Non-union of the odontoid process. An experimental investigation.

A clinical study of fractures of the odontoid process showed a 62 per cent failure on union. Of many features studied, only displacement and its direction had a definite bearing on non-union of the fracture. Blood supply was considered as a possible etiological factor. The blood supply of the human odontoid was elucidated by means of studying 19 human autopsies by means of microangiography. The dog was found to have a similar blood supply and hence, was used as the experimental model. Two osteotomies were performed; one below the accessory ligaments and one above the accessory ligaments. All osteotomies carried out below the accessory ligament united and all carried out above failed to unite. Microangiographic studies revealed however, that avascular necrosis was not the cause of non-union. Further anatomic studies revealed that different size gaps occurred depending on the level of the osteotomy, with a large gap occurring in the osteotomy of the odontoid which was performed above the accessory ligament. First the result of immobilization and the healing of the odontoid osteotomy was derived, by carrying out an instant occipito-cervical fusion by means of wire loops and methylmethacrylate. In distinction to the free floating apical segment of non-immobolized spines, with a high odontoid osteotomy, the occipito-cervical fusion resulted in a dense fibrous tissue stabilizing the apical fragment but at no time was bony union observed. A final attempt was to secure immobilization and reduce the gap, by carrying out only a partial osteotomy. The osteotomy was performed in such a way that the posterior cortex was left intact. Union occurred in all instances. The fracture gap and movement play a definite role in the pathogenesis of pseudarthrosis of the odontoid process. Where gap and movement were eliminated, union occurred. It is difficult to transpose this experimental situation to the clinical one, for the assessment of gap clinically, is only radiographic. This method is much too imprecise to be of value in this assessment. The fact however, that a high rate of non-union is associated with a high degree of displacement, supports the experimental thesis.