This study presents epidemiologic data on atlas fractures over an observation period of 10 years.
It was found that the majority of patients with atlas fractures are male and that most cases occurred in the elderly population with a mean age of 72 years. This is in line with findings of previous epidemiologic studies on this topic that confirm a higher incidence of C1 fractures among the elderly [5, 11].
The main cause of trauma was a low-energy trauma which typically occurs in the older population where frequent falls combined with degenerative changes and reduced mobility of the lower cervical spine increase the risk of upper cervical fractures.
Only 33% of all atlas fractures were isolated injuries. About two-thirds of the cases were diagnosed with one or more concomitant head and spine injuries, and especially combined atlanto-axial fractures were seen in more than half of the patients. Previous studies have shown that atlas fractures rarely occur in isolation but often in association with other injuries of the spine [12, 13]. The combination of C1 and C2 injuries has been reported for 40–44% of the patients with atlas fractures in the literature [1, 9]. This was confirmed by findings of the present study, even though the incidence of concomitant injuries was noticeably higher compared to previously published results. Possible explanations include that CT imaging increasingly has become a standard diagnostic procedure for injuries of the cervical spine. The high percentage of concomitant intracranial bleedings may also be a result of the increased prevalence of oral anticoagulants in elderly patients. In fact, the high co-prevalence of occiput and C2 fractures and intracranial bleedings found in this series of cohort with atlas fractures strongly suggest a low threshold for performing CT imaging in these cases.
The distribution of fracture patterns according to Gehweiler showed that the majority of the injuries are type 3 fractures, followed by type 1 and 2, and then less frequently type 4 articular fractures. The transverse process fractures (Gehweiler 5) were found to be a rarity, which may be rooted in the rather subtle clinical symptoms they cause. This distribution of fracture patterns is consistent with data from previous studies [14, 15].
While isolated atlas fractures of type 1, 2 and 5 were preferably managed non-operatively, type 3 fractures and type 4 fractures were more often treated surgically in our study. This is in congruence with current treatment guidelines [16]. Despite inconsistent recommendations in the literature, fractures Gehweiler type 1, 2 and 5 are consistently being treated with conservative therapy.
Likewise, most studies report similar rates of surgical treatment of unstable type 3 (“Jefferson”) fractures [9, 14]. In contrast to the literature, however, almost every fourth Gehweiler 4 fracture in our population was treated operatively. The literature mainly recommends conservative treatment for isolated fractures of the lateral masses [10]. However, recent studies observed frequently a lateral displacement of the lateral mass fragments associated with subluxation of the occipital condyle resulting in neck pain, head malposition and impaired head rotation [17]. Hence, some guidelines advocate for surgery in Gehweiler 4 fractures with joint incongruency [9, 18]. In this context, it must be noted that the decision for and the type of surgery was frequently made based on concomitant injuries of the adjacent cervical spine segments.
The in-hospital complications observed in this study were in the range of what is known from other case series [1, 10]. Nosocomial infections as well as the development of a delirium must be considered as high-risk complications after surgery of elderly patients. In our cohort, patients who were treated non-operatively had a shorter hospital stay but were more likely to die during the hospitalization. Hence, as in other studies, the impact of a surgical intervention on the occurrence of complications was minor, as most part in-hospital complications were determined by the prevalence of preexisting comorbidities [13, 19].
The limitations of this study are associated with the retrospective design and the lack of follow-up examinations. It is difficult to assess post hoc what influence the patients’ illness had in some cases on the decision for or against surgical treatment. It might be that patients with estimated higher perioperative morbidity were frequently treated non-operatively even though the fracture pattern itself would have suggested surgery. An indication for this is the higher in-hospital mortality in the non-operative group. In fact, some patients might have died before surgery was possible.
We also did not perform a radiographic assessment of fracture union rates. The lack of complications or nonunions does not necessarily mean that these patients had good functional outcomes. The definition of “low-energy trauma” used in this retrospective analysis is very broad. Especially in elderly patients, it can often be difficult to precisely assess how deep and with how much energy the patient fell. This is a problem also known from thoraco-lumbar fractures in elderly patients.
Even though a large sample in total, the number of patients in each fracture pattern group did not allow statistic comparisons of operative versus non-operative treatment across single fracture types. Thus, the findings of this study cannot be used to support operative versus non-operative treatment for specific fracture patterns of the atlas. Future prospective comparative studies need to further investigate the potential benefit of operative versus non-operative treatment of atlas fractures in elderly patients.