methods of evaluating lumbar and cervical...

9
Review Article Methods of evaluating lumbar and cervical fusion Jordan A. Gruskay, BA, Matthew L. Webb, AB, Jonathan N. Grauer, MD* Department of Orthopaedics and Rehabilitation, Yale University School of Medicine, PO Box 208071, New Haven, CT 06520-8071, USA Received 10 September 2012; revised 12 June 2013; accepted 21 July 2013 Abstract Introduced in 1911, spinal fusion is now widely used to stabilize the cervical, thoracic, and lumbar spine. Despite advancements in surgical techniques, including the use of instrumentation and opti- mizing bone graft options, pseudarthrosis remains one of the most significant causes of clinical fail- ure following attempted fusion. Diagnosis of this common complication is based on a focused clinical assessment and imaging studies. Pseudarthrosis classically presents with the onset of or re- turn of axial or radicular symptoms during the first postoperative year. However, this diagnosis is complicated because other diagnoses can mimic these symptoms (such as infection or adjacent seg- ment degeneration) and because many cases of pseudarthrosis are asymptomatic. Computed tomog- raphy and assessment of motion on flexion/extension radiographs are the two preferred imaging modalities for establishing the diagnosis of pseudarthrosis. The purpose of this article was to review the current status of imaging and clinical practices for assessing fusion following spinal arthrode- sis. Ó 2014 Elsevier Inc. All rights reserved. Keywords: Pseudarthrosis; Cervical fusion; Lumbar fusion; Diagnostics Introduction It has been more than a century since Albee [1] and Hibbs [2] introduced spinal fusion in 1911 for the treatment of tuberculosis of the spine. Since then, it has become a commonly used procedure for the treatment of degenera- tive spinal conditions, as well as deformity, traumatic insta- bility, and a range of other spinal disorders. Despite its long history and widespread usage, failure of spinal arthrodesis or fusion resulting in pseudarthrosis is a common complica- tion following spine surgery. Literally meaning ‘‘false joint,’’ pseudarthrosis refers to a failure of osseous bridging at more than 1 year after surgery [3]. Although the use of instrumentation and opti- mizing bone graft options have improved fusion rates, pseudarthrosis remains a problem, with incidence rates reported in the literature ranging from 0% to 56%, varying with site, approach, fusion material, instrumentation, pa- tient factors, and year of the study [4–8]. However, as many patients with pseudarthrosis remain asymptomatic, the true incidence is likely underestimated by the literature. Many different imaging techniques are used for evalua- tion of pseudarthrosis, including static and dynamic plain ra- diographs, computed tomography (CT), magnetic resonance imaging (MRI), bone scintigraphy, and radiostereometric as- sessment (RSA) [9]. However, none is perfect, and no univer- sally agreed on imaging criterion exists. Interpreting fusion status is often difficult, and findings are skewed by surgeon bias. In assessing fusion on dynamic radiographs for anterior cervical discectomy and fusion (ACDF) patients 6 months postoperatively, Skolasky et al. [10] showed poor agreement (kappa50.308) between the treating surgeon and an indepen- dent review panel of peers blinded to the patient’s clinical sta- tus. Agreement was even worse when assessing a patient with improving clinical symptoms. Open surgical exploration remains the ‘‘gold standard’’ for the diagnosis of pseudarthrosis, but noninvasive methods of establishing the status of fusion are preferable and even open exploration is not always with definitive re- solve [11]. This makes interpretation of any study assessing union status challenging, as the absolute answer of union status is not always readily apparent. FDA device/drug status: Not applicable. Author disclosures: JAG: Nothing to disclose. MLW: Nothing to dis- close. JNG: Consulting: Harvard Clinical Research Institute (D), Affinergy (D), Alphatec (E), Depuy (C), KCI (B), Medtronic (B), Powered Research (A), Stryker (E), Vital 5 (B); Grants: Medtronic (F, Paid directly to insti- tution/employer), Smith and Nephew (None). The disclosure key can be found on the Table of Contents and at www. TheSpineJournalOnline.com. * Corresponding author. Department of Orthopaedics and Rehabilita- tion, Yale University School of Medicine, PO Box 208071, New Haven, CT 06520-8071, USA. Tel.: (203) 737-7463; fax: (203) 785-7132. E-mail address: [email protected] (J.N. Grauer) 1529-9430/$ - see front matter Ó 2014 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.spinee.2013.07.459 The Spine Journal 14 (2014) 531–539

Upload: others

Post on 12-Mar-2020

4 views

Category:

Documents


0 download

TRANSCRIPT

The Spine Journal 14 (2014) 531–539

Review Article

Methods of evaluating lumbar and cervical fusion

Jordan A. Gruskay, BA, Matthew L. Webb, AB, Jonathan N. Grauer, MD*Department of Orthopaedics and Rehabilitation, Yale University School of Medicine, PO Box 208071, New Haven, CT 06520-8071, USA

