Sage Journals: Discover world-class research

Abstract

Study Design

Single-center retrospective cohort study

Objectives

Type II odontoid fractures occur disproportionately among elderly populations and cause significant morbidity and mortality. It is a matter of debate whether these injuries are best managed surgically or conservatively. Our goal was to identify how treatment modalities and patient characteristics correlated with functional outcome and mortality.

Methods

We identified adult patients (>60 years) with traumatic type II odontoid fractures. We used multivariate regression controlling for patient demographics, Glasgow Coma Scale (GCS) score, Charlson Comorbidity Index (CCI), modified Rankin Scale (mRS) score, modified Frailty Index (mFI-5 and mFI-11), fracture displacement, and conservative vs operative treatment.

Results

Of the 59 patients (mean age 77.9 years), 24 underwent surgical intervention and 35 underwent conservative management. Operatively managed patients were younger (73.4 vs 80.6 years, P < .001) and had higher degree of fracture displacement (3.5 vs 1.0 mm, P = .002) than conservatively managed patients but no other differences in baseline characteristics. Twenty-four patients (40.7%) died within the study period (median time to death: 376 days). There were no differences between treatment groups in functional outcomes (mRS or Frankel Grade) or mortality (33.3% in operative group vs 45.7%, P = .34). There was a statistically significant correlation between higher presentation mRS score and subsequent mortality on multivariate analysis (OR = 2.06, 95% CI 1.04-4.10, P = .039), whereas surgical intervention, age, GCS score, CCI, mFI-5, mFI-11, sex, and fracture displacement were not significantly correlated.

Conclusions

Mortality after type II odontoid fractures in elderly patients is common. mRS score at presentation may help predict mortality more accurately than other patient factors.

Keywords

odontoid mortality elderly frailty modified Rankin Scale score type II odontoid fracture

Introduction

The incidence of cervical fractures is biphasic in distribution, most commonly affecting individuals between 20 and 30 or 70 and 80 years old.^1-3 In the younger population, high-energy trauma is the most common cause, whereas the elderly are susceptible to low-energy trauma due to age-related changes in bone density. Of all cervical spine injuries in older patients, C2 fractures are the most common, with type II odontoid fractures of the Anderson D’Alonzo Classification⁴ representing the most prevalent subtype.^1,5,6

Type II odontoid fractures occur at the base of the odontoid process and are inherently unstable. These injuries disproportionately affect the elderly population, who are at higher risk of subsequent morbidity and mortality.⁷ Globally, there is an expected increasing life expectancy over the next decades with the population of individuals over 80 years old in North America alone anticipated to double by 2050.⁸ As these numbers rise, the incidence of these fractures will increase. Thus, developing an appropriate treatment algorithm is essential. Despite the relatively common nature of these injuries, there continues to be debate as to whether Type II odontoid fractures are best managed surgically or conservatively in this elderly population.

Expected functional status and mortality are vital considerations when contemplating treatment options and selecting surgical candidates. Various scores such as the Charlson Comorbidity Index (CCI),⁹ modified Rankin Scale (mRS) score,¹⁰ and modified Frailty Index (mFI-5 and mFI-11)^11,12 have been developed to stratify a patient’s condition, aid in determining a patient’s candidacy for surgery, and predict outcome. We, therefore, performed a retrospective review that included patient demographics, treatment modality, and multiple patient comorbidity and frailty indices, which were then controlled for in a multivariate model to identify factors associated with functional outcome and mortality.

Methods

Study Design and Patient Population of Interest

After receiving institutional review board approval with a waiver of informed consent (IRB#115230), we performed a retrospective single-institution cohort study comprising tertiary-care hospital record data from August 2015 to July 2020. Health records, including death reports, were available for at least 2 years after presentation for each patient. Clinical and radiographic follow-up ranged from 2 days (in the case of in-hospital death) to 4.3 years. Patients included for analysis were over the age of 60 years with traumatic type II odontoid fractures as defined by the Anderson and D’Alonzo classification method.⁴ Variables of interest relating to fracture presentation included patient age at presentation, mechanism of injury, sex, body mass index (BMI), fracture displacement (category and distance of displacement), operative intervention related to odontoid fracture, and neurologic deficit at presentation. Fifty-nine patients met criteria for inclusion and were categorized into 2 groups based on operative or nonoperative management. Twenty-four patients were treated operatively while 35 were treated with conservative measures. Patients who were initially treated nonoperatively but went on to surgery were evaluated in the “conservatively managed” group throughout the study. The STROBE guidelines were used for reporting of this study.

Functional Status and Comorbidity/Frailty Variables of Interest

Functional status and comorbidity variables of interest included presenting mRS score, CCI, Glasgow Coma Scale (GCS) score, and mFI-5 and -11. The CCI is a validated measure of patient comorbidity⁹ that has been used previously to predict mortality after odontoid fractures in the elderly.^13-15 The mFI-5¹¹ and mFI-11¹⁶ are validated frailty indices developed from the American College of Surgeons National Surgical Quality Improvement Program database to predict mortality after surgical procedures; these have been used previously to assess frailty in elderly patients with odontoid fracture.^17,18

Outcomes of Interest

Outcomes of interest were obtained at different follow-up intervals based on available hospital documentation. Mortality was determined for all patients >2 years after initial presentation as summarized by hospital and death records and did not rely on neurosurgical follow-up. Clinical outcomes were obtained from neurosurgery clinic notes. Patients with imaging obtained more than 30 days postoperatively were assessed for radiographic fracture stability relative to the initial imaging, which was defined as no new odontoid displacement or abnormal motion on radiographic imaging (Figures 1 and 2). Fourteen patients did not have repeat imaging more than 30 days after presentation and were therefore excluded from the fracture stability analysis (median imaging follow-up duration 116 days, IQR 249). Radiographic findings of stability were corroborated with the surgeon’s imaging interpretation recorded in final clinic follow-up notes. Other outcomes of interest included related readmission, discharge disposition, Clavien Dindo score¹⁹ at most recent clinic follow-up for patients who underwent surgical intervention, need for intensive care unit (ICU) stay, length of stay, Frankel grade²⁰ at last clinic follow-up, and post-intervention chronic pain and chronic opioid use.

