Orthopaedics & Traumatology: Surgery & Research (2013) 99, S301—S312

Available online at

www.sciencedirect.com

ORIGINAL ARTICLE

Complications of upper cervical spinetrauma in elderly subjects. A systematicreview of the literature

P. Jubert, G. Lonjon, C. Garreau de Loubresse ∗, the Bone andJoint Trauma Study Group (GETRAUM)1

Service de chirurgie orthopédique, hôpital Raymond-Poincaré, 104, boulevard Raymond-Poincaré, 92380Garches, France

Accepted: 30 April 2013

KEYWORDSSpinal injuries;Axis;Cervical atlas;Complications;Systematic review

Summary The frequency of cervical spine trauma in elderly patients is increasing with mostinjuries occurring in the upper cervical spine. These fractures are associated with a risk ofsometimes life-threatening complications, although very few studies have specifically analyzedthis. The goal of this study was to identify the incidence of complications in the literature(mortality and morbidity) following upper cervical spine trauma in elderly patients.Methods: A systematic search was performed on the MEDLINE database without limiting thesearch by language or date to identify all studies reporting the rate of complications afterupper cervical spine trauma in patients over the age of 60.Results: Twenty-four observational studies were included, four were comparative. These stud-ies included a total of 857 patients, mean age 76. Nearly all traumas were odontoid processfractures, and most were treated surgically (57%). The median mortality rate was 9.2% (Q1—Q3:2.5—19.6) and the median rate of short-term complications was 15.4% (Q1—Q3: 5.8—26.9). Themain late stage complication was nonunion, which developed in a mean 10 to 12% dependingon the type of treatment.Conclusion: Complications following cervical spine trauma are frequent in elderly patients

whatever the type of treatment. Knowledge of the rate of complications in the literature and the potential risk factors is essepatients and to prevent complicType of Study: Systematic revie© 2013 Elsevier Masson SAS. All

∗ Corresponding author. Tel.: +33 01 47 10 77 36.E-mail address: ch.garreau@rpc.aphp.fr (C. Garreau de Loubresse).

1 56, rue Boissonade, 75014 Paris, France.

1877-0568/$ – see front matter © 2013 Elsevier Masson SAS. All rights rehttp://dx.doi.org/10.1016/j.otsr.2013.07.007

ntial for the clinician to improve the information provided toations.w of the literature. Level of evidence IV.rights reserved.

served.

S

I

TpyvCosa

ohlscnsbitaa

ishhiUni

s

(i

M

Tg

owmboiee

uiiP‘i‘

acwitof the population, type of fracture, therapeutic strategy,

FftFC

302

ntroduction

he frequency of traumatic cervical spine injuries inatients over 65 is increasing, while it is decreasing inounger patients [1—3]. Most injuries occur in the upper cer-ical spine (UCS) especially in the odontoid process [4—7].lassification of odontoid fractures is based on the locationf the fracture and the direction of the fracture line. Clas-ifications by Roy Camille, Anderson and D’Alonzo or Grauerre the most frequently used scores (Fig. 1) [8].

The association of osteoporosis of the UCS andsteoarthritis of the lower cervical spine (LCS) explains theigh frequency of these injuries, which usually occur duringow energy traumas [4,9—11]. In most cases the neurologicaltatus of the patient is unaffected and there are very fewlinical signs, thus explaining the frequent delay in diag-osis [11]. Two therapeutic strategies are possible. First,urgical treatment by anterior or posterior approach to sta-ilize the fracture. The other is conservative treatment bymmobilization in an external brace. The goal of both thesereatments is to stabilize the injuries to obtain union andllow patients to rapidly reach the same level of autonomys before the injury.

Unlike fractures of the upper end of the femur (FUEF)n the elderly, which have been the subject of numeroustudies and guidelines, UCS injuries are associated with aigh rate of mortality and morbidity, whose risk factorsave not yet all been identified. [12,13]. There is no exist-ng consensus on the therapeutic management of traumaticCS injuries. Like FUEF, guidelines could be drafted if theegative predictive factors of survival in patients with UCS

njuries were known.

Thus, the goal of this study was to perform a systematictudy of the literature to evaluate the complication rate

lrc

igure 1 Types of classification used for odontoid fractures. Anderacture. C. type III fracture). Raymond Roy Camille Classification. (Aype fracture [slanted fracture line below and behind]. C. Fracture de. de Peretti, M. Challali. Traumatismes du rachis cervical supérieur.

opyright© 2012 Elsevier Masson SAS. All rights reserved.

P. Jubert et al.

mortality and morbidity) after treatment of traumatic UCSnjuries in patients over the age of 60.

aterials and methods

his review was performed according to PRISMA [14] groupuidelines and was not saved in a database.

All studies published on traumatic UCS injuries in patientsver 60 in English or in French were searched for. Studiesere selected with no time limit, and had to report theorbidity and mortality of these injuries. The studies coulde randomized trials, cohort studies, case control studiesr series of cases with at least 10 included patients. Clin-cal cases and systematic reviews of the literature werexcluded. Studies that did not report on follow-up werexcluded.

A search was performed in the MEDLINE database (lastpdate: September 2012) with an association of words todentify the elderly population with upper cervical spinenjury, with no time limitation. The terms searched for inubmed were ‘‘odontoid process’’ ‘‘axis’’ ‘‘cervical atlas’’‘atlato-axial joint’’ (MeSH), ‘‘fracture bone’’ ‘‘spinalnjuries’’ ‘‘spinal fracture’’ ‘‘aged’’ (MeSH) ‘‘elderly’’ and‘fracture fixation’’ (text).