Received 10 September 2012; revised 12 June 2013; accepted 21 July 2013

Abstract Introduced in 1911, spinal fusion is now widely

FDA device/drug

Author disclosure

close. JNG: Consultin

(D), Alphatec (E), De

(A), Stryker (E), Vita

tution/employer), Smi

The disclosure key

TheSpineJournalOnlin

* Corresponding a

tion, Yale University

CT 06520-8071, USA

E-mail address: jo

1529-9430/$ - see fro

http://dx.doi.org/10.10

used to stabilize the cervical, thoracic, and lumbarspine. Despite advancements in surgical techniques, including the use of instrumentation and opti-mizing bone graft options, pseudarthrosis remains one of the most significant causes of clinical fail-ure following attempted fusion. Diagnosis of this common complication is based on a focusedclinical assessment and imaging studies. Pseudarthrosis classically presents with the onset of or re-turn of axial or radicular symptoms during the first postoperative year. However, this diagnosis iscomplicated because other diagnoses can mimic these symptoms (such as infection or adjacent seg-ment degeneration) and because many cases of pseudarthrosis are asymptomatic. Computed tomog-raphy and assessment of motion on flexion/extension radiographs are the two preferred imagingmodalities for establishing the diagnosis of pseudarthrosis. The purpose of this article was to reviewthe current status of imaging and clinical practices for assessing fusion following spinal arthrode-sis. � 2014 Elsevier Inc. All rights reserved.

Keywords: Pseudarthrosis; Cervical fusion; Lumbar fusion; Diagnostics

Introduction

It has been more than a century since Albee [1] andHibbs [2] introduced spinal fusion in 1911 for the treatmentof tuberculosis of the spine. Since then, it has becomea commonly used procedure for the treatment of degenera-tive spinal conditions, as well as deformity, traumatic insta-bility, and a range of other spinal disorders. Despite its longhistory and widespread usage, failure of spinal arthrodesisor fusion resulting in pseudarthrosis is a common complica-tion following spine surgery.

Literally meaning ‘‘false joint,’’ pseudarthrosis refersto a failure of osseous bridging at more than 1 year aftersurgery [3]. Although the use of instrumentation and opti-mizing bone graft options have improved fusion rates,pseudarthrosis remains a problem, with incidence rates

status: Not applicable.

s: JAG: Nothing to disclose. MLW: Nothing to dis-

g: Harvard Clinical Research Institute (D), Affinergy

puy (C), KCI (B), Medtronic (B), Powered Research

l 5 (B); Grants: Medtronic (F, Paid directly to insti-

th and Nephew (None).

can be found on the Table of Contents and at www.

e.com.

uthor. Department of Orthopaedics and Rehabilita-

School of Medicine, PO Box 208071, New Haven,

. Tel.: (203) 737-7463; fax: (203) 785-7132.

[email protected] (J.N. Grauer)

nt matter � 2014 Elsevier Inc. All rights reserved.

16/j.spinee.2013.07.459

reported in the literature ranging from 0% to 56%, varyingwith site, approach, fusion material, instrumentation, pa-tient factors, and year of the study [4–8]. However, asmany patients with pseudarthrosis remain asymptomatic,the true incidence is likely underestimated by theliterature.

Many different imaging techniques are used for evalua-tion of pseudarthrosis, including static and dynamic plain ra-diographs, computed tomography (CT), magnetic resonanceimaging (MRI), bone scintigraphy, and radiostereometric as-sessment (RSA) [9]. However, none is perfect, and no univer-sally agreed on imaging criterion exists. Interpreting fusionstatus is often difficult, and findings are skewed by surgeonbias. In assessing fusion on dynamic radiographs for anteriorcervical discectomy and fusion (ACDF) patients 6 monthspostoperatively, Skolasky et al. [10] showed poor agreement(kappa50.308) between the treating surgeon and an indepen-dent review panel of peers blinded to the patient’s clinical sta-tus. Agreement was evenworsewhen assessing a patient withimproving clinical symptoms.

Open surgical exploration remains the ‘‘gold standard’’for the diagnosis of pseudarthrosis, but noninvasivemethods of establishing the status of fusion are preferableand even open exploration is not always with definitive re-solve [11]. This makes interpretation of any study assessingunion status challenging, as the absolute answer of unionstatus is not always readily apparent.

532 J.A. Gruskay et al. / The Spine Journal 14 (2014) 531–539

There is a question of whether radiographic findings inan asymptomatic patient even matter. The diagnosis ofpseudarthrosis in the absence of symptoms is not typicallyby itself an indication for surgical revision. Several studieshave shown that in the short term (1–3 years) radiographicfindings of nonunion have no impact on patient outcomes[6,12]. However, on a long-term basis (5þ years), manypatients with radiographic pseudarthroses have inferior out-comes and require further intervention [13–15]. One canargue that this suggests that it is important to radiographi-cally screen patients, even asymptomatic ones, forpseudarthrosis.

Overall, it is routine practice to follow all fusion patientsfor several years after surgery. Follow-up visits often in-clude a focused history and physical, and plain radiographs.Depending on the findings elicited from the visit, furthertests and imaging may be considered.

Classification

Heggeness and Esses [16] classified patients with spinalpseudarthrosis according to the appearance of the attemptedfusion on imaging studies into four categories: atrophic,transverse, shingle, and complex (Table). The idea behindthis system was to allow for future analyses of different fu-sion techniques and the types of nonunions formed. How-ever, this classification system has yet to find a significantclinical application.