Figure 1.

Lateral cervical x-rays from an 84-year-old woman who presented after a ground-level fall and was found to have a displaced, unstable type II odontoid fracture. She was managed in a collar and was determined to have achieved a stable nonunion at final radiographic follow-up. Left panel: lateral cervical x-ray obtained on presentation. Right panel: x-rays obtained after 12 weeks in a cervical collar.

Figure 2.

Lateral cervical x-rays from an 82-year-old woman who presented after a ground-level fall and was found to have a displaced type II odontoid fracture. She was treated with a C1-C3 posterior spinal fusion and achieved a stable bony fusion at final radiographic follow-up. Left panel: preoperative lateral cervical x-ray. Right panel: postoperative lateral cervical x-ray demonstrating bridging bone within previous fracture site, suggesting fusion.

Statistical Analysis

Descriptive statistics were performed for all baseline patient characteristic variables, with categorical variables expressed as n (%) and continuous variables expressed as mean ± standard deviation (SD). A univariate analysis was performed to compare patients undergoing operative vs nonoperative fracture management. Patients <75 and ≥75 years of age were also compared using a univariate analysis. In both univariate analyses, Fisher’s exact test was used for categorical variables, and Student’s t-test was used for continuous variables. Variables that were independently associated with mortality at last clinical follow-up were identified with a multivariate binary logistic regression analysis. A receiver operating characteristic (ROC) curve was generated to assess model predictive power with calculation of the corresponding area under the curve (AUC). Alpha for significance was set to P < .05. All analyses were performed using SPSS v. 27 (IBM, Armonk, NY).

Results

Fifty-nine patients were included for analysis (Table 1). Patient age at presentation was normally distributed, with an average age of 77.9 years. The majority of patients were female (59.3%). The most common mechanism of injury was a ground-level fall (91.5%), with all other accidents each contributing 1.7%. Most patients had non-displaced fractures (54.2%), with a mean fracture displacement for the entire cohort of 2.04 mm (SD ± 2.94). Six patients (10.2%) had a neurologic deficit at presentation. Three patients with neurologic deficit underwent operative intervention. Two of these patients presented with bilateral weakness of the upper extremities. After posterior cervical fusion, one experienced resolution of weakness and the other had some residual weakness in the left arm. The third operatively managed patient presented with right arm weakness that improved after surgery. Among the three patients managed conservatively, one presented with bilateral upper-extremity weakness and one with right upper-extremity numbness; both patients experienced resolution by final clinic follow-up. The third patient in this conservatively managed group had right arm weakness and numbness that improved but did not resolve by final clinic follow-up. Twenty-four patients (40.7%) underwent operative fracture fixation.

Table 1.

Demographic and Comorbidity Variables for the Whole Cohort.

Variable	Value
Total patients	59
Age (y)	77.9 ± 8.4
Sex
Male	24 (40.7%)
Female	35 (59.3%)
BMI	26.7 ± 5.4
Mechanism of injury
Fall	54 (91.5%)
ATV accident	1 (1.7%)
Vehicle vs bike	1 (1.7%)
Horse kick	1 (1.7%)
Idiopathic	1 (1.7%)
Other	1 (1.7%)
Fracture displacement
Displaced fracture	27 (45.8%)
Non-displaced fracture	32 (54.2%)
A-P fracture displacement (mm)	2.04 ± 2.94
Fractures treated operatively	24 (40.7%)
mRS score at presentation	2.05 ± 1.31
Charlson Comorbidity Index	3.34 ± 3.15
GCS score at presentation	14.6 ± 1.6
Neurologic deficit at presentation	6 (10.2%)
mFI-5 at presentation	1.75 ± 1.23
mFI-11 at presentation	2.78 ± 2.04

Value reported as mean±SD or no. (%).

BMI, body mass index; ATV, all-terrain vehicle; A-P, anterior-posterior; mRS, modified Rankin Scale; GCS, Glasgow Coma Scale, MFI, modified frailty index.

Table 2 presents outcome variables for the entire cohort. The median clinic follow-up duration was 175 days with an interquartile range (IQR) of 355. The mean length of stay was 3.8 days (SD ± 3.9). Most patients were discharged home (52%), followed in frequency by assisted living (22%) and skilled nursing facility (10.2%). Forty-five patients had radiographic follow-up more than 30 days postoperatively (76.3%). Of these, 41 (91.1%) demonstrated evidence of stability, including 3 with confirmed stable nonunion. Seventeen patients were assessed for radiographic evidence of bony fusion at their last follow-up appointment, 10 (58.8%) of whom demonstrated bridging trabecular bone between the fracture endplates, indicating fusion. There were 7 patients with confirmed nonunion, 4 of which were characterized as unstable. The median time from presentation to final imaging was 116 days (IQR = 376).