Studies were selected based on the title and the abstract,nd then two independent readers read the entire arti-le. Moreover, the references of the identified studiesere evaluated to search for missing studies. The follow-

ng descriptive features were extracted from all studies:ype of study, size of the study sample, characteristics

ength/duration of patient follow-up. The mortality rate wasecorded at 3 months and at the final follow-up. The compli-ation rate was noted as well as the date this was evaluated.

rson and D’Alonzo Classification. (A. type I fracture. B. type II. OBAV type fracture [oblique line below and in front]. B. OBAR

type HTAL [horizontal fracture line]). This figure is taken from:EMC-Appareil locomoteur 2012;7(4):1—8 [Article 15-825-A-10].

wrsrcTmaatosm

M

To(mpcph(

utInsisrietT(g

waaw7tttwigtp(t

Upper cervical spine trauma in elderly subjects

Because there were no homogenous randomized or non-randomized comparative studies, a meta-analysis could notbe performed to compare the two types of treatment. Onlya descriptive analysis of the mortality and morbidity ratescould be obtained with this systematic review of the litera-ture.

Results

Three hundred thirty three MEDLINE references were iden-tified (Fig. 2). After reading the title and the abstract,47 studies were selected. Twenty-five studies were excludedafter reading the entire article, leaving 22 studies in the finalanalysis. Two studies were added after the bibliographic ref-erences had been evaluated. Thus a total of 24 studies wereincluded in this systematic review of the literature.

Study characteristics

All selected studies were retrospective [15—38] and only4 were comparative [23,27,34,38] (Table 1). The selectedstudies described the management of UCS fractures in857 patients over 60 years old. The mean age in 22/24studies was 76. The gender ratio based on 20 studieswas 1.2 women/1 man (333 W/274 M). Follow-up was alwaysreported.

The type of fracture was reported in 23/24 studies(846 patients). Fractures were Anderson and D’Alonzo type IIin 803/846 cases (95%).

The mechanism of injury was reported in 16/24 studies(415 patients). The mechanism was a low energy trauma(falling from standing height) in 305/415 patients. The non-neurological clinical signs associated with injury were onlyreported in 1/24 studies and was neck pain in all cases [29].

Nineteen studies (587 patients) reported the presence ofconcomitant injuries in 110 patients: 61 injuries of the UCSother than an odontoid process fracture (58 C1 fractures,one C1—C2 dislocation and one C2 lateral mass fracture),nine LCS injuries and 40 extraspinal injuries.

Nineteen studies reported the initial neurological status(726 patients). Only one study had the presence of asso-ciated neurological symptoms as exclusion criteria. [15]. Atotal of 692 patients (55 ASIA E, 98 Frankel E, 539 with noclinical deficit) had no neurological deficits. A total of onlysix patients had tetraparesis [21,23,28,29]. The presence ofsecondary cervical myelopathy was reported in 3 patients[28,29].

The presence of comorbidities was reported in 14 stud-ies (497 patients). Four studies used the ASA score (Americansociety of Anesthesiologists) [18,20,26,37], one the Charlsoncomorbidity index (CCI) [36] and a Cumulative Illness Rat-ing Scale for Geriatrics (CIRS-G) [30]. Three studies reportedthe mean number of comorbidities per patient without usinga valid score [21,34,38]. The four remaining studies specif-ically described the different comorbidities of the patients[16,24,32,35]. Finally 1 study reported the presence ofcomorbidities in six patients without providing a specific

description [23].

The type of treatment was described in all the selectedstudies (857 patients) (Table 2). It was surgical in 485patients and conservative in 372. Anterior screw fixation

D

Ut

S303

as the most frequent surgical technique (75%), and aigid neck brace was the most frequent type of con-ervative external immobilization (65%). Fourteen studieseported the treatment choice criteria for surgical versusonservative treatment [15,17—20,22,23,25,28—31,37,38].he three most frequent criteria were the number of co-orbidities or the ASA status (6/14 studies), the patients

nd/or the surgeon’s choice (6/14 studies) and the char-cteristics of the fracture (7/14 studies). In case of surgicalreatment, internal fixation by anterior screw fixation of thedontoid process was chosen in 75% of cases. In case of con-ervative treatment a rigid or semi-rigid neck brace was theost frequently prescribed type of immobilization (84%).

orbidity and mortality

he median rate of short term complications, all typesf treatment combined was 15.4% (Q1—Q3: 5,8—26,9)Tables 2—4). The main complications of surgical treat-ent were dysphagia (30/323 patients) and respiratoryroblems (27/323 patients). The main complications ofonservative treatment were local complications (14/372atients) (mobilization or infection of the pin sites of thealo brace, skin reactions) and respiratory decompensation7/372 patients).

Long-term complications were mainly difficulties withnion and nonunion (Table 3). Twenty-one studies reportedhe rate of nonunion in odontoid fractures (768 patients).n the group of surgically treated patients (396 patients),onunion occurred in 49 (12.4%). In the group receiving con-ervative treatment (372 patients), nonunion was reportedn 40 (10.8%) and only one study reported the treatmenttrategy if union was not obtained [18]. Only 4 studieseported on whether nonunion should be treated or not whent was observed [16,23,24,33]. The studies reported a cat-gory of patients in whom stable nonunion developed dueo a fibrotic reaction of the fracture site (551 patients).his rate of ‘‘stable fibrosis’’ in the surgery group was 12.9%51/396 patients) and 25.8% in the conservative treatmentroup (40/155 patients).

