Geriatric Odontoid Fractures: Treatment Algorithms of the German Society for Orthopaedics and Trauma Based on Expert Consensus and a Systematic Review

Abstract

Study Design

Systematic review/expert consensus.

Objectives

Fractures of the axis represent the most frequent injury of the spine in elderly patients. Both, operative and non-operative treatment are associated with a high rate of complications and mortality. The aim of this article was to summarize the current literature on the management of odontoid fractures in geriatric patients and to weigh it based on an expert consensus process.

Methods

In a joint consensus process, members of the Spine Section of the German Orthopaedic and Trauma Society (DGOU) aimed to formulate recommendations for the diagnostic workup and treatment of odontoid fractures in geriatric patients. Based on the previously published recommendations, this article is an updated version with incorporating a systematic review of the recent literature.

Results

Based on the new data available, the recommendations established in the initial consensus process were adapted.

Conclusions

Computed tomography represents the diagnostic standard for patients with suspected injuries of the upper cervical spine. Anderson/D’Alonzo odontoid fractures type 1, non-displaced type 2, and type 3 can be treated conservatively. Even non-unions do not necessarily result in poor clinical outcome. In Anderson/D’Alonzo type 2 fractures, surgical therapy offers the advantage of relatively safe osseous healing with no increased complication rate even in elderly patients and can thus be recommended. In very high aged patients, however, a case-by-case decision should be made. When surgical stabilization of osteoporotic odontoid fractures is indicated, posterior techniques are biomechanically advantageous and can be considered the standard.

Introduction

Fractures of the dens axis are the most common type of injury of the cervical spine in patients over 65 years of age.^1,2 In view of the demographic development of the last decades, a significant incidence increase has been observed. ³ Especially in geriatric patients, fractures of the dens axis present a particular challenge. Usually, patients present with poor bone quality based on characteristic local degenerative changes. Due to unfavorable biomechanics, small fracture surface, and poor blood supply to the dens base, odontoid fractures poorly heal in elderly patients.

One-year mortality rates of 18%-45% have been described for odontoid fractures in geriatric populations.^4-6 The most important factor for the increased mortality is the high number of comorbidities in elderly patients, evenly increasing the risk of perioperative complications. On the other hand, non-union rates of up to 80% have been reported with conservative therapy.^7-9 There has been ongoing debate whether to treat these fractures operatively or non-operatively. ¹⁰

This paper summarizes the current state of the literature on odontoid fractures in the elderly patients and provides recommendations for their management based on expert consensus and a systematic review.

Methods

Consensus Process

A joint consensus process involving the members of the Spine Section of the German Society for Orthopaedics and Trauma Surgery (DGOU) had been performed resulting in recommendations on the diagnosis and treatment of odontoid fractures published previously in 2020. ¹¹ Based on these previously published recommendations, this article is an updated version with systematically incorporating the recent literature.

For this, a systematic review of the literature was conducted according to the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-analyses) checklist and algorithm, ⁶ with the goal of finding all articles describing the diagnostic workup and treatment odontoid fractures during the last 5 years.

Search Strategy

Investigations between 08/2017 and 08/2022 were included. Prospective and retrospective observational investigations were considered for analysis. The authors performed a systematic an initial search of PubMed/Medline databases for eligible investigations. The search terms were (odontoid) AND (fracture*). Studies were selected according the following inclusion criteria: a) case series, cohort studies, clinical trials, or registry studies regarding the epidemiology, the diagnostic workup or the treatment of odontoid fractures in patients older than 60 years, b) articles published in English or German language. Exclusion criteria were: a) cadaver studies, biomechanical studies, anatomical studies, systematic or narrative reviews, meta-analyses, case reports, uncontrolled case series, case series reporting only descriptive outcome data of a single treatment arm, case series with fewer than 100 patients or fewer than 20 patients in one treatment arm, studies reporting on only radiographic measures, recommendations and guidelines, surveys, and technical notes, b) studies investigating atypical fractures of C2, combined fractures of C1 and C2, non-unions, or metastatic fractures.