Clinical presentation

Asymptomatic pseudarthrosis has been reported in asmany as 50% of patients with pseudarthrosis, making diag-nosis based solely on clinical information difficult [12]. Forpatients presenting with symptoms, axial or radicular painis the most common complaint associated with pseudarth-rosis. Clauditory or myelopathic symptoms may be presentas well.

Classically, patients do quite well for some time aftergetting through the perioperative period. This period ofimprovement (referred to as the ‘‘honeymoon’’ period) isbelieved to be associated with the temporary stability pro-vided by instrumentation; however, without solid osseousunion, implant loosening and increased segmental motion

Table

Classification system for pseudoarthroses of the lumbar spine

Morphologic category Description

Atrophic Most severe case. Involves gross atrophy and resorptio

Complex Least common. Unique for the presence of more than

Shingle Substantial mass of matured bone graft is present, but

sagittal plane. Creates the impression of an onion-sk

Transverse Most common. Substantial mass of viable remodeled b

horizontal or transverse discontinuity.

Data from Heggeness MH, Esses SI. Classification of pseudarthroses of the

over months to years can lead to symptoms. In the caseof persistent pain without a ‘‘honeymoon’’ period or newneurologic symptoms soon after an attempted arthrodesis,alternative or concomitant pathology, such as infection,hardware issues, residual pathology, or adjacent segmentprocesses must also be considered.

Physical examination is often nonspecific, but can occa-sionally identify gross motion, surgical site infection, lossof alignment, or new neurologic symptoms. Tenderness orpain with motion at the affected segment is a positive butnot specific finding for nonunion. Patients with positive ex-amination findings and/or persistent pain and discomfortshould receive a more complete workup, including imagingstudies. Examination findings or symptoms alongside ra-diographic confirmation of nonunion may encourage a moreextensive evaluation.

Static radiographs

Plain radiographs are routinely obtained at follow-upvisits to evaluate fusion and instrumentation status overtime. To that end, 96% of surgeons take anteroposteriorand lateral films as part of routine follow-up after ACDF[17]. Although mainly used to rule out hardware failureor alignment-related issues, the study provides some infor-mation about the progression of arthrodesis. Progressiveconsolidation is determined by increased opacificationand potentially bridging trabecular bone at the margins ofthe graft. This should be seen between 6 and 12 weeks post-operatively and typically follows a brief period of resorp-tion in the early stages of healing (Fig. 1) [18].

The existence of any radiolucent line within the fusionmass is considered diagnostic of pseudarthrosis (Fig. 2,Left) [19–21]. Other clues suggesting nonunion includeprogressive settling or deformity (Fig. 3, Left). Radiolu-cency or haloing around implants is indicative of loosening(Fig. 3, Right), but generally better seen on CT [22]. De-layed failure of implants is suggestive of continued implantloading and pseudarthrosis.

In most cases, instrumentation makes assessment ofthese films difficult. In fact, plain films have been shownto correlate with surgical exploration only 43% to 82% ofthe time with a high rate of false-negative studies, makingplain film relatively insensitive in diagnosing pseudarthro-sis [7,9,21,23,24].

n of the bone graft.

one adjacent defect in the fusion mass.

a defect is present in the fusion mass, which passes obliquely through the

in construction of the fusion mass.

one, continuous with the fusion mass of adjacent levels that possesses a

lumbar spine. Spine 1991;16:S449–S454.

Fig. 1. Lateral radiograph of a patient after anterior lumbar interbody fu-

sion (ALIF) demonstrating bridging bone and an anterior sentinel sign

(arrow).

533J.A. Gruskay et al. / The Spine Journal 14 (2014) 531–539

Dynamic radiographs

Dynamic radiographs (lateral flexion and extension)may provide further useful information for evaluatingfusion status. As an example, nearly half of surgeons ob-tain flexion/extension films at routine follow-up afterACDF [17]. Assessment of fusion status using dynamic

Fig. 2. Neutral, flexion, and extension lateral films of a patient status post C45,

sistent with union, C56 appears consistent with pseudarthrosis. Neutral (Left), fle

tension films demonstrate greater than 2-mm increase in distance between spino

radiographs remains an evolving practice, and definitionsvary by approach over time and between studies.

The US Food and Drug Administration (FDA) has re-quired assessment of dynamic films for evaluation of newdevices. In the 2000 Guidance Document for Preparationof IDEs for Spinal System, the FDA set radiographic guide-lines for successful fusion as less than 3 mm translationalmotion and less than 5� angular motion for the lumbarspine. No specific measurements were given for assessingcervical fusion. However, there is controversy surroundinghow much motion and what type of motion between flexionand extension indicates nonunion, and the increasing use ofinstrumentation in fusion procedures means that lack ofmovement does not necessarily indicate the presence of os-seous union [3].