All the studies reported the mortality rate in patientsith UCS injuries. The median mortality rate three monthsfter injury was 9.2% (Q1—Q3: 2.5—19.6), or 120 patients,ll types of treatment combined. Forty-three of the patientsho died had received surgical treatment (485 patients) and7 conservative treatment (372 patients). Fifty-one addi-ional deaths occurred after 3 months, including seven inhe surgical treatment group and 44 in the conservativereatment group. In the surgical treatment group, mortalityas reported in relation to the type of surgical technique

n certain cases. Twenty-two deaths were reported in theroup with odontoid screw fixation (323 patients) and 12 inhe group that was treated by posterior approach (162atients). The main cause of death was cardiopulmonary68/171patients) occurring less than 3 months after therauma in 89.7% of the cases (61/68 patients) (Table 4).

iscussion

CS fractures in general and of the odontoid process, in par-icular, are becoming a frequent ‘‘geriatric’’ trauma, and

S304

P. Jubert

et al.

Table 1 Characteristics of included studies.

Authors, year Type of study Number of patients Meanage

Force oftrauma, n (%)

Neurologicalstatusreported(type)

Odontoidfracture, n(%)

Associated UCSlesion, n (%)

Treatment Comorbidities

Hanigan et al.,1993 [15]

Retrospective 19 > 80 Low, 15 (79) Yes Ot II: 16(84.2), OtIII: 3 (15.8)

No Mixed NA

Berleman andSchwarzenbach,1997 [16]

Retrospective 19 75 Low, 19 (100) Yes Ot II: 19(100)

Fr. C1: 4 (21) Surg. 3 patients withcomorbidities

Müller et al., 1999[17]

Retrospectiveandcomparative

23 80.9 Low, 21 (91) Yes(Frankel)

Ot II: 22(95.7), OtIII: 1 (4.3)

Fr. C1: 13 (56.5) Mixed NA

Andersson et al.,2000 [18]

Retrospective 29 78 Low, 19 (65.5) Yes(Frankel)

Ot II: 24(82.8), OtIII: 5 (17.2)

Fr.C1: 3 (10) Mixed ASA Score I: 4 II: 20 III:5

Harrop et al.,2000 [19]

Retrospective 10 80 Low, 8 (80) Yes (ASIA) Ot II: 10(100)

Fr.C1: 4 (40) Surg. NA

Börm et al., 2003[20]

Retrospectiveandcomparative

15 81 NA Yes(Frankel)

Ot II: 15(100)

0 Surg. ASA Score mean: 3

Tashjian et al.,2006 [21]

Retrospective 78 80.7 NA Yes Ot II: 60(77), Ot III:18 (23)

Fr.C1: 10 (12.8) Mixed Mean number: 2.1

Frangen et al.,2007 [22]

Retrospective 27 85.5 Low, 27 (100) Yes Ot IIB: 27(100)

0 (exclusioncriteria)

Surg. NA

Platzer et al.,2007 [23]

Retrospectiveetcomparative

41 71 Low, 33 (80) Yes Ot II: 41(100)

NA Surg. 6 patients withcomorbidities

Collins et al., 2008[24]

Retrospective 15 69 Low, 9 (60) Yes Ot IIB: 15(100)

No Surg. 8 patients withcomorbidities

Koech et al., 2008[25]

Retrospective 42 80 Low, 23 (55) No Ot II: 42(100)

No Ortho. NA

Smith et al., 2008[26]

Retrospective 72 > 80 NA Yes Ot II: 72(100)

No Mixed Mean ASA score: 2,3and 2 (Ortho. andSurg.)

Daentzer etFlörkemeier,2009 [27]

Retrospectiveandcomparative

11 75.4 Low, 5 (45) No UCS NA Ortho. NA

Upper

cervical spine

trauma

in elderly

subjects

S305

Table 1 (Continued)

Authors, year Type of study Number of patients Meanage

Force oftrauma, n (%)

Neurologicalstatusreported(type)

Odontoidfracture, n(%)

Associated UCSlesion, n (%)

Treatment Comorbidities

Omeis et al., 2009[28]

Retrospective 29 80 Low, 24 (82.8) Yes Ot II: 29(100)

Fr.C1: 5 (17.2) Surg. NA

Lefranc et al.,2009 [29]

Retrospective 27 80.7 Low, 20 (74) Yes Ot II: 18(66.7), OtIII: 9 (33.3)

Fr.C1: 3 (11.1),Fr.C2: 2 (7.4),ldislocationC1C2: 1 (3.7)

Mixed NA

Chaudhary et al.,2010 [30]

Retrospective 20 > 70 Low, 16 (80) Yes Ot II: 20(100)

0 (exclusioncriteria)

Mixed NA

Dailey et al., 2011[31]

Retrospective 57 81.2 Low, 36 (63) Yes Ot II: 54(95), Ot III:3 (5)

Fr.C1: 11 (19)Fr.C2: 1 (2)

Surg. NA

Hou et al., 2011[32]

Retrospective 43 80.6 NA Yes(Frankel)

Ot II: 43(100)

NA Surg. 17 patients withcomorbidities

Mayer et al., 2011[33]

Retrospective 18 78 NA No Ot II: 14(77.8), OtIII: 4 (22.2)

Fr.C1: 2 (11) Surg. NA

Molinari et al.,2011 [34]

Retrospectiveandcomparative

34 84 NA No Ot II: 34(100)

NA Ortho. 10 patients aveccomorbidités

Osti et al., 2011[35]

Retrospective 35 79.6 Low, 26 (74) Yes (ASIA) Ot II: 35(100)

No Surg. 15 patientswithcomorbidities

Schoenfel, 2011 Retrospective 156 82 NA Yes Ot II: 156(100)

NA Mixed NA

Hénaux et al.,2012 [37]

Retrospective 11 85.4 Low, 11 (100) Yes (ASIA) Ot IIB: 11(100)

Fr.C1: 3 (27) Surg. Score ASA I: 0 II: 5 III: 6

Molinari et al.,2012 [38]

Retrospectiveandcomparative

26 79 NA No Ot II: 26(100)

NA Surg. 10 patients withcomorbidities

Ot II/III: ondontoid fracture type II/III from Anderson and D’Alonzo Classification; Ot IIB: odontoid fracture type IIB, Grauer classification; UCS: upper cervical spine; Fr.: fracture; Surg.:surgical treatment; Ortho.: conservative treatment; Mixed: conservative OR surgical treatment.