Table 2.

Outcome Variables for the Entire Cohort.

Variable	Value
Length of stay (d)	3.8 ± 3.9
Discharge disposition
Home	31 (52.5%)
Acute rehabilitation	2 (3.4%)
SNF	6 (10.2%)
Assisted living	13 (22.0%)
Deceased/hospice	5 (8.5%)
NA	2 (3.4%)
Stable fracture at last follow-up^a	41 (91.1%)
Evidence of fusion^b	10 (58.8%)
Related readmission	3 (5.1%)
Reoperation	3 (5.1%)
Mortality	24 (40.7%)
Time from presentation to death (n = 24)	1.3 ± 1.2 years
mRS score at last f/u	2.23 ± 2.49
Frankel grade at last f/u
B (complete motor loss, incomplete sensory)	2 (3.4%)
C (incomplete motor loss)	1 (1.7%)
D (incomplete motor loss but able to ambulate)	9 (15.3%)
E (no neurologic deficit)	47 (79.7%)
Patient-reported outcome measures at last f/u^c
NDI (n = 27)	45.8 ± 18.0
Neck pain VAS (n = 13)	1.6 ± 2.3
PROMIS (n = 6)	55.3 ± 12.2
Neurosurgery clinic follow-up 30 days post-injury	50 (84.7%)

Value reported as mean±SD or no. (%).

Hospital records were accessed in all 59 patients for a f/u duration of at least 2 years for the main outcome of interest: mortality (range from 2 days in the case of in-hospital death to 4 years). Median clinic follow-up duration = 175 days (IQR = 355). Median imaging follow-up interval = 116 days (IQR = 376). Median follow-up documenting pain scales = 116 days (IQR = 249).

14 patients had no radiographic outcomes after 30 days postoperatively, meaning 69.5% of total patients with or without radiographic follow-up had stable fractures at last f/u. Stability was defined as no new displacement or motion on cervical x-rays.

17 patients were assessed for evidence of bony fusion in addition to stability at last f/u. Fusion was defined by evidence of bridging trabecular bone between fracture endplates on imaging.

13 patients had no f/u patient-reported outcome measures.

SNF, skilled nursing facility; NA, not available; mRS, modified Rankin scale; f/u, follow-up; NDI, neck disability index; VAS, visual analogue scale; PROMIS, Patient-reported Outcomes Measurement Information System.

Three patients (5.1%) had a related readmission. One patient treated with rigid bracing fell without his collar, leading to reinjury of his neck three months after initial presentation. Two surgical patients had a related readmission: one presented three months postoperatively with hardware failure and the second returned within 30 days for altered mental status thought to have occurred in the setting of postoperative medications. The patient who experienced hardware failure underwent reoperation. Two patients (5.7%) who were initially managed conservatively underwent subsequent operative intervention for treatment failure in the form of ongoing clinically significant fracture instability. According to institutional and outside records, 24 patients (40.7%) died. Death occurred at an average of 1.3 years after initial presentation for these 24 patients. At last clinical follow-up, the majority of patients had no neurologic deficit (79.7%), and 15.3% were at least able to ambulate. Thirty-nine (66.1%) patients in our cohort had standardized pain scales recorded at a median of 116 days (IQR = 249) postoperatively. The average neck disability index was 45.8 (SD ± 18; n = 27), mean neck pain visual analog scale score was 1.6 (SD ± 18; n = 13), and mean Patient-Reported Outcomes Measurement Information System (PROMIS) form v1.0 pain interference score was 55.3 (SD ± 18; n = 6).

Table 3 displays a comparative analysis of patients using operative treatment vs nonoperatively treated patients. Operatively managed patients were younger on average (73.4 vs 80.6 years, P < .001) and more likely to be male, but this difference was not significant (54.2% vs 31.4%, P = .081). Operative patients had a greater degree of fracture displacement at presentation (3.45 vs .96 mm displacement, P = .002). Conservatively managed patients were braced for an average of 12 weeks (SD ± 6.7). C1-C2 posterior spinal fusion was performed in 62.5% of operative patients, followed in frequency by posterior fusion of C1-C3 (16.7%) and C1-C4 (8.3%), with C1-T1, occiput-C2, and occiput-C6 fusion performed in one patient each. There were no anterior odontoid screw fixations, the criteria for which are the presence of a well-aligned acute fracture that is perpendicular to the screw trajectory, an intact transverse ligament, and a body habitus that allows for appropriate screw trajectory. Three patients who were managed conservatively and none of the surgical candidates died within 30 days of presentation (8.6% vs 0%, P = .141). One patient already on hospice with severe Alzheimer disease returned to hospice and died within 30 days. The other 2 deaths occurred in the hospital. One patient experienced a type II odontoid fracture resulting from collapse during a cardiac arrest, which led to his death later during the hospitalization, and the second patient had severe Alzheimer disease and sepsis and was transitioned to comfort measures and later died.

Table 3.

Comparison of Characteristics for Operatively vs Nonoperatively Managed Patients.