A standard method for obtaining, measuring, and as-sessing angles on dynamic radiographs has not been firmlyestablished. Motion can be assessed by the change in Cobbangle on flexion and extension. However, the motions seenare typically too small to be accurately detected via thismethod. Accordingly, some controversy exists as to the al-lowable degrees of motion to identify nonunion. For lum-bar interbody fusion, Kuslich et al. [25] proposed thecriterion of angular change of greater than 7� indicatingnonunion, change less than 3� indicating union, andchange of 3� to 7� indicating indeterminate fusion status.Other recommended criteria for motion indicating solid fu-sion have varied from 1� to 5� [26]. For cervical fusioncases, the criteria of an angle change greater than 2� indi-cating nonunion is commonly used [27–29]. Cannada et al.[29] reported a specificity of 39% and a sensitivity of 82%in their study. Interobserver agreement is often low in theassessment of dynamic films. A 2007 study by Taylor et al.[30] of cervical fusions demonstrated a kappa value of 0.17among three radiologists, three orthopedic surgeons, andone neurosurgeon.

C56 anterior cervical discectomy, and fusions (ACDF). C45 appears con-

xion (Middle), and extension (Right) films are shown. The flexion and ex-

us processes consistent with pseudarthrosis.

Fig. 3. Plain films of a patient with a C45 stand-alone fusion device above a prior C56 anterior cervical discectomy, and fusions (ACDF) (anteroposterior

film [Left], and lateral film [Right]). Although the C56 appears consistent with fusion (integrated bone on the lateral), C45 appears suggestive of nonunion

based on settling and haloing seen around the anterior plate portion of the construct at that level (arrows).

534 J.A. Gruskay et al. / The Spine Journal 14 (2014) 531–539

Based on the limitation of using Cobb angle for assess-ing union status, Cannada et al. [29] proposed a method ofmeasuring the change in distance between any two spinousprocesses on cervical flexion and extension. With a greaterthan 2 mm change indicating nonunion, they reported 89%specificity and 91% sensitivity in diagnosing pseudarthrosiswith an interobserver reliability of a50.95 [29] (Fig. 2,Middle and Right).

Computerized quantitative motion analysis methodshave been developed for evaluating the presence of solidfusion in flexion/extension films using the previously men-tioned criteria. These specialized computer programs usepattern-matching software to assess flexion/extension filmswith little input from the user. This technology has beenfound to greatly improve agreement between observersand accuracy of the assessment [28,30,31]. In 2011, usingcomputer technology, Ghiselli et al. [28] tested various cri-teria to identify nonunion and found that angular change ofless than 1� most accurately indicates solid fusion in thecervical spine when compared with open exploration. Bonoet al. [32] used a complex model of simulated lumbar mo-tion and, despite wide variation due to technique, deter-mined that less than 4.1� of motion was most predictiveof fusion in lumbar cases.

Overall, for angular motion assessment, computerizedanalysis of plain films has clearly been shown to be optimalfor assessing plain films, if available. Alternatively, measur-ing the change in distance between spinous processes isa more reliable indicator of fusion status than Cobb anglemeasurements and has been recommended for assessingfor pseudarthrosis [33].

Although often discussed as a method of assessing fu-sion, sagittal plane motion appears to be less sensitive thanangular motion assessment. If used criteria indicating

nonunion varying from 2 to 5 mm of motion have been sug-gested [34,35].

Recently, the utility of routine postoperative plain filmsin asymptomatic patients has been questioned, as the prac-tice exposes patients to unneeded radiation, increases costs,and rarely guides treatment course [36–38]. One study re-ported that radiographs alone led to a change in treatmentin only 0.9% (6 of 659) of asymptomatic office visits,whereas costing more than $35,000 to perform [36]. Onthe flip side, the potential risk of missing important pathol-ogy in even a small percentage of patients may be too great,especially in an increasingly litigious health-care environ-ment. And although plain radiographs do have low specific-ity, their high relative sensitivity and low cost make theman effective and important screening test [7,29].

Computed tomography

CT is now considered the ‘‘gold standard’’ imaging testfor assessment of fusion status and evaluating for bridgingbone following equivocal plain radiographic findings [28].There are some reports in the literature to perform CT scansat 3, 6, 12, and 24 months after fusion or until solid arthrod-esis is seen [39], but most surgeons obtain scans less fre-quently or not at all unless pseudarthrosis is clinicallysuspected [17].

Indicators of fusion by CT include the presence of bonytrabeculation across the fusion level with a lack of bonylucency at the graft/vertebral body junction. Device subsi-dence, cystic changes on the endplates, and haloing sur-rounding instrumentation (indicating loosening) (Fig. 3,Middle) are important signs of nonunion [39]. CT moreclearly demonstrates the existence or absence of bridging

Fig. 4. Computed tomography evidence of pseudarthrosis. (Left) A radiolucent line on a mid-sagittal image after anterior cervical discectomy, and fusions

(ACDF) (arrow). (Right) Haloing around lumbar pedicle screws on an axial image after posterior instrumentation construct.

535J.A. Gruskay et al. / The Spine Journal 14 (2014) 531–539

bone, making bony nonunion easier to visualize than onplain radiographs (Figs. 4 and 5, Left).

CT has been found to correlate better with surgical explo-ration than plain films in numerous studies [4,7,40–47]. Intheir 2008 study of anterior cervical fusion, Buchowskiet al. [24] found that CT findings agreed with intraoperativefindings in 78.6% to 85.7% of cases, with a Kappa statistic(measure of interrater agreement) of 0.81 compared withplain radiography, which had a 71.4% to 92.8% agreementand a Kappa statistic of only 0.67. These results were consis-tent with a 2007 study by Carreon et al. [11] that reported74% to 96% accuracy of CT following posterolateral lumbarinstrumented fusion.