S306

P. Jubert

et al.

Table 2 Rate of complications in relation to type of treatment.

Authors, Year Number ofpatients

Number ofpatientstreatedsurgically(type)

Number ofpatientstreated con-servatively(type)

Number of complicationsafter surgical treatment(type)

Number ofcomplicationsafter conservativetreatment (type)

Total Overallincidence%

CI 95%

Hanigan et al.,1993 [15]

19 5 (posterior route) 14 (2 halo brace, 11neck braces, 1traction)

0 0 0 0.0 [0—20.9]

Berleman andSchwarzenbach,1997 [16]

19 19 (anterior route) 5 (3 respiratory failure,1 urinary infection, 1hematoma)

5 26.3 [10.1—51.4]

Müller et al., 1999[17]

23 5 (anterior route) 18 (13 halo brace, 5neck braces)

0 4 (loss ofreduction)

4 17.4 [5.7—39.5]

Andersson et al.,2000 [18]

29 18 (11 anterior route,7 posterior route)

11 (1 halo brace, 10neck braces)

1 (MCI) 1 (bedsore) 2 6.9 [1.2—24.2]

Harrop et al.,2000 [19]

10 10 (anterior route) 2 (1 pneumopathy, 1 MCI) 2 20.0 [3.5—55.7]

Börm et al., 2003[20]

15 15 (anterior route) 2 (respiratory failure) 2 13.3 [2.3—41.6]

Tashjian et al.,2006 [21]

78 13 (NA) 65 (38 halo brace, 27neck braces)

Details Chir. et Ortho.Not given (16pneumopathy, 12 ACR, 3DVT/PE, 40 urinaryinfection)

72 92.3 [83.4—96.8]

Frangen et al.,2007 [22]

27 27 (posterior route) NA

Platzer et al.,2007 [23]

41 41 (anterior route) 10 (2 cardiac failure, 3respiratory failure 3pneumopathy, 1 severeinfection, 1 DVT)

10 24.4 [12.9—40.6]

Collins et al., 2008[24]

15 15 (anterior route) 0 0 0.0 [0—25.3]

Koech et al., 2008[25]

42 42 (32 halo brace, 10neck brace)

4 (3 bedsores, 1pin mobilization)

4 9.5 [3—23.5]

Smith et al., 2008[26]

72 32 (10 anterior route,22 posterior route)

40 (16 halo brace, 24neck brace)

20 (10 respiratoryfailure, 10 dysphagia)

12 (7 respiratoryfailure, 5 urinaryinfections)

32 44.4 [32.8—56.5]

Upper

cervical spine

trauma

in elderly

subjects

S307

Table 2 (Continued)

Authors, Year Number ofpatients

Number ofpatientstreatedsurgically(type)

Number ofpatientstreated con-servatively(type)

Number of complicationsafter surgical treatment(type)

Number ofcomplicationsafter conservativetreatment (type)

Total Overallincidence%

CI 95%

Daentzer etFlörkemeier,2009 [27]

11 11 (halo brace) 8 (4 pin siteinfection, 1bedsore, 3problem withmaterial)

8 72.7 [39.3—92.6]

Lefranc et al.,2009 [29]

27 11 (8 anterior route,3 posterior route)

16 (neck brace) 0 0 0 0.0 [0—15.5]

Omeis et al., 2009[28]

29 29 (15 anterior route,14 posterior route)

2 (1 Preoperative CRF, 1respiratory failure)

2 6.9 [1.2—24.2]

Chaudhary et al.,2010 [30]

20 11 (anterior route) 9 (neck brace) NA NA

Dailey et al., 2011[31]

57 57 (anterior route) 34 (20 dysphagia, 11pneumopathies, 3 MCI)

34 59.6 [45.8—72.1]

Hou et al., 2011[32]

43 43 (42 anterior route,1 posterior route)

NA

Osti et al., 2011[35]

35 35 (anterior route) 10 (Cardiorespiratorydistress, pneumopathy,kidney failure: nodetails)

10 28.6 [15.2—46.5]

Mayer et al., 2011[33]

18 18 (anterior route) 1 (Postoperativegastrointestinalhemorrhage)

1 5.6 [0.2—29.3]

Molinari et al.,2011 [34]

34 34 (neck brace) 2 (bedsore) 2 5.9 [1—21.1]

Schoenfeld et al.,2011 [36]

156 44 (NA) 112 (NA) NA NA

Hénaux et al.,2012 [37]

11 11 (anterior route) 0 0 0.0 [0—32.1]

Molinari et al.,2012 [38]

26 26 (posterior route) 5 (1 worsening ofneurological status, 1dural breach, 1respiratory failure, 2dysphagia)

5 19.2 [7.3—39.9]

MDI: myocardial infarction; CRF: cardiorespiratory failure; DVT/PE: deep vein thrombosis, pulmonary embolism; Surg.: surgical treatment; Ortho.: conservative treatment.