Titles and abstracts were reviewed. Duplicates were removed and full texts were checked for suitability. The remaining full-length articles were then screened by all authors and disagreement was resolved by consensus. The final decision was made based on the analysis of the full text. In doubt, articles were included into the next stage. A flowchart of the filtering stages (titles, abstracts, full-length texts) is shown in Figure 1.

Figure 1.

Literature research flow chart.

Synthesis of Results

We extracted data concerning study characteristics including authors’ names, title, year of publication, journal of publication, number of patients, time of follow-up and type of study. For the description of the study population, the number of patients and their age were collected. Outcome parameters were analyzed according to the inclusion criteria. For all included studies we used the Oxford Centre for Evidence-Based Medicine 2011 for defining the level of evidence. ⁷

Results

Systematic Review

Twelve articles met the inclusion criteria (Figure 1), among these 5 were retrospective case series and seven were studies using data from national databases (Table 1). Publication dates ranged from 2017 to 2022 and data from totally 48,316 patients with an odontoid fracture were reported. Nine studies compared some operative vs non-operative treatment, while three studies compared different operative techniques. The results of the analyzed studies were therefore incorporated into the following chapters on Operative versus non-operative management and Operative therapy.

Table 1.

Studies Included into the Systematic Review.

1st Author	Year	Design	N	Patient Age	Treatment	Main Outcome
Patterson	2017	Retrospective, national database	141	78	Operative: Anterior (34%) vs posterior (66%)	Anterior odontoid fracture stabilization associated with shorter operative times, greater risk of unplanned readmissions and revision operations within 30 days of surgery
Robinson	2018	Retrospective, national database	3375	83	Operative (22%) vs non-operative (78%)	Adjusted for age, gender, comorbidity, and year of injury, surgically treated patients had greater survival than non-surgically treated patients. Among those >88 years, surgical treatment lost its effect on survival
Bhimani	2018	Retrospective, national database	165	70	Operative: C1-C2 vs C0-cervical	Higher risk of reoperation for occiputo-cervical (9%) vs C1-C2 fixation (2%)
De Bonis	2019	Retrospective, multicenter	147	≥65	Operative (54%) vs non-operative (46%)	Modified Rankin scale and Charlson comorbidity Index are independent predictors for functional outcome and posttreatment disability. No significant differences between patients of 65 to 79 years and ≥80 years
Gembruch	2019	Retrospective, monocentric	125	86	Operative (78%) vs non-operative (22%)	Fracture union in 91% after surgery and 70% after conservative treatment. 3-month mortality 28% in the surgical group and 20% in the non-op group
Shousha	2019	Retrospective, monocentric	133	74 anterior 78 posterior	Operative: Anterior (35%) vs posterior (65%)	Less blood loss and shorter hospital stay with anterior fixation but higher rates for non-union and revision surgery compared to posterior fixation
Honda	2020	Retrospective, national database	3167	70	Operative (22%)2 vs non-operative (78%)	No significant difference in in-hospital mortality between conservative treatment (4.6%) vs surgery (3.0%), also in patients aged ≥80 years (7.2% vs 5.4%)
Lukins	2020	Retrospective, monocentric	141	≥65	Operative (28%) vs non-operative (72%)	Age, male gender, and injury severity are predictors of 12-month mortality, choice of treatment is not
Rizvi	2020	Retrospective, monocentric	282	77	Operative (68%) vs non-operative (32%)	Secondary surgery in 14.4% after initial non-op treatment and 9.5% after initial operative treatment. Higher fusion rates with surgery
Alluri	2021	Retrospective, national database	32,419	77	Operative (67%) vs non-operative (32%)	Operative treatment associated with higher odds of complications (pulmonary, gastrointestinal, renal). Inpatient mortality 3.6% for operative and 5.9% for nonoperative treatment, but more preexisting comorbidities in non-operative patients
von Glinski	2021	Retrospective, national database	2921	77 non-op 70 operative	Operative (19%) vs non-operative (81%)	Higher overall 90-day readmission rate of after nonoperative (29.4%) vs operative (16.4%) treatment
Smith	2022	Retrospective, national database	5300	77	Operative (8%) vs non-operative (92%)	Surgery was associated with decreased one-year mortality in all patients, patients between the ages of 65 and 74, and patients over the age of 75

