Occipitoatlantoaxial malformations in equids

The case report by Mikic et al. (2025) described in this issue of Equine Veterinary Education details the clinical findings and imaging results of a 3-year-old Poitou donkey with occipitoatlantoaxial malformation (OAAM). The donkey was presented with limited neck mobility and a mild to moderate degree of proprioceptive ataxia. Advanced imaging, using computed tomography (CT), identified an asymmetric malformation of the occiput and C1 with extradural cervical static stenosis. There is a genetic test for OAAM currently available in Arabian horses (Bordbari et al., 2017); however, this donkey was not genotyped for the OAAM1 mutation.

While the term OAAM has been used to describe various malformations of the occiput, C1 and C2, the term was originally coined to specifically define symmetric familial occipitalisation of the atlas with atlantisation of the axis in Arabian horses (Mayhew et al., 1978). These malformations were characterised by an anterior developmental ‘shift’, whereby C1 resembled the occiput and C2 resembled C1. Clinically, these foals would either be dead at birth, demonstrate tetraparesis at birth, or develop a progressive ataxia as foals, with upper motor neuron signs and general proprioceptive deficits referable to a cervical lesion. Foals would often demonstrate an extended neck posture and an audible ‘click’ when the neck moved. At necropsy, these foals had symmetric malformations of the occiput, C1 and C2, with symmetrical compression of the cervical spinal cord. Pedigree analysis revealed common ancestry among these Arabian foals (Mayhew et al., 1978; Watson & Mayhew, 1986).

In 2017, we identified a genetic deletion in a highly conserved region of the genome in an Arabian foal affected with this ‘classic’ type of Arabian OAAM (Bordbari et al., 2017). The 2.7 kb deletion is located between two homeobox genes, HOXD4 and HOXD3. Homeobox genes direct the formation of body structures during early embryonic development. HOXD3 encodes for the formation of the occiput, atlas and axis, and HOXD4 encodes for the formation of the cervical vertebrae 1–3 and parts of the sternum. Mice that have had Hoxd3 deleted (Hoxd3^−/−) display a very similar phenotype to familial Arabian horse OAAM (Condie & Capecchi, 1993). Therefore, the deletion that was identified with familial Arabian OAAM (OAAM1) likely removes a conserved regulatory element that controls HOXD3 and perhaps even HOXD4. While this OAAM1 mutation defined the clinical disease in one case and additional carriers have been identified, it does not explain all cases of familial Arabian OAAM. It is likely that additional genetic mutations, particularly in the HOXD3/4 region of the genome, may be present in other cases of familial OAAM in Arabian horses.

In addition to the familial symmetric Arabian horse OAAM, there have been many other variations on this phenotype described. Many of these phenotypic variations occurred in Arabian or half-Arabian horses and include duplications of the atlas ± the axis (de Lahunta et al., 1989; Watson & Mayhew, 1986) and one case with subluxation of the atlantooccipital joint with fusion of the atlas and axis, lateral deviation of the atlantoaxial joint and 20° rotation of the atlas (Blikslager et al., 1991). Asymmetrical atlantooccipital fusion has been reported where the entire atlas was rotated on its vertical axis with no evidence of spinal cord constriction in an unknown breed (Mayhew et al., 1978).

Congenital symmetric OAAM have also been described in non-Arabian breeds, including Quarter Horse and related breeds (Wilson et al., 1985), Friesians (Bell et al., 2007) and Miniature horses (Rosenstein et al., 2000). Congenital asymmetric OAAM in non-Arabian breeds occurred in a Standardbred and a Morgan, both of which were presented with cervical scoliosis and head deviation, with no evidence of spinal cord disease (Mayhew et al., 1978). These horses had fusion between the occiput and atlas with missing occipital condyles. The scoliosis observed in the mid-cervical region was due to a wedge-shaped piece of vertebrae fused to the caudal edge of the axis. A similar malformation was subsequently reported in a Miniature horse foal; however, there was no scoliosis (Rosenstein et al., 2000). Based on these classifications, the donkey described in this case report would be classified as a congenital asymmetric OAAM case. This malformation is likely congenital but nonfamilial based on the information provided in the case report.

Many of the case reports in Arabian and non-Arabian breeds cite recumbency as foals or an early-onset progressive ataxia and/or weakness associated with the malformations. In the case report described here, the donkey was 3 years old at the time of presentation and had a 4-week history of spinal ataxia. This is a later onset than what is typically observed in published cases. However, we have identified a family of Arabian horses with C2/C3 malformations that result in cervical stiffness with only very mild ataxia, and these cases were not identified until the horses were adults. Thus, equids presenting at older ages with proximal neck stiffness ± ataxia should be evaluated for congenital OAAM as signs may be apparent until later in life.

There are no funders to report for this submission.

No conflicts of interest have been declared.

Not required for this clinical commentary.