myelography for diagnosis. Unfortunately, for most inter-vertebral sites, myelography results in a high number of false positive and false negative results (van Biervliet et al. 2004). Myelography remains however a prerequisite if surgical intervention is considered a viable option on the basis of severity of signs and the owner‟s wishes and expectations. This is because plain standing radiography does not definitively pinpoint the actual site of the compressive lesion(s) (Moore et al. 1994). Note that neck flexion and extension while under anaesthesia are contraindicated if there is evidence for compression on the initial neutral views. Ventrodorsal projections may be attempted in small or young animals, especially in the cranial neck, and may demonstrate an asymmetric compressive lesion that might otherwise account for some false negative diagnoses in larger horses. A variety of techniques have been used for interpreting equine myelograms. It is common for the ventral dye column to be lost, but for there to be no or minimal change to the width of the dorsal dye column in normal horses; consequently, often at inter-vertebral sites between C2 and C6, diagnosis of compression is made on the basis of a 50% or greater decrease in the sagittal width of the dorsal and ventral contrast columns in comparison with the column width at the immediate cranial or caudal mid-vertebral site. At C6-C7, a reduction of more than 20% of the dural diameter measured in the mid body region is best used to diagnose compression, since this measurement has a relatively high sensitivity and specificity. Clinicians may favour use of different “cut off” values for exclusion or inclusion of diagnosis based on the consequence of the derived decision (i.e. possible euthanasia or surgery). In a similar manner to interpreting plain radiographs, the positive predictive value for myelography is likely higher in ataxic horses from 36