Despite its strengths, CT is interpreted subjectively andis vulnerable to both type I and type II errors. Comparedwith computerized motion analysis of flexion-extensioncervical radiographs with a nonunion defined as greaterthan 1� motion, CT had the same sensitivity and positivepredictive value, and a slightly less negative predictive

Fig. 5. Nonunion after cervical anterior cervical discectomy, and fusions (ACD

raphy reconstruction (Left). (Middle and Right) Magnetic resonance imaging findi

in the subchondral marrow of adjacent vertebrae in T2-weighted images (Middl

value [28]. Although CT is clearly a more accurate imagingmodality, when advanced imaging analysis software thatcan detect minute motions is applied, radiographs can pro-vide approximating information regarding pseudarthrosis[3].

Artifact from metallic implants can be an issue with anyadvanced imaging modality. With CT, scatter from metallicimplants can obscure some findings, but later-generation ti-tanium implants have significantly less artifact than stain-less steel implants [39]. Clearly CT has markedly lessscatter than MRI. In situations of prior surgery when poten-tial neural element compression needs to be assessed, CTmyelogram has potential distinct advantages of assessingfusion, implants, and stenosis.

A limitation of CT is the radiation exposure and associ-ated costs. These factors are not to be taken lightly in an erain which we are beginning to appreciate the significant ef-fective doses of radiation imparted by these medical proce-dures. One study estimated that the effective dose of

F). An absence of bridging bone is seen on mid-sagittal computed tomog-

ngs classically consistent with pseudarthrosis: high signal intensity changes

e), and low signal intensity changes in T1-weighted images (Right).

Fig. 6. Decision-making algorithm for assessing fusion status. A plus (þ) sign represents a positive finding and a negative (�) sign represents a negative

finding. CBC, complete blood count; CRP, C-reactive protein; CT, computed tomography; ESR, erythrocyte sedimentation rate; MRI, magnetic resonance

imaging.

536 J.A. Gruskay et al. / The Spine Journal 14 (2014) 531–539

radiation associated with a CT of the cervical spine isequivalent to approximately 55 chest radiographs, a CT ofthe thoracic spine is equivalent to approximately 225 chestradiographs, and a CT of the lumbar spine is equivalent toapproximately 240 chest radiographs [48]. Another analysisfound CTs of both the cervical and lumbar spine to be as-sociated with on average 6 mSv of radiation, or the equiv-alent of 300 posteroanterior chest radiographs (0.02 mSv)[49]. The risks and benefits of any such study must be con-sidered when using such studies.

Magnetic resonance imaging

MRI is less accurate than CT and not routinely used inassessing fusion status because it is highly susceptible to ar-tifact from metallic implants and less useful for bony as-sessment [3,24,45,50]. Most of the time when MRI is

used in the assessment of a patient with suspected pseu-darthrosis, it is to delineate potential residual, recurrent,or new stenosis that might need attention if revision surgeryis considered. MRI has the additional utility of assessingadjacent pathology that could be the cause of clinical symp-toms of pseudarthrosis.

That said, low signal intensity changes in the subchon-dral marrow of adjacent fused vertebrae in T2-weighted im-ages and high signal intensity changes in T1-weightedimages are suggestive of fusion. Conversely, high signal in-tensity changes on T2-weighted images and low signal in-tensity changes on T1-weighted images are suggestive ofnonunion [22] (Fig. 5, Left and Right). Of further note, fastspin-echo sequences are often helpful due to the minimizedmetallic artifact.

Buchowski et al. [24] confirmed this limitation in 2008when they reported 66.7% average agreement betweenMRI and intraoperative findings, with a Kappa statistic of

537J.A. Gruskay et al. / The Spine Journal 14 (2014) 531–539

0.48. Another study, however, found a Kappa statistic of0.88 when carbon fiber cages were used for posterior lum-bar interbody fusion supplemented with posterior instru-mentation [51] and there are several such compositedevices in use today that produce fewer artifacts on MRIsthan conventional metallic devices [39].

Overall, MRI is inferior to other imaging modalities inthe assessment of fusion. This is not the study of choicefor such assessment.

Other imaging techniques

Bone scan

Bone scintigraphy using technetium Tc-99m with singlephoton emission computer tomography (SPECT) can pro-vide information on the current metabolic activity of thespine that would be suggestive of nonunion. Increased bio-logical activity and blood supply in areas of healing causemore radiomarker to be absorbed, which is reflected on thescan.

Of note, because nonunion is not considered until 1 yearafter surgery, increased uptake in scans done before thistime is not necessarily suggestive of pseudarthrosis. Usingthis modality 6 months after index surgery has been associ-ated with a 50% false-positive rate [52]. In scans done after1 year, increased signal is thought to indicate continuedbone activity and is thus suggestive of nonunion.

Unfortunately, this modality has been shown to be oflimited utility in diagnosing pseudarthrosis at this timepoint as well. In their 1998 study, Albert et al. [53] reportedthat SPECT has a sensitivity of 50% and a specificity of58% (false-positive rate of 42%) in diagnosing nonunionwhen compared with open surgical exploration.