Imaging

In elderly patients, odontoid fractures usually occur as a result of low-energy trauma. Patients often present after a fall from standing height including some kind of head impact and clinical sings of head or face injury. Patients report with neck pain and limited motion of the head and cervical spine. Neurologic deficits are rare but may be fatal, because of the localization at the level of the cervical medulla. Clinical examination is therefore an inaccurate measure for detecting odontoid fractures.

In accordance with the C-spine rules, radiologic imaging including CT-scan of the cervical spine is always recommended in patients over 65 years of age with traumatically induced neck pain or after head impact. ¹²

Conventional radiographs of the cervical spine in antero-posterior and lateral projection as well as a transoral odontoid radiographs continue to be part of basic diagnostics because of their availability, low cost, and low radiation exposure.

However, because computed tomography (CT) is clearly superior to conventional radiographs in the cervical spine in terms of sensitivity and specificity,^13,14 it is considered the gold standard, especially in cases of suspected upper cervical spine injury.

Magnetic resonance imaging (MRI) is used only in cases of neurologic deficits, suspected disco-ligamentous injury, or to differentiate between fresh and old fractures. ¹³

Native CT should be supplemented by CT/MRI angiography to exclude lesions of the vertebral artery and to detect an irregular course of the vessels preoperatively. This is recommended especially if CT shows fracture extensions into the vertebral canal or if surgical anterior or posterior C1/2 fusion is planned.^15,16

Angulation (>11°) and displacement (>5 mm) on CT scans are strong predictors for fracture instability. ¹⁷ Dynamic fluoroscopic examination can be performed in borderline cases or to detect disco-ligamentous C1/2 instability or to differentiate between stable or unstable non-union. The examination should be performed by a physician and with the patient awake in guided passive flexion and extension.

Classifications

Odontoid fractures are usually classified according to the system described by Anderson and D'Alonzo ¹⁸ (Figure 2). The classification distinguishes three morphological types: Type 1 is a fracture of the odontoid tip in terms of a bony avulsion of the ala ligaments, type 2 is a fracture in the odontoid base region, and type 3 is a fracture in the odontoid base region extending into the body of the axis.

Figure 2.

Classification of odontoid fractures by Anderson and D’Alonzo. ¹⁸

According to Eysel and Roosen, type 2 can be further divided into three subtypes (Figure 3): ¹⁹ type 2A has a horizontal fracture course, in type 2B the fracture proceeds from antero-superior to postero-inferior, and in type 2C the fracture proceeds from antero-inferior to postero-superior. The fracture course in type 2A and 2B allows anterior screw osteosynthesis, whereas this is not possible in type 2C. A nearly similar subclassification was described by Grauer et al in 2005. ²⁰

Figure 3.

Sub-Classification of type 2 fractures by Eysel and Roosen. ¹⁹

The Upper Cervical Spine Injury Classification System of the AOSpine summarizes odontoid fractures under type A (“bony injury only without significant ligamentous, tension band, discal injury”; www.aospine.org/classification).

Treatment

Operative Versus Non-Operative Management

The decision for or against a surgical procedure in odontoid fractures of elderly patients is influenced by several factors. In addition to fracture morphology and concomitant injuries in the cervical spine, the patient’s age and comorbidities must be considered.