	Operative (n = 24)	Nonoperative (n = 35)	P-value
Mean age (y)	73.4 ± 7.6	80.6 ± 7.8	<.001
Sex
Male	13 (54.2%)	11 (31.4%)	.08
Female	11 (45.8%)	24 (68.6%)	.08
Mean BMI	26.4 ± 5.2	27.2 ± 5.8	.58
Mean A-P fracture displacement (mm)	3.45 ± 3.21	.96 ± 2.09	.002
Median time from presentation to operating room (d)	11	NA	NA
Mean duration of conservative bracing management (weeks)	NA	12.0 ± 6.7	NA
Operative intervention
C1-2 PSF	15 (62.5%)	NA	NA
C1-3 PSF	4 (16.7%)
C1-4 PSF	2 (8.3%)
C1-T1 PSF	1 (4.2%)
O-C2 PSF	1 (4.2%)
O-C6 PSF	1 (4.2%)
Mean mRS score at presentation	1.75 ± 1.15	2.29 ± 1.36	.12
Mean mRS score at last f/u	2.21 ± 2.54	2.29 ± 2.49	.91
Mean mFI-5 at presentation	1.67 ± 1.38	1.77 ± 1.11	.75
Mean mFI-11 at presentation	2.58 ± 1.74	2.86 ± 2.20	.61
Mean CCI	3.17 ± 2.33	3.23 ± 3.46	.94
Other mobility-limiting injury^b	1 (4.2%)	5 (14.3%)	.21
Mortality at last f/u	8 (33.3%)	16 (45.7%)	.34
30-day mortality	0 (0%)	3 (8.6%)	.14
Related readmission	2 (8.3%)	1 (2.9%)	.57
Discharge disposition
Home	13 (54.2%)	18 (51.4%)	.58
Acute rehab	1 (4.2%)	1 (2.9%)
SNF	4 (16.7%)	2 (5.7%)
Assisted living	4 (16.7%)	9 (25.7%)
Deceased/hospice	1 (4.2%)	4 (11.4%)
NA	1 (4.2%)	1 (2.9%)
Clavien Dindo score at last f/u
0	19 (79.2%)	NA	NA
1	2 (8.3%)
2	1 (4.2%)
3a	1 (4.2%)
3b	1 (4.2%)
Mean length of stay (d)	4.7 ± 3.7	2.8 ± 3.8	.07
ICU need	9 (37.5%)	7 (20.0%)	.14
Fracture instability at last radiographic f/u^a	1 (4.8%)	3 (13.6%)	.61
Reoperation	1 (4.2%)	2 (5.7%)	1.00
Frankel grade at last f/u
B (complete motor loss, incomplete sensory)	1 (4.2%)	1 (2.9%)	.63
C (incomplete motor loss)	1 (4.2%)	0 (0%)
D (incomplete motor loss but able to ambulate)	3 (12.5%)	6 (17.1%)
E (no neurologic deficit)	19 (79.2%)	28 (80.0%)
Mean NDI at last f/u (n = 27 total)	46.2 ± 21.6	43.4 ± 15.7	.71
Mean neck pain VAS at last f/u (n = 13 total)	2.3 ± 2.8	.2 ± 0.4	.09
Chronic pain			.62
Chronic pain with chronic opioid use	10 (41.7%)	12 (34.3%)
Chronic pain without opioid use	6 (41.7%)	13 (37.1%)

14 patients who were missing radiographic outcomes >30 days postoperatively were excluded from the fracture stability analysis. Fracture instability was defined as new or worsening fracture displacement or motion on imaging.

6 patients had other injuries that potentially limited their mobility including pelvic fractures (n = 3), multiple metatarsal factures (n = 1), lumbar compression fracture (n = 1), and a left tibial plateau fracture (n = 1).

BMI, body mass index; A-P, anterior-posterior; PSF, posterior spinal fusion; mRS, modified Rankin scale; f/u, follow-up; mFI, modified frailty index; CCI, Charlson Comorbidity Index; SNF, skilled nursing facility; NA, not available; ICU, intensive care unit; NDI, neck disability index; VAS, visual analogue scale.

Operative patients were more likely to return for a related readmission (16.7% vs 2.9%), but this difference was not statistically significant (P = .148). Fracture instability as noted in the surgeon’s final follow-up note occurred in one operatively managed patient (4.2%) and three conservatively managed patients (13.6%) (P = .609). Two patients in the conservatively managed group and one in the operative group underwent subsequent surgery in the cervical spine (5.7% and 4.2%, respectively, P = 1.0). BMI, mRS score at presentation, mRS score at last clinic follow-up, mFI-5 at presentation, mFI-11 at presentation, CCI, the presence of other mobility-limiting injuries, mortality, discharge disposition, ICU need, length of stay, rate of fracture instability, Frankel Grade, neck pain, and the presence of chronic pain and chronic opioid use were not different between the treatment groups at last clinic follow-up.

As part of our analysis, we dichotomized the cohort around the age of 75 to evaluate patient characteristics and outcomes in older and younger patients (Table 4). There were statistically significantly more females than males in the ≥75 group (74% vs 29%, P = .015). Patients who were younger than 75 years old were more likely to undergo operative intervention (60% vs 29%, P = .019). mRS score at presentation was higher among the ≥75 age group (2.5 ± 1.3 vs 1.4 ± 1.0, P = .001). Variables that did not differ between age groups include BMI, mean anterior-posterior fracture displacement, fracture stability at final radiographic follow-up, MFI-5, MFI-11, CCI, neck disability index, visual analog scale score, PROMIS, length of stay, readmissions, and mRS score and mortality.

Table 4.

Patient Characteristics and Outcomes Dichotomized Around the Age of 75.