Ultrasound

Ultrasonography has also been suggested to be of po-tential utility in the evaluation of pseudarthrosis [54]. Ina small study of 10 patients by Jacobson et al. [54] com-paring ultrasound results with findings at revision surgeryfor posterolateral fusion with instrumentation, all surgi-cally determined nonunions were correctly identified pre-operatively by ultrasound. Fusion was determined by theexistence of a hyperechoic, shadowing interface (bonemass) that bridged between contiguous vertebral seg-ments. Meanwhile, pseudarthrosis was suspected in casesin which there was no evidence of bridging interface, or inthe presence of scattered and nonbridging echogenic fociat the fusion site.

This technique seems to have promise, especially for pa-tients whose other imaging studies might be obscured by me-tallic implants. Of course, because the ultrasound is able tovisualize only posterior elements of the spinal column, thismodality would likely have limited usage for assessing fu-sion methods using nonposterior approaches. Unfortunately,

no subsequent studies have appeared in the literature to ex-pand on the previous findings.

Radiostereometric analysis

RSA provides three-dimensional imaging of spinal mo-tion in vivo [55,56]. At the index surgery, small metallicbeads are implanted on the spine above and below a levelof interest. Orthogonal films are later taken with dynamicradiographs and computer-assisted motion analysis, allow-ing for a three-dimensional reconstruction of spinal motion.

For the most part, despite its high reported accuracy, thismethod has tended to be used for research more than inclinical practice. Nonetheless, this modality has shownpromise in conjunction with plain radiographs in assessingfusion after ACDF [57,58] and lumbar fusion [55].

Decision-making algorithm

Because of the lack of an ideal imaging modality to diag-nose pseudarthrosis, multiple imaging studies are typicallyused. Decision-making algorithms have been proposed inthe literature previously [18]. The algorithm that we havedeveloped (shown in Fig. 6) should be useful when decidingon a strategy to assess for fusion status.

Conclusion

Diagnosis of pseudarthrosis following cervical and lum-bar spine surgery remains a challenge. Typically, clinicalpresentation in conjunction with various imaging studieshave been used to make a diagnosis, but no definitivemethod exists for assessing nonunion.

Based on the data available, thin-cut CT scan and com-puterized motion analysis of dynamic plain films are thebest imaging modalities, whereas surgical exploration re-mains the gold standard. Nonetheless, the delineation of fu-sion versus pseudarthrosis remains challenging. This is ofsignificance in research and the clinical setting.

References

[1] Albee FH. Transplantation of a portion of the tibia into the spine for

Pott’s disease: a preliminary report 1911. Clin Orthop Relat Res

2007;460:14–6.

[2] Hibbs RA. An operation for progressive spinal deformities: a prelim-

inary report of three cases from the service of the orthopaedic

hospital. 1911. Clin Orthop Relat Res 2007;460:17–20.

[3] Raizman NM, O’Brien JR, Poehling-Monaghan KL, Yu WD. Pseu-

darthrosis of the spine. J Am Acad Orthop Surg 2009;17:494–503.

[4] Rothman SL, Glenn WV Jr. CT evaluation of interbody fusion. Clin

Orthop Relat Res 1985;47–56.

[5] Kaiser MG, Mummaneni PV, Matz PG, et al. Management of ante-

rior cervical pseudarthrosis. J Neurosurg Spine 2009;11:228–37.

[6] Herkowitz HN, Kurz LT. Degenerative lumbar spondylolisthesis

with spinal stenosis. A prospective study comparing decompression

with decompression and intertransverse process arthrodesis. J Bone

Joint Surg Am 1991;73:802–8.

538 J.A. Gruskay et al. / The Spine Journal 14 (2014) 531–539

[7] Brodsky AE, Kovalsky ES, Khalil MA. Correlation of radiologic as-

sessment of lumbar spine fusions with surgical exploration. Spine

1991;16:S261–5.

[8] Farey ID, McAfee PC, Davis RF, Long DM. Pseudarthrosis of the

cervical spine after anterior arthrodesis. treatment by posterior

nerve-root decompression, stabilization, and arthrodesis. J Bone

Joint Surg Am 1990;72:1171–7.

[9] Sugiyama S, Wullschleger M, Wilson K, et al. Reliability of clinical

measurement for assessing spinal fusion: an experimental sheep

study. Spine 2012;37:763–8.

[10] Skolasky RL, Maggard AM, Hilibrand AS, et al. Agreement be-

tween surgeons and an independent panel with respect to surgical

site fusion after single-level anterior cervical spine surgery: a pro-

spective, multicenter study. Spine 2006;31:E503–6.

[11] Carreon LY, Djurasovic M, Glassman SD, Sailer P. Diagnostic

accuracy and reliability of fine-cut CT scans with reconstructions

to determine the status of an instrumented posterolateral fusion

with surgical exploration as reference standard. Spine 2007;32:

892–5.

[12] Fischgrund JS, Mackay M, Herkowitz HN, et al. 1997 Volvo award

winner in clinical studies. degenerative lumbar spondylolisthesis

with spinal stenosis: a prospective, randomized study comparing de-

compressive laminectomy and arthrodesis with and without spinal

instrumentation. Spine 1997;22:2807–12.