There is consensus for conservative therapy for Anderson and D’Alonzo type 1 and type 3 fractures - if there is no gross dislocation or concomitant ligamentous instability.^21,22

In contrast, the treatment of type 2 fractures has been subject of controversial debates in the literature. Type 2 fractures are localized at the base of the dens, where the shear forces of the articulating atlas ring are particularly strong. Stress shielding by limited degenerative C1/2 motion of the dens base leads to a localized osteopenic situation. In combination with the weak blood supply to the dens base and a frequently existing anklyosis of the anterior arch of the atlas with the odontoid tip in older individuals, this leads to non-union rates of up to 80% under conservative therapy. ⁹ In contrast, very high osseous union rates were reported with surgical therapy.^23,24 Meta-analyses show significantly higher fracture healing in surgically treated patients, especially for fractures with a displacement of more than 4 mm.^22,25 However, odontoid non-unions do not necessarily lead to a worse functional outcome.^26,27 Cases of secondary myelopathy in odontoid non-union have mostly been described for younger and more active individuals but rarely occur in elderly patients.^28,29 A recent systematic review concluded that it is safe to forgo surgery for carefully selected patients with non-united odontoid fractures when near-anatomic alignment is present, dynamic instability is lacking on imaging studies, the neurologic examination findings are normal, and the risk of neck injury is low. ³⁰

The complication rates in geriatric patients under 85 years of age treated operatively or conservatively do not differ in terms of frequency - but they do in terms of complication patterns.^26,31 The overall 90-day readmission rate after non-operative treatment is almost twice as high as after operative treatment. ³² Surgically treated patients over 80 years of age suffer from an increased incidence of internal complications,^22,33 while 10% of patients who are treated conservatively with a cervical orthosis develop pressure ulcers.^34,35

In elderly patients, odontoid fractures are associated with high mortality regardless of the therapy chosen.^4-6,36,37 For non-operative treatment, 1-year mortalities of up to 45% have been described.^36,38 Similar mortality rates have been described for operative therapy, even in patients over 80 years of age.^33,39 A retrospective multicenter study of the AOSpine North America reported significantly lower mortality in surgically vs non-operatively treated patients (7% vs 22%). This difference was still evident at 2-year follow-up. ⁴ Improved 1-year survival after surgical management was also reported in a recent national database study analyzing data of 5300 patients with an odontoid fracture. ⁴⁰ In contrast a single center study reported an increased mortality with surgery, ⁴¹ However, systematic meta-analyses of numerous studies have so far failed to detect significant differences in mortality between conservatively and surgically treated elderly patients with odontoid fractures. ²² Most recent studies with large cohorts report that age, gender and comorbidities are the most relevant risk factors for mortality and that choice of treatment is not.^31,37,42,43 A recent retrospective national database analysis of 3375 patients reported that - when adjusted for age, gender, comorbidity, and year of injury - surgically treated patients had greater survival than non-surgically treated patients. ⁴⁴ Among those aged over 88 years, surgical treatment lost its effect on survival, though. ⁴⁴

No significant differences in mortality can be found between conservatively and surgically treated patients over 80 years of age.^22,45,46A problem with all studies on geriatric odontoid fractures, however, is the age-related high drop-out rate, which is likely to lead to a relevant selection bias, especially when comparing surgically vs non-operatively treated patients. Patients with more comorbidities are more likely to be treated non-operatively.^31,40

However, the data published so far indicate that the general frailty and the comorbidities of the patients responsible for it probably play a greater role in morbidity and mortality than age.^32,46,47 It can be assumed, though, that patients older than 88 years are very likely to be frail. Therefore, most authors agree that in very old patients, surgical therapy is only rarely indicated in selected cases.^32,47

Non-Operative Therapy

The non-surgical treatment of odontoid fractures in geriatric patients includes medicinal pain therapy according to WHO and immobilization in a semi-rigid cervical orthosis for six weeks. There is no advantage of wearing hard cervical orthoses in terms of achieving fracture union and these should be worn for a maximum of 24 hours as they otherwise may cause pressure ulcers.^35,48 Even with semi-rigid orthoses the fit should be checked regularly. In patients who are impaired or in need of care, the use of an orthosis should be questioned to avoid dysphagia and aspiration.