Variable	Age <75 years (n = 25)	Age ≥75 years (n = 34)	P-value
Age (y)	69.4 ± 3.5	83.6 ± 5.1	<.001^a
Sex			.015a
Male	15 (60%)	10 (29%)
Female	10 (40%)	24 (74%)
BMI	27.7 ± 6.2	26.3 ± 4.9	.328
A-P fracture displacement (mm)	2.1 ± 2.9	2.0 ± 2.9	.923
Operative intervention	15 (60%)	10 (29%)	.019a
Fracture stability at last f/u*	20 (80%)	21 (61.8%)	1.000
mFI-5	1.6 ± 1.2	1.8 ± 1.3	.558
mFI-11	2.5 ± 1.7	3.0 ± 2.6	.397
CCI	3.8 ± 3.8	3.0 ± 3.6	.316
mRS score at presentation	1.4 ± 1.0	2.5 ± 1.3	.001^a
mRS score at last f/u	1.8 ± 2.5	2.5 ± 2.5	.306
NDI	45.6 ± 19.3	44.6 ± 19.9	.886
Neck pain VAS	2.4 ± 3.0	.7 ± 1.5	.230
Length of stay (d)	3.0 ± 4.0	3.9 ± 3.8	.389
Related readmission	1 (4%)	2 (5.9%)	1.000
Mortality	9 (36.0%)	15 (44.1%)	.531

14 patients who were missing radiographic outcomes >30 days postoperatively were excluded.

BMI, body mass index; A-P, anterior-posterior; f/u, follow-up; mFI, modified frailty index; CCI, Charlson Comorbidity Index; mRS, modified Rankin scale; NDI, neck disability index; VAS, visual analogue scale.

The multivariate analysis of mortality outcomes results is shown in Table 5. Female sex was found to be protective of all-cause mortality (OR .142, 95% CI .034-.596, P = .008). mRS score at presentation (OR 2.06, 95% CI 1.03-4.10, P = .039) was independently associated with mortality. Age, mFI-11, CCI, fracture displacement, and operative intervention were not independently significant. Figure 3 displays the ROC curve, with an AUC of .793.

Table 5.

Binary Logistic Regression Multivariate Analysis of Variables Associated With Mortality.

Variable	P-value	OR	95% CI for OR
Variable	P-value	OR	Lower	Upper
Age at presentation (y)	.752	1.016	.923	1.118
Female sex	.008	.142	.034	.596
mFI-11 at presentation (calculated, 0-11)	.598	.886	.564	1.391
CCI (calculated)	.267	1.187	.877	1.607
AP displacement on cervical spine CT (mm)	.585	.922	.688	1.235
Operative intervention	.645	.676	.128	3.565
mRS score (at presentation)	.039	2.063	1.038	4.100

CI, confidence interval; OR, odds ratio; mFI, modified frailty index; CCI, Charlson Comorbidity Index; A-P, anterior-posterior; CT, computed tomography; mRS, modified Rankin scale.

Figure 3.

ROC analysis for the multivariate model. ROC area under the curve = .793.

Table 6 demonstrates the characteristics of patients who had died during the study period. Of these characteristics, female sex was found to be protective of all-cause mortality after type II odontoid fracture (71.4% of patients alive during the study period vs 41.7% of patients who died, P = .022). Notably, neurologic deficit was not associated with mortality in this study (8.6% vs 12.5%, P = .624). Evidence of bony fusion did not differ between groups, but the sample size was small. The time between presentation and surgical intervention for patients who were alive during the study period tended to be longer but this difference was not statistically significant (mean rank 14.22 vs 9.06 days, P = .093, Mann-Whitney U test).

Table 6.

Characteristics Among Patients Who Died by Last Follow-Up.

Variable	Alive at last f/u (n = 35)	Deceased at last f/u (n = 24)	P-value
Operative intervention	16 (45.7%)	8 (33.3%)	.342
Female sex	25 (71.4%)	10 (41.7%)	.022
Age >75 years	19 (54.3%)	15 (62.5%)	.531
ICU stay	8 (22.9%)	8 (33.3%)	.374
Neurologic deficit at presentation	3 (8.6%)	3 (12.5%)	.624
Other mobility-limiting injuries	4 (11.4%)	2 (8.3%)	.699
Fusion (n = 17)	5 (14.3%)	5 (20.8%)	.763
Time to osteosynthesis in days (median)^a	14.22 (17.0)	9.06 (1.5)	.093

^aMann-Whitney U test.

f/u, follow-up; ICU, intensive care unit.

Discussion

We studied the impact of patient factors and treatment on outcomes. Our multivariate analysis found that although treatment modality had no significant impact on all-cause mortality within the study period, mRS score at presentation was a significant predictor of patient mortality. In an aging population, the question of surgical vs conservative management of Type II odontoid fractures is of growing import.

Surgical management of Type II odontoid fractures may involve anterior fixation of the odontoid process to the C2 vertebral body. The anterior approach is associated with higher rates of nonunion among older patients and is favored in minimally displaced fractures.^21,22 C1-C2 posterior fusion boasts a superior rate of bony fusion, especially in the elderly, and minimizes complications like dysphagia compared with the anterior screw osteosynthesis approach.^21,23,24 Patients treated conservatively may be placed in a semi-rigid cervical collar for 12 weeks or in halo immobilization. Surgical candidates in our cohort each underwent posterior fusion, and conservatively managed patients were in a rigid collar for an average of 12 weeks. In the literature, treatment success has historically hinged upon radiographic evidence of osseous fusion or assessments of subsequent morbidity and mortality.