[13] Kornblum MB, Fischgrund JS, Herkowitz HN, et al. Degenerative

lumbar spondylolisthesis with spinal stenosis: a prospective long-

term study comparing fusion and pseudarthrosis. Spine 2004;29:

726–33; discussion 733–4.

[14] Phillips FM, Carlson G, Emery SE, Bohlman HH. Anterior cervical

pseudarthrosis. Natural history and treatment. Spine 1997;22:

1585–9.

[15] Tsutsumimoto T, Shimogata M, Yoshimura Y, Misawa H. Union

versus nonunion after posterolateral lumbar fusion: a comparison

of long-term surgical outcomes in patients with degenerative lumbar

spondylolisthesis. Eur Spine J 2008;17:1107–12.

[16] Heggeness MH, Esses SI. Classification of pseudarthroses of the

lumbar spine. Spine 1991;16:S449–54.

[17] Bohl DD, Hustedt JW, Blizzard DJ, et al. Routine imaging for ante-

rior cervical decompression and fusion procedures: a survey study

establishing current practice patterns. Orthopedics 2012;35:

e1068–72.

[18] Hilibrand AS, Dina TS. The use of diagnostic imaging to assess spi-

nal arthrodesis. Orthop Clin North Am 1998;29:591–601.

[19] Lauerman WC, Bradford DS, Transfeldt EE, Ogilvie JW. Manage-

ment of pseudarthrosis after arthrodesis of the spine for idiopathic

scoliosis. J Bone Joint Surg Am 1991;73:222–36.

[20] Larsen JM, Rimoldi RL, Capen DA, et al. Assessment of pseudarth-

rosis in pedicle screw fusion: a prospective study comparing plain

radiographs, flexion/extension radiographs, CT scanning, and bone

scintigraphy with operative findings. J Spinal Disord 1996;9:

117–20.

[21] Kant AP, Daum WJ, Dean SM, Uchida T. Evaluation of lumbar

spine fusion. Plain radiographs versus direct surgical exploration

and observation. Spine 1995;20:2313–7.

[22] Lee C, Dorcil J, Radomisli TE. Nonunion of the spine: a review.

Clin Orthop Relat Res 2004:71–5.

[23] Albert TJ, Pinto M, Denis F. Management of symptomatic lum-

bar pseudarthrosis with anteroposterior fusion. A functional and

radiographic outcome study. Spine 2000;25:123–9; discussion

130.

[24] Buchowski JM, Liu G, Bunmaprasert T, et al. Anterior cervical fu-

sion assessment: surgical exploration versus radiographic evalua-

tion. Spine 2008;33:1185–91.

[25] Kuslich SD, Ulstrom CL, Griffith SL, et al. The Bagby and Kuslich

method of lumbar interbody fusion. History, techniques, and 2-year

follow-up results of a United States prospective, multicenter trial.

Spine 1998;23:1267–78; discussion 1279.

[26] McAfee PC, Boden SD, Brantigan JW, et al. Symposium: a critical

discrepancy—a criteria of successful arthrodesis following inter-

body spinal fusions. Spine 2001;26:320–34.

[27] Lee CS, Chung SS, Choi SW, et al. Critical length of fusion requir-

ing additional fixation to prevent nonunion of the lumbosacral junc-

tion. Spine 2010;35:E206–11.

[28] Ghiselli G, Wharton N, Hipp JA, et al. Prospective analysis of im-

aging prediction of pseudarthrosis after anterior cervical discectomy

and fusion: computed tomography versus flexion-extension motion

analysis with intraoperative correlation. Spine 2011;36:463–8.

[29] Cannada LK, Scherping SC, Yoo JU, et al. Pseudoarthrosis of the

cervical spine: a comparison of radiographic diagnostic measures.

Spine 2003;28:46–51.

[30] Taylor M, Hipp JA, Gertzbein SD, et al. Observer agreement in as-

sessing flexion-extension x-rays of the cervical spine, with and with-

out the use of quantitative measurements of intervertebral motion.

Spine J 2007;7:654–8.

[31] Goldberg G, Albert TJ, Vaccaro AR, et al. Short-term comparison of

cervical fusion with static and dynamic plating using computerized

motion analysis. Spine 2007;32:E371–5.

[32] Bono CM, Khandha A, Vadapalli S, et al. Residual sagittal motion

after lumbar fusion: a finite element analysis with implications on

radiographic flexion-extension criteria. Spine 2007;32:417–22.

[33] Kaiser MG, Mummaneni PV, Matz PG, et al. Radiographic assess-

ment of cervical subaxial fusion. J Neurosurg Spine 2009;11:221–7.

[34] Burkus JK, Foley K, Haid RW, LeHuec JC. Surgical interbody re-

search group—radiographic assessment of interbody fusion devices:

fusion criteria for anterior lumbar interbody surgery. Neurosurg Fo-

cus 2001;10:E11.

[35] Kanayama M, Hashimoto T, Shigenobu K, et al. A prospective ran-

domized study of posterolateral lumbar fusion using osteogenic

protein-1 (OP-1) versus local autograft with ceramic bone substi-

tute: emphasis of surgical exploration and histologic assessment.