S308 P. Jubert et al.

ram

sPts[

Figure 2 Flow diag

tudies on this topic are increasing in the literature [39—42].

aradoxically, available published studies often lack statis-ical power even though the number of recently publishedtudies is a sign of the growing interest in this subject31—33,36—38,42].

htpa

of article selection.

The studies included in this review of the literature

ave different methodologies, study populations, types ofreatment and reported follow-up. Moreover, there were norospective studies. Many of the series in the review includet least 20 patients [16,20,24,37]. Often, information such

Upper cervical spine trauma in elderly subjects S309

Table 3 Rate of nonunion in relation to type of treatment.

Surgery Conservative

Anterior route Posterior route

Number of patients 323 105 372Number of patients with information 313 83 155Union 202 (64.5%) 43 (51.8%) 53 (34.2%)Nonunion 35 (11.2%) 14 (16.8%) 40 (25.8%)Stable fibrosis 37 (11.8%) 14 (16.8%) 40 (25.8%)Death 22 (7%) 12 (14.6%) 19 (12.2%)

aa[mttcaownsctcatpc(t

pvcctOtnssc

idwTs

cT

Lost to follow-up 17 (5.5%)

as the initial neurological status, comorbidities, associatedinjuries and the level of autonomy before injury was notreported [25,27,33,36,38].

Many of these studies confirmed that the most frequentinjuries of the UCS were type II odontoid process fracturesand that there were two peaks in the frequency of thisentity. The first was at 30 years old and the other around 75[34]. Moreover, the category ‘‘elderly subjects’’ includes,in fact, two subcategories, less than and more than 74 yearsold, with different characteristics in each. UCS injurieswere more frequent and the mortality rate was higher inpatients over 74 [4,7,15,36]. The studies in this analysisobserved that modified anatomical structures (osteoporo-sis, osteoarthritis) were responsible for the traumatic injuryassociated with a low energy trauma, like most other trau-matic injuries in the elderly. [3,6,7,10,43].

Despite their frequency, these injuries often go unnoticedresulting in delayed treatment [5,33,35,37]. Nevertheless,two North American studies, the National Emergency Xradiography Utilization Study (NEXUS) and the Canadian C-Spine Rule Study have described clinical criteria to preventa delayed diagnosis and guidelines in the prescription of cer-vical spine X-rays following trauma [44,45]. More recently,the validity of these criteria has been questioned, in partic-ular in elderly subjects. Systematic CT Scan of the neck isrecommended in these cases [46,47].

The type of treatment was often the only cause ofmortality mentioned, while comorbidities and the patientspre-existing condition were not always described, eventhough they have been identified as a risk factor in othertraumatic injuries in the elderly [20,21,48]. Only 58% of thestudies in this review reported comorbidites and only 17%defined them. This lack of information as well as number ofdifferent scores used (ASA, CCI, CIRS-G) made it difficult toperform a statistical analysis of the risk factors of mortalityand morbidity.

The most frequent negative predictive factors in geriatrictraumatology are comorbidities and injuries concomitant tothe initial trauma [49]. Like comorbidities, these concomi-tant injuries were only reported in 62.5% of the studies.Thus, there were concomitant C1 injuries in 10% of cases,LCS injuries in only 1.5% and extraspinal injuries in 6.8%.Patients with concomitant injuries were excluded in certain

studies such as Chaudhary et al. [30] and Frangen et al. [22].Except in the study by Lefranc, the exact type of concomi-tant injuries were usually not described, whether they wereof the cervical spine or peripheral injuries [29].

v(fg

0 (0%) 3 (2%)

The presence of a secondary neurological deficit couldlso be considered a risk factor of mortality even if thessociation of a neurological deficit and UCS injury is rare3,6,10]. Decompensation of a cervical myelopathy with aotor deficit after a low energy trauma has been shown

o increase the mortality rate in elderly subjects [3,11]. Inhe same way, the mortality rate in elderly subjects with aervical spine fracture and a neurological deficit increasedt 3 months and 1 year of follow-up [10]. In this reviewf the literature, the post-traumatic neurological statusas reported in 85% of the studies but the description wasot homogenous. Certain studies used scores with detailedcales (ASIA, Frankel) while others indicated the results oflinical examinations that nevertheless made it impossibleo compare the different series. A UCS fracture was asso-iated with a neurological deficit in less than 5% of casesnd with tetraparesia in less than 1% in all the studies inhis review. This percentage seems low compared to resultsublished in certain studies. Patel recently found neurologi-al deficits in 9.6% of odontoid fractures, all ages combined42—89 years old) with a high mortality rate because 50% ofhese patients died within 3 months after injury [50].

An analysis of the literature showed that 57% of theatients received surgical management and 43% conser-ative treatment. Treatment selection criteria were notlearly defined but were based on the patient’s pre-existingondition, comorbidities, the characteristics of the frac-ure as well as the surgeon’s and patient’s preferences.dontoid screw fixation was the most frequent surgicalechnique. There were no studies showing that this tech-ique was better than others. Moreover, there are numerousurgical techniques by posterior approach (C1—C2 arthrode-is with or without a graft, wiring) that are difficult toompare.

External immobilization with a rigid neck brace was usedn most cases (65%) of conservative treatment. The meanuration of immobilization with a neck brace or a halo corsetas often long, sometimes as long as 24 weeks all [25,27,34].his long immobilization and its effect on autonomy caneem surprising in a population of elderly patients.