Especially in the beginning, clinical and radiological follow-up should be frequently performed over the first 6 weeks to detect secondary displacement. In case of doubt, a CT scan or a dynamic examination should be ordered at a low level.

Right from the beginning of conservative therapy, isometric exercises assisted by physiotherapists should be started. Active physiotherapeutic exercise for mobility should only be started after radiological confirmation of fracture healing or if there is no indication of secondary dislocation in the conventional radiographs 6 weeks after trauma.

Treatment with halo fixation is associated with significantly increased morbidity and mortality in geriatric patients and should be avoided, ⁴⁹ but may serve as an alternative in individual cases in patients with unstable, highly displaced odontoid fractures whose general condition does not permit general anesthesia. ⁵⁰

Operative Therapy

Biomechanical studies have shown that posterior instrumentation of C1/2 using a posterior internal fixator according to Goel/Harms is the most biomechanically stable method for treating odontoid fractures.^51,52 In addition, it allows for direct open reduction and true osseous fusion of the C1/2 segment. For this technique, screws are placed bilaterally into the lateral masses of C1 and either into the pedicles or the pars of C2 and connected with rods. Hence, it can be considered the work horse for most geriatric odontoid fractures (Figure 4).

Figure 4.

Recommended treatment algorithm.

An alternative of almost similar initial biomechanical strength is posterior transarticular screw fixation of C1/2 as described by Magerl in combination with a Gallie cerclage and bone graft fusion through an open approach. ⁵³ However, the true advantage of the posterior transarticular screw fixation is that it can be applied percutaneously, even though the stability achieved is not the same as with posterior Goel/Harms fixation. ⁵¹ It can also not be used in patients with a high-riding vertebral artery.^15,16 It is only in these cases that a posterior occipito-cervical fixation may be used as this techniques has a higher risk for reoperation compared to posterior C1/2 fixation. ⁵⁴

In patients with healthy bone and fractures according to Eysel/Roosen type 2A and 2B, anterior screw fixation of odontoid fractures represents a less-invasive alternative that – in contrast to posterior procedures – theoretically maintains mobility in the segment C1/2. In osteoporotic bone, stand-alone anterior Odontoid screw fixation has been shown to be associated with high failure and revision rates compared to posterior fixaton.^55,56 However, a combination of anterior odontoid screw fixation with an anterior transarticular screw fixation of C1/2 can achieve similar stability as posterior transarticular screw fixation. ⁵⁷ Therefore, this type of combined anterior odontoid/transarticular screw fixation appears to offer an alternative that is particularly suitable for patients in whom surgery in the prone position would be associated with increased perioperative risk.

Conclusion

Odontoid fractures in elderly patients are increasing in incidence and relevance. A high degree of suspicion is key and the threshold for advanced imaging diagnostics should be low. CT is the gold standard for primary diagnosis of suspected upper cervical spine injuries and should be combined with an angiography when operative treatment is considered.

Anderson/D'Alonzo type 1 and type 3 odontoid fractures and non-displaced type 2 fractures can be treated conservatively. Even non-unions do not necessarily result in poor clinical outcome. In Anderson/D'Alonzo type 2 fractures, surgical therapy offers the advantage of relatively safe osseous healing with no increased complication rate even in elderly patients and can thus be recommended. In very high aged patients, though, a case-by-case decision should be made.

When surgical stabilization of osteoporotic odontoid fractures is indicated, posterior procedures are biomechanically advantageous and can be considered the standard.

Anterior combined odontoid/transarticular screw fixation are a less invasive alternative when strictly indicated and technically well executed. However, anterior procedures are associated with an increased risk of dysphagia. ⁵⁸

ORCID iDs

Georg Osterhoff https://orcid.org/0000-0001-5051-0998

Alexander C. Disch https://orcid.org/0000-0001-8414-4041

Max J. Scheyerer https://orcid.org/0000-0003-1392-3990