Spine 2006;31:1067–74.

[36] Shau DN, Bible JE, Samade R, et al. Utility of post-operative radio-

graphs for cervical spine fusion: a comprehensive evaluation of op-

erative technique, surgical indication, and duration since surgery.

Spine 2012;37:1994–2000.

[37] Ugokwe KT, Kalfas IH, Mroz TE, Steinmetz MP. A review of the

utility of obtaining repeated postoperative radiographs following

single-level anterior cervical decompression, fusion, and plate

placement. J Neurosurg Spine 2008;9:175–9.

[38] Bartels RH, Beems T, Schutte PJ, Verbeek AL. The rationale of

postoperative radiographs after cervical anterior discectomy with

stand-alone cage for radicular pain. J Neurosurg Spine 2010;12:

275–9.

[39] Williams AL, Gornet MF, Burkus JK. CT evaluation of lumbar in-

terbody fusion: current concepts. AJNR Am J Neuroradiol

2005;26:2057–66.

[40] Pai VS, Hodgson B. Assessment of bony union following surgical

stabilisation for lumbar spondylolysis: a comparative study between

radiography and computed tomography. J Orthop Surg (Hong Kong)

2006;14:17–20.

[41] Ploumis A, Mehbod A, Garvey T, et al. Prospective assessment of

cervical fusion status: plain radiographs versus CT-scan. Acta Or-

thop Belg 2006;72:342–6.

[42] Lang P, Genant HK, Steiger P, et al. 3-dimensional computed to-

mography and multiplanar CT-reformations in lumbar spondylode-

sis. Rofo 1988;148:524–9.

[43] Shah RR, Mohammed S, Saifuddin A, Taylor BA. Comparison of

plain radiographs with CT scan to evaluate interbody fusion follow-

ing the use of titanium interbody cages and transpedicular instru-

mentation. Eur Spine J 2003;12:378–85.

[44] Epstein NE, Silvergleide RS. Documenting fusion following ante-

rior cervical surgery: a comparison of roentgenogram versus two-

dimensional computed tomographic findings. J Spinal Disord Tech

2003;16:243–7.

539J.A. Gruskay et al. / The Spine Journal 14 (2014) 531–539

[45] Epstein NE, Silvergleide RS, Black K. Computed tomography

validating bony ingrowth into fibula strut allograft: a criterion for

fusion. Spine J 2002;2:129–33.

[46] Zinreich SJ, Long DM, Davis R, et al. Three-dimensional CT imag-

ing in postsurgical ‘‘failed back’’ syndrome. J Comput Assist To-

mogr 1990;14:574–80.

[47] Siambanes D, Mather S. Comparison of plain radiographs and CT

scans in instrumented posterior lumbar interbody fusion. Orthope-

dics 1998;21:165–7.

[48] Biswas D, Bible JE, Bohan M, et al. Radiation exposure from mus-

culoskeletal computerized tomographic scans. J Bone Joint Surg

Am 2009;91:1882–9.

[49] Fazel R, Krumholz HM, Wang Y, et al. Exposure to low-dose ion-

izing radiation from medical imaging procedures. N Engl J Med

2009;361:849–57.

[50] Sutter B, Friehs G, Pendl G, Tolly E. Bovine dowels for anterior

cervical fusion: experience in 66 patients with a note on postoper-

ative CT and MRI appearance. Acta Neurochir (Wien) 1995;137:

192–8.

[51] Kroner AH, Eyb R, Lange A, et al. Magnetic resonance imaging

evaluation of posterior lumbar interbody fusion. Spine 2006;31:

1365–71.

[52] Lin PM. Posterior lumbar interbody fusion technique: Complica-

tions and pitfalls. Clin Orthop Relat Res 1985:90–102.

[53] Albert TJ, Pinto M, Smith MD, et al. Accuracy of SPECT scanning

in diagnosing pseudoarthrosis: a prospective study. J Spinal Disord

1998;11:197–9.

[54] Jacobson JA, Starok M, Pathria MN, Garfin SR. Pseudarthrosis: US

evaluation after posterolateral spinal fusion: work in progress. Radi-

ology 1997;204:853–8.

[55] Johnsson R, Stromqvist B, Aspenberg P. Randomized radiostereo-

metric study comparing osteogenic protein-1 (BMP-7) and autograft

bone in human noninstrumented posterolateral lumbar fusion: 2002

Volvo award in clinical studies. Spine 2002;27:2654–61.

[56] Johnsson R, Selvik G, Stromqvist B, Sunden G. Mobility of the

lower lumbar spine after posterolateral fusion determined by roent-

gen stereophotogrammetric analysis. Spine 1990;15:347–50.

[57] Nabhan A, Pape D, Pitzen T, et al. Radiographic analysis of fusion

progression following one-level cervical fusion with or without

plate fixation. Zentralbl Neurochir 2007;68:133–8.

[58] Park SA, Fayyazi AH, Ordway NR, et al. Correlation of radioster-

eometric measured cervical range of motion with clinical radio-

graphic findings after anterior cervical discectomy and fusion.

Spine 2009;34:680–6.