Weighing expected outcomes is vital to surgical decision-making. Most of the literature on the management of type II odontoid fractures in the elderly has sought to answer the question of surgical vs conservative therapy. The results are mixed, and the subject remains controversial. Proponents of surgical intervention cite superior rates of bony fusion and decreased likelihood of treatment failure.^23,25-28 They tend to agree that patients who are younger, have fewer comorbidities, and have significant fracture displacement are more likely to benefit from undergoing surgery over conservative management.^6,21,29 Authors who argue for a selective conservative approach have indicated that surgery carries a potentially higher complication risk and provides equivalent functional outcomes.^27,30-33 Some articles propose that functional status is more critical than radiographical evidence of bony fusion.³¹ In addition, they claim that fibrous union resulting from conservative treatment provides sufficient stability for positive clinical outcomes.^27,31,34 Patients in the present study achieved similar radiographical and clinical evidence of stability regardless of surgical intervention.

When dichotomized around the age of 75 years, older patients in our study had a higher mRS score at presentation but mortality did not differ between groups based on age alone. Of note, patients chosen for surgery in our cohort tended to be younger (73.4 y vs 80.6 y, P < .001). As discussed above, this patient selection pattern aligns with the favored practice as described in the current literature, which states that patients above the age of 80 years are less likely to benefit from surgery.^7,24,32,33 Graffeo et al⁷ demonstrated no survival benefit after surgical management of patients over the age of 80 years. Fan et al²⁴ showed that patients younger than age 70 years experience higher rates of union and lower rates of mortality after surgical intervention. Barlow et al³⁵ revealed that operative management of Type II odontoid fracture was cost-effective for elderly patients below the age of 84 years using $100,000/QALY. Above the age of 84 years, surgeries were comparatively more costly and less effective. In addition to the potential for higher complications, cost-effectiveness is a vital measure to consider when outcomes are seemingly equal. Our results and those of the authors cited above suggest that the older patients are at presentation, the more likely they are to be best served by conservative therapy.

Male sex was found to be statistically significantly associated with all-cause mortality after a Type II odontoid fracture. This finding agrees with the findings from a prior study by Chapman et al,²⁶ who identified an elevated mortality risk in males over females after odontoid fractures (52% vs 38%, P = .013). Similarly, a study reported in JAMA in 2009 found that that older males have a higher risk of mortality after any major fracture.³⁶

Evidence of fracture stability at final radiographic follow-up did not significantly differ between operatively and conservatively managed patients; however, individuals who underwent surgical stabilization in our cohort also had a statistically significantly larger degree of fracture displacement (3.45 vs .96 mm displacement, P = .002). Fractures with more significant displacement carry a lower likelihood of eventual osseous fusion and stable clinical outcomes without surgical intervention.^21,37 The literature also suggests that patients with minimal fracture displacement more readily arrive at a clinically stable outcome with conservative therapy.³⁸

We found no relationship between treatment modality and functional outcome. In this case, other patient characteristics such as age and degree of fracture displacement should play a role in determining appropriate intervention. Additionally, patients may be stratified based on an assessment of neurologic deficits. Although no association between neurologic deficit and mortality was found in our study, Patel et al³⁹ have pointed out that patients who present with neurologic deficit are at an increased risk of death after Type II odontoid fractures.

Our study demonstrates that frailty and disability are vital to preoperative assessment. We recorded each patient’s mRS score, CCI, mFI-5, and mFI-11, which have proven utility in patient prognostication. Carlstrom et al.¹⁸ demonstrated the utility of the mIF-5, modified CCI, and Davies frailty indices in predicting mortality and selecting operative vs nonoperative treatment for elderly patients with Type II odontoid fractures. Their results support our claim that assessing patient frailty in addition to age is critical in determining which patients will be best served by surgical intervention. Few other studies have focused the role of frailty in determining patient mortality after Type II odontoid fracture.^26,40 To this literature, we add the utility of mRS score in the assessment of elderly patients with Type II odontoid fracture.

The use of mRS score to predict patient outcomes has been validated in other pathologies such as stroke.⁴¹ Our multivariate regression revealed a correlation between a patient’s presenting mRS score and their mortality within the study period (OR 2.24, 95% CI 1.13-4.43, P = .020). mRS score may predict death after Type II odontoid fracture regardless of age, severity of fracture displacement, operative intervention, or neurological deficit. After initially selecting patients based on age, comorbidities, and degree of fracture displacement, we propose using a patient’s mRS score as a potential branch point in determining operative vs conservative management.

Limitations

This is a retrospective study representing the experience of a single institution. Patient death occurred most frequently at an outside facility, which limited our ability to accurately record the cause of death. Post-treatment radiographical assessment was often limited to x-ray, which is inherently less reliable than other imaging modalities when assessing stability and bony fusion. Pain scales were often not recorded upon presentation, limiting our analysis of pre- and post-intervention pain. Future studies will ideally use a randomized controlled design and include more patients across multiple institutions.

Conclusion

Type II odontoid fractures are disproportionately morbid in the elderly population. Choosing optimal candidates for either surgical or conservative management is essential. Patients in this cohort were selected for operative intervention in a manner consistent with the literature, which states that patients who are younger and experience a higher degree of fracture displacement may benefit from operative fracture treatment. Mortality, fracture stability, mRS score, mFI-5, mFI-11, CCI, and pain at last clinic follow-up did not differ based on treatment modality. Our study shows that male sex and an elevated mRS at presentation correlate with greater mortality risk after Type II odontoid fracture. Other frailty and comorbidity measures failed to predict outcome. Additional studies are needed to further characterize which patients are the best candidates for surgical or conservative interventions in the setting of Type II odontoid fracture.