In the surgery group union was obtained in 62% of patientsompared to only 34% in the conservative treatment group.he rate of ‘‘stable fibrosis’’ was twice as high in the conser-

ative treatment group as in the ‘‘surgical treatment’’ group25.8% and 12.9% respectively), which can explain why dif-erence in the rate of nonunion was low between the tworoups. Stability of the fracture site provided by the fibrous

S310

P. Jubert

et al.

Table 4 Rate of mortality and causes of death.

Authors, year Number ofPatients

Mean durationof follow-up(months)

Number ofdeaths at3 months

Mortality rateat 3 months

Confidenceinterval 95%

Number of additionaldeaths at the finalfollow-up

Cause of death

Hanigan et al., 1993 [15] 19 28 5 26.3 [10.12—51.42] 0 5 cardiopulomonaryBerleman and

Schwarzenbach, 1997[16]

19 54 1 5.3 [0.28—28.11] 1

Müller et al., 1999 [17] 23 44.3 8 34.8 [17.19—57.18] 0 7 cardiopulmonary,head trauma

Harrop et al., 2000 [19] 10 10 1 10.0 [0.52—45.88] 1 1 cardiopulmonary,1 NA

Andersson et al., 2000[18]

29 51 0 0.0 [0—14.56] 10 NA

Borm. 2003 15 16.6 1 6.7 [0.35—33.97] 1 1 cardiopulmonary,1 NA

Tashjian et al., 2006 [21] 78 NA 24 30.8 [21.08—42.38] NA 24 cardiopulmonaryPlatzer et al., 2007 [23] 41 > 24 4 9.8 [3.17—24.06] NA NAFrangen et al., 2007 [22] 27 3 6 22.2 [9.38—42.73] NA 6 cardiopulmonarySmith et al., 2008 [26] 72 NA 10 13.9 [7.22—24.52] NA NACollins et al., 2008 [24] 15 18 0 0.0 [0—25.35] 0Koech et al., 2008 [25] 42 24 0 0.0 [0—10.44] 0Daentzer et Flörkemeier,

2009 [27]11 6.6 0 0.0 [0—32.14] 0

Lefranc et al., 2009 [29] 27 12 5 18.5 [7.03—38.75] 0 NAOmeis et al., 2009 [28] 29 13.5 1 3.4 [0.18—19.63] 0 1 cardiopulmonaryChaudhary et al., 2010

[30]20 > 3 4 20.0 [6.61—44.27] 0 4 cardiopulmonary

Dailey et al., 2011 [31] 57 15 5 8.8 [3.27—20.04] 0 4 cardiopulmonary,1 systemic

Osti et al., 2011 [35] 35 74 3 8.6 [2.24—24.19] 0 3 cardiopulmonarySchoenfeld et al., 2011

[36]156 36 33 21.2 [15.2—28.56] 28 NA

Hou et al., 2011 [32] 43 21.3 0 0.0 [0—10.21] 8 7 cardiopulmonary,1 systemic

Mayer et al., 2011 [33] 18 75.7 0 0.0 [0—21.88] 0Molinari et al., 2011 [34] 34 3 2 5.9 [1.03—21.06] 2 NAHénaux et al., 2012 [37] 11 34 2 18.2 [3.21—52.25] 0 NAMolinari et al., 2012 [38] 26 13.7 5 19.2 [7.31—39.98] 0 5 cardiopulmonary

[

[

[

[

[

[

[

[

[

[

[

[

[

Upper cervical spine trauma in elderly subjects

callus could prevent neurological decompensation in thesepatients and thus, the necessity of revision surgery.

Unfortunately no comparative statistical analyses couldbe performed to evaluate these heterogenous retrospectiveseries.

The optimal choice of treatment for these injuries is stillcontroversial. [51]. For Smith, for example, the outcomeis better after conservative than after surgical treatment[26]. While for Andersson, anterior screw fixation is a tech-nique which is associated with an unacceptably high rate ofcomplications [18]. On the other hand, Omeis compared thedifferent surgical techniques and did not find any significantdifference in results or complications [28].

The median mortality rate at 3 months was 9.1% alltreatments combined. Deaths were more frequent in theconservative treatment group than in the surgical treatmentgroup (21% versus 8.9%). After 3 months the mortality rateincreased in the conservative treatment group (32.5% versus10.3%). However, no direct comparison could be performedbecause no statistical comparison was possible betweenthe surgical and conservative treatment groups. The causesof death were not reported in 40% of the cases and thecause was cardiopulmonary (myocardial infarction, throm-boembolic complications, pneumopathies) in most reportedcases. This review also could not identify the risk factorsof mortality. Boakye claims that mortality is independent ofthe type of treatment and that age is the main risk factor[52]. For others age, a neurological deficit and concomitantinjuries may be risk factors of mortality and the type oftreatment alone was not the only cause of death. [13]. Nev-ertheless these conclusions were not statistically validated.

Conclusion

Traumatic injuries of the upper cervical spine are fre-quent in elderly subjects. The mechanism of injury is lowenergy trauma, which explains the delayed diagnosis of thisserious entity. There is no consensus on management. Con-servative and surgical treatments are both associated withcomplications and mortality whose risk factors have notbeen statistically validated in the literature. Immobiliza-tion by a neck brace or halo corset should be limited tostable lesions. Unstable fractures should be treated surgi-cally. Surgery seems to result in a better rate of union andfaster return to the pre-existing level of autonomy in theseelderly patients. Prospective descriptive studies are neededthat would make it possible to draft future guidelines forthe management of this complex, frequent, and sometimeslife threatening entity.

Disclosure of interest

The authors declare that they have no conflicts of interestconcerning this article.

References

[1] Wang H, Li C, Xiang Q, Xiong H, Zhou Y. Epidemiology ofspinal fractures among the elderly in Chongqing, China. Injury2012;43:2109—16.

[

S311

[2] Roche SJ, Sloane PA, McCabe JP. Epidemiology of spinetrauma in an Irish regional trauma unit: a 4-year study. Injury2008;39:436—42.

[3] Fassett DR, Harrop JS, Maltenfort M, Jeyamohan SB, Ratliff JD,Anderson DG, et al. Mortality rates in geriatric patients withspinal cord injuries. J Neurosurg Spine 2007;7:277—81.

[4] Watanabe M, Sakai D, Yamamoto Y, Sato M, Mochida J. Uppercervical spine injuries: age-specific clinical features. J OrthopSci 2010;15:485—92.

[5] Spivak JM, Weiss MA, Cotler JM, Call M. Cervical spine injuriesin patients 65 and older. Spine 1994;19:2302—6.

[6] Malik SA, Murphy M, Connolly P, O’Byrne J. Evaluation of mor-bidity, mortality and outcome following cervical spine injuriesin elderly patients. Eur Spine J 2008;17:585—91.

[7] Lomoschitz FM, Blackmore CC, Mirza SK, Mann FA. Cervicalspine injuries in patients 65 years old and older: epidemiologicanalysis regarding the effects of age and injury mechanismon distribution, type, and stability of injuries. AJR Am JRoentgenol 2002;178:573—7.

[8] Anderson LD, D’Alonzo RT. Fractures of the odontoid processof the axis. J Bone Joint Surg Am 1974;56:1663—74.

[9] Lakshmanan P, Jones A, Howes J, Lyons K. CT evaluation ofthe pattern of odontoid fractures in the elderly—relationship toupper cervical spine osteoarthritis. Eur Spine J 2005;14:78—83.

10] Harris MB, Reichmann WM, Bono CM, Bouchard K, Corbett KL,et al. Mortality in elderly patients after cervical spine frac-tures. J Bone Joint Surg Am 2010;92:567—74.

11] Carlisle E, Truumees E, Herkowitz H. Cervical spine traumain arthritic, stiff, or osteoporotic patients. Semin Spine Surg2005;17:100—5.

12] Van Middendorp JJ, Albert TJ, Veth RPH, Hosman AJF. Method-ological systematic review: mortality in elderly patients withcervical spine injury: a critical appraisal of the reporting ofbaseline characteristics, follow-up, cause of death, and anal-ysis of risk factors. Spine 2010;35:1079—87.

13] Abrahamsen B, Van Staa T, Ariely R, Olson M, Cooper C. Excessmortality following hip fracture: a systematic epidemiologicalreview. Osteoporos Int 2009;20:1633—50.

14] Moher D, Liberati A, Tetzlaff J, Altman DG. Preferred reportingitems for systematic reviews and meta-analyses: the PRISMAstatement. Int J Surg 2010;8:336—41.

15] Hanigan WC, Powell FC, Elwood PW, Henderson JP. Odontoidfractures in elderly patients. J Neurosurg 1993;78:32—5.

16] Berlemann U, Schwarzenbach O. Dens fractures in the elderly.Results of anterior screw fixation in 19 elderly patients. ActaOrthop Scand 1997;68:319—24.

17] Müller EJ, Wick M, Russe O, Muhr G. Management of odontoidfractures in the elderly. Eur Spine J 1999;8:360—5.

18] Andersson S, Rodrigues M, Olerud C. Odontoid fractures: highcomplication rate associated with anterior screw fixation in theelderly. Eur Spine J 2000;9:56—9.

19] Harrop JS, Przybylski GJ, Vaccaro AR, Yalamanchili K. Efficacyof anterior odontoid screw fixation in elderly patients with TypeII odontoid fractures. Neurosurg Focus 2000;8:e6.

20] Börm W, Kast E, Richter HP, Mohr K. Anterior screw fixationin type II odontoid fractures: is there a difference in outcomebetween age groups? Neurosurgery 2003;52:1089—92 [discus-sion 1092—4].

21] Tashjian RZ, Majercik S, Biffl WL, Palumbo MA, Cioffi WG. Halo-vest immobilization increases early morbidity and mortality inelderly odontoid fractures. J Trauma 2006;60:199—203.

22] Frangen TM, Zilkens C, Muhr G, Schinkel C. Odontoid frac-tures in the elderly: dorsal C1/C2 fusion is superior to halo-vestimmobilization. J Trauma 2007;63:83—9.

23] Platzer P, Thalhammer G, Ostermann R, Wieland T, VécseiV, Gaebler C. Anterior screw fixation of odontoid frac-tures comparing younger and elderly patients. Spine 2007;32:1714—20.

S

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

312

24] Collins I, Min WK. Anterior screw fixation of type II odontoidfractures in the elderly. J Trauma 2008;65:1083—7.

25] Koech F, Ackland HM, Varma DK, Williamson OD, Malham GM.Nonoperative management of type II odontoid fractures in theelderly. Spine 2008;33:2881—6.

26] Smith HE, Kerr SM, Maltenfort M, Chaudhry S, Norton R, AlbertTJ, et al. Early complications of surgical versus conservativetreatment of isolated type II odontoid fractures in octoge-narians: a retrospective cohort study. J Spinal Disord Tech2008;21:535—9.

27] Daentzer D, Flörkemeier T. Conservative treatment of uppercervical spine injuries with the halo vest: an appropriate optionfor all patients independent of their age? J Neurosurg Spine2009;10:543—50.