Footnotes

Acknowledgments

We thank Kristin Kraus, MSc, for editorial support.

Declaration of Conflicting Interests

The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: Dr. Dailey is a consultant for Zimmer Biomet Spine. Dr. Bisson is a consultant for MiRus, Stryker Spine, and nView and has equity interest in MiRus. The other authors have no potential conflicts to report.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Kyril Cole

Marcus D. Mazur

Andrew T. Dailey

Erica F. Bisson

References

Robinson

Moller

Robinson

, et al. Fracture subtypes, incidence, and treatment allocation change with age: A retrospective cohort study of 233 consecutive cases. Biomed Res Int. 2017;2017:8321680. doi:10.1155/2017/8321680.

Muller

Wick

Russe

, et al. Management of odontoid fractures in the elderly. Eur Spine J. 1999;8:360-365. doi:10.1007/s005860050188.

Pepin

Bourne

Hawkins

. Odontoid fractures, with special reference to the elderly patient. Clin Orthop Relat Res. 1985;193:178-183.

Anderson

D’Alonzo

. Fractures of the odontoid process of the axis. J Bone Joint Surg Am. 1974;56:1663-1674.

Robinson

Olerud

Robinson

. Epidemiology of C2 fractures in the 21st century: a national registry cohort study of 6,370 patients from 1997 to 2014. Adv Orthop. 2017;2017:6516893. doi:10.1155/2017/6516893.

Rizvi

SAM

Helseth

Ronning

, et al. Odontoid fractures: Impact of age and comorbidities on surgical decision making. BMC Surg. 2020;20:236. doi:10.1186/s12893-020-00893-7.

Graffeo

Perry

Puffer

, et al. Deadly falls: operative versus nonoperative management of Type II odontoid process fracture in octogenarians. J Neurosurg Spine. 2017;26:4-9. doi:10.3171/2016.3.SPINE151202.

United Nations Department of Economic and Social Affairs--Population Division . World Population Ageing 2019. New York:United Nations;2020. ST/ESA/SER.A/444.

Charlson

Pompei

Ales

, et al. A new method of classifying prognostic comorbidity in longitudinal studies: Development and validation. J Chronic Dis. 1987;40:373-383. doi:10.1016/0021-9681(87)90171-8.

10.

Rankin

. Cerebral vascular accidents in patients over the age of 60. II. Prognosis. Scott Med J. 1957;2:200-215. doi:10.1177/003693305700200504.

11.

Subramaniam

Aalberg

Soriano

, et al. New 5-factor modified frailty index using American College of Surgeons NSQIP data. J Am Coll Surg. 2018;226:173-181. doi:10.1016/j.jamcollsurg.2017.11.005.

12.

Velanovich

Antoine

Swartz

, et al. Accumulating deficits model of frailty and postoperative mortality and morbidity: Its application to a national database. J Surg Res. 2013;183:104-110. doi:10.1016/j.jss.2013.01.021.

13.

Chibbaro

Mallereau

Ganau

, et al. Odontoid type II fractures in elderly: What are the real management goals and how to best achieve them? A multicenter European study on functional outcome. Neurosurg Rev. 2021;45:709-718. doi:10.1007/s10143-021-01594-2.

14.

De Bonis

Iaccarino

Musio

, et al. Functional outcome of elderly patients treated for odontoid fracture: A multicenter study. Spine (Phila Pa 1976). 2019;44:951-958. doi:10.1097/BRS.0000000000002982.

15.

Lofrese

Musio

De Iure

, et al.

Type II odontoid fracture in elderly patients treated conservatively: Is fracture healing the goal?

Eur Spine J. 2019;28:1064-1071. doi:10.1007/s00586-019-05898-2.

16.

Karam

Tsiouris

Shepard

, et al. Simplified frailty index to predict adverse outcomes and mortality in vascular surgery patients. Ann Vasc Surg. 2013;27:904-908. doi:10.1016/j.avsg.2012.09.015.

17.

Alas

Segreto

Chan

, et al. Association between frailty status and odontoid fractures after traumatic falls: investigation of varying injury mechanisms among 70 elderly odontoid fracture patients. J Orthop Trauma. 2019;33:e484-e488. doi:10.1097/BOT.0000000000001597.

18.

Carlstrom

Helal

Perry

, et al. Too frail is to fail: Frailty portends poor outcomes in the elderly with type II odontoid fractures independent of management strategy. J Clin Neurosci. 2021;93:48-53. doi:10.1016/j.jocn.2021.08.027.

19.

Dindo

Demartines

Clavien

. Classification of surgical complications: A new proposal with evaluation in a cohort of 6336 patients and results of a survey. Ann Surg. 2004;240:205-213. doi:10.1097/01.sla.0000133083.54934.

20.

Frankel

Hancock

Hyslop

, et al. The value of postural reduction in the initial management of closed injuries of the spine with paraplegia and tetraplegia. I. Paraplegia. 1969;7:179-192. doi:10.1038/sc.1969.30.

21.

Joaquim

Patel

. Surgical treatment of Type II odontoid fractures: Anterior odontoid screw fixation or posterior cervical instrumented fusion? Neurosurg Focus. 2015;38(4):E11. doi:10.3171/2015.1.FOCUS14781.

22.

Tian

, et al. Pooled analysis of non-union, re-operation, infection, and approach related complications after anterior odontoid screw fixation. PLoS One. 2014;9:e103065. doi:10.1371/journal.pone.0103065.

23.

Andersson

Rodrigues

Olerud

. Odontoid fractures: High complication rate associated with anterior screw fixation in the elderly. Eur Spine J. 2000;9:56-59. doi:10.1007/s005860050009.

24.

Fan

Huang

, et al. Surgery vs conservative treatment for type II and III odontoid fractures in a geriatric population: A meta-analysis. Medicine (Baltimore). 2019;98:e10281. doi:10.1097/MD.0000000000010281.

25.

Kuntz

C 4th

Mirza

Jarell

, et al. Type II odontoid fractures in the elderly: Early failure of nonsurgical treatment. Neurosurg Focus. 2000;8(6):e7. doi:10.3171/foc.2000.8.6.8.

26.

Chapman

Smith

Kopjar

, et al. The AO Spine North America geriatric odontoid fracture mortality study: A retrospective review of mortality outcomes for operative versus nonoperative treatment of 322 patients with long-term follow-up. Spine (Phila Pa 1976). 2013;38:1098-1104. doi:10.1097/BRS.0b013e318286f0cf.

27.

Seybold

Bayley

. Functional outcome of surgically and conservatively managed dens fractures. Spine (Phila Pa 1976). 1998;23:1837-1845. doi:10.1097/00007632-199809010-00006.

28.

Bednar

Parikh

Hummel

. Management of type II odontoid process fractures in geriatric patients; a prospective study of sequential cohorts with attention to survivorship. J Spinal Disord. 1995;8:166-169.

29.

Blauth

Richter

Kiesewetter

, et al. [Operative versus non operative treatment of odontoid non unions. How dangerous is it not to stabilize a non union of the dens?]. Chirurg. 1999;70:1225-1238. doi:10.1007/s001040050774.

30.

Muller

Schwinnen

Fischer

, et al. Non-rigid immobilisation of odontoid fractures. Eur Spine J. 2003;12:522-525. doi:10.1007/s00586-003-0531-1.

31.

Pal

Sell

Grevitt

. Type II odontoid fractures in the elderly: An evidence-based narrative review of management. Eur Spine J. 2011;20:195-204. doi:10.1007/s00586-010-1507-6.

32.

Perry

Graffeo

Carlstrom

, et al. Fusion, failure, fatality: Long-term outcomes after surgical versus nonoperative management of type II odontoid fracture in octogenarians. World Neurosurg. 2018;110:e484-e489. doi:10.1016/j.wneu.2017.11.020.

33.

Borsotti

Starnoni

Ecker

, et al.

One-year follow-up for type II odontoid process fractures in octogenarians: Is there a place for surgical management?

Surg Neurol Int. 2020;11:285. doi:10.25259/SNI_376_2020.

34.

Clark

White

3rd . Fractures of the dens. A multicenter study. J Bone Joint Surg Am. 1985;67:1340-1348.

35.

Barlow

Higgins

Ozanne

, et al. Cost effectiveness of operative versus non-operative treatment of geriatric type-II odontoid fracture. Spine (Phila Pa 1976). 2016;41:610-617. doi:10.1097/BRS.0000000000001275.

36.

Bliuc

Nguyen

Milch

, et al. Mortality risk associated with low-trauma osteoporotic fracture and subsequent fracture in men and women. JAMA. 2009;301:513-521. doi:10.1001/jama.2009.50.

37.

Platzer

Thalhammer

Sarahrudi

, et al. Nonoperative management of odontoid fractures using a halothoracic vest. Neurosurgery. 2007;61:522-529. doi:10.1227/01.NEU.0000290898.15567.21.

38.

Gonschorek

Vordemvenne

Blattert

, et al. Treatment of odontoid fractures: Recommendations of the Spine Section of the German Society for Orthopaedics and Trauma (DGOU). Global Spine J. 2018;8:12S-17S. doi:10.1177/2192568218768227.

39.

Patel

Smith

Radcliff

, et al. Odontoid fractures with neurologic deficit have higher mortality and morbidity. Clin Orthop Relat Res. 2012;470:1614-1620. doi:10.1007/s11999-011-1994-8.

40.

Shafafy

Valsamis

Luck

, et al. Predictors of mortality in the elderly patient with a fracture of the odontoid process. Bone Joint J. 2019;101-B:253-259. doi:10.1302/0301-620X.101B3.BJJ-2018-1004.R1.

41.

Banks

Marotta

. Outcomes validity and reliability of the modified Rankin scale: Implications for stroke clinical trials: A literature review and synthesis. Stroke. 2007;38:1091-1096. doi:10.1161/01.STR.0000258355.23810.c6.

Using Frailty Measures to Predict Functional Outcomes and Mortality After Type II Odontoid Fracture in Elderly Patients: A Retrospective Cohort Study

Abstract

Study Design

Objectives

Methods

Results

Conclusions

Keywords

Introduction

Methods

Study Design and Patient Population of Interest

Functional Status and Comorbidity/Frailty Variables of Interest

Outcomes of Interest

Statistical Analysis

Results

Discussion

Limitations

Conclusion

Footnotes

Acknowledgments

Declaration of Conflicting Interests

Funding

ORCID iDs

References