28] Omeis I, Duggal N, Rubano J, Cerabona F, Abrahams J, FinkM, et al. Surgical treatment of C2 fractures in the elderly:a multicenter retrospective analysis. J Spinal Disord Tech2009;22:91—5.

29] Lefranc M, Peltier J, Fichten A, Desenclos C, Toussaint P,Le Gars D. [Odontoid process fracture in elderly patientsover 70 years: morbidity, handicap, and role of surgical treat-ment in a retrospective series of 27 cases]. Neurochirurgie2009;55:543—50.

30] Chaudhary A, Drew B, Orr RD, Farrokhyar F. Management oftype II odontoid fractures in the geriatric population: outcomeof treatment in a rigid cervical orthosis. J Spinal Disord Tech2010;23:317—20.

31] Dailey AT, Hart D, Finn MA, Schmidt MH, Apfelbaum RI. Ante-rior fixation of odontoid fractures in an elderly population. JNeurosurg Spine 2010;12:1—8.

32] Hou Y, Yuan W, Wang X. Clinical evaluation of anterior screwfixation for elderly patients with type II odontoid fractures. JSpinal Disord Tech 2011;24:E75—81.

33] Mayer M, Zenner J, Auffarth A, Atzwanger J, Romeder F, HitzlW, et al. Efficacy of anterior odontoid screw fixation in theelderly patient: a CT-based biometrical analysis of odontoidfractures. Eur Spine J 2011;20:1441—9.

34] Molinari RW, Dahl J, Gruhn WL, Molinari WJ. Functionaloutcomes, morbidity, mortality, and fracture healing in26 consecutive geriatric odontoid fracture patients treatedwith posterior fusion. J Spinal Disord Tech 2013;26:119—26.

35] Osti M, Philipp H, Meusburger B, Benedetto KP. Analysis offailure following anterior screw fixation of Type II odontoidfractures in geriatric patients. Eur Spine J 2011;20:1915—20.

36] Schoenfeld AJ, Bono CM, Reichmann WM, Warholic N, Wood KB,Losina E, et al. Type II odontoid fractures of the cervical spine:do treatment type and medical comorbidities affect mortality

in elderly patients? Spine 2011;36:879—85.

37] Hénaux PL, Cueff F, Diabira S, Riffaud L, Hamlat A, Brassier G,et al. Anterior screw fixation of type IIB odontoid fractures inoctogenarians. Eur Spine J 2012;21:335—9.

P. Jubert et al.

38] Molinari RW, Khera OA, Gruhn WL, McAssey RW. Rigid cervicalcollar treatment for geriatric type II odontoid fractures. EurSpine J 2012;21:855—62.

39] Clement ND, Aitken S, Duckworth AD, McQueen MM, Court-Brown CM. Multiple fractures in the elderly. J Bone Joint SurgBr 2012;94:231—6.

40] Jacobs DG. Special considerations in geriatric injury. Curr OpinCrit Care 2003;9:535—9.

41] Longo UG, Denaro L, Campi S, Maffulli N, Denaro V. Upper cer-vical spine injuries: indications and limits of the conservativemanagement in Halo vest. A systematic review of efficacy andsafety. Injury 2010;41:1127—35.

42] Guyatt G, Gutterman D, Baumann MH, Addrizzo-Harris D, HylekEM, Phillips B, et al. Grading strength of recommendationsand quality of evidence in clinical guidelines: report froman American college of chest physicians task force. Chest2006;129:174—81.

43] Ersmark H, Dalen N, Kalen R. Cervical spine injuries: a follow-up of 332 patients. Paraplegia 1990;28:25—40.

44] Hoffman JR, Mower WR, Wolfson AB, Todd KH, Zucker MI,National Emergency X-Radiography Utilization Study Group.Validity of a set of clinical criteria to rule out injury to thecervical spine in patients with blunt trauma. N Engl J Med2000;343:94—9.

45] Stiell IG, Wells GA, Vandemheen KL, Clement CM, Lesiuk H, DeMaio VJ, et al. The Canadian C-spine rule for radiography inalert and stable trauma patients. JAMA 2001;286:1841—8.

46] Schrag SP, Toedter LJ, McQuay Jr N. Cervical spine fractures ingeriatric blunt trauma patients with low-energy mechanism:are clinical predictors adequate? Am J Surg 2008;195:170—3.

47] Dickinson G, Stiell IG, Schull M, Brison R, Clement CM, Van-demheen KL, et al. Retrospective application of the NEXUSlow-risk criteria for cervical spine radiography in Canadianemergency departments. Ann Emerg Med 2004;43:507—14.

48] Bass E, French DD, Bradham DD, Rubenstein LZ. Risk-adjustedmortality rates of elderly veterans with hip fractures. Ann Epi-demiol 2007;17:514—9.

49] Gubler KD, Davis R, Koepsell T, Soderberg R, Maier RV, RivaraFP. Long-term survival of elderly trauma patients. Arch Surg1997;132:1010—4.

50] Patel A, Smith HE, Radcliff K, Yadlapalli N, Vaccaro AR. Odon-toid fractures with neurologic deficit have higher mortality andmorbidity. Clin Orthop Relat Res 2012;470:1614—20.

51] Denaro V, Papalia R, Di Martino A, Denaro L, Maffulli N. Thebest surgical treatment for type II fractures of the dens is stillcontroversial. Clin Orthop Relat Res 2011;469:742—50.

52] Boakye M, Arrigo RT, Kalanithi PSA, Chen YR. Impact of age,

injury severity score, and medical comorbidities on earlycomplications after fusion and halo-vest immobilization for C2fractures in older adults: a propensity score matched retro-spective cohort study. Spine 2012;37:854—9.