Failures of cervical spine instrumentation may arise from an incomplete understanding of the biomechanics of the cervical spine, injury patterns, or instrumentation systems. Anterior, posterior, and combined approaches all have recognizable patterns of failure. Modern systems have evolved impressive user-friendliness and performance capabilities; however, they are never foolproof. Misunderstandings of biomechanical forces and overwhelming the capabilities of instrumentation is generally responsible for shorter term failure. Failure to achieve solid osseous union often underlies longer term complications.
Maintenance or restoration of physiologic cervical lordosis limits undue implant stresses.
Understanding limitation of instrumentation systems is crucial to their appropriate application.
Failure to achieve solid osseous union underlies longer term failures.
Many anterior and posterior systems are available for a wide variety of situations and can be used with a high degree of success.
Anterior instrumentation is designed to function as a graft buttress and to resist extension.
Posterior instrumentation relies on the structural integrity of posterior elements and resists flexion forces primarily, although some posterior constructs resist multiplanar forces.
Select implants neutralize or counteract the deforming force involved in a specific case.
Careful and proper arthrodesis technique is the most useful way to avoid pseudoarthrosis and resulting implant failure.
Cervical spinal instrumentation is designed to increase the rate of spinal fusion, maintain deformity correction, facilitate solid arthrodesis in the presence of high-grade instability or extensile neural element decompression, decrease or eliminate the need for external immobilization, and allow for early mobilization of a postoperative patient. In most cervical spine applications, instrumentation is a temporary adjunct to solid osseous union, the absence of which will lead to a high incidence of implant failure over time.
All discussions assume the proper technique of implantation, so as not to discuss the large number of complications related to improper implant placement, inadvertent damage to critical structures, or surgical exposure complications.
Since initial reports of anterior cervical spine surgery in the 1950s, a dramatic evolution of spinal implant in general and cervical spine implants specifically has occurred. Although these developments have brought increased user-friendliness, potential benefits to patient safety and modest improvement in biomechanical performance, these developments have also come with substantial increases in implant costs. Modern anterior plates have shown clear evidence of superiority over initial devices adapted from long-bone fracture care or uninstrumented fusion surgery. However, clinical results that would substantiate the preferred use of these later-generation devices over second- and third-generation devices remain incomplete.
Any treatise on failed treatment of any sort must first define the principles of successful treatment. A surgeon utilizing cervical spine instrumentation ideally should have extensive and detailed knowledge of the appropriate indications and intended use of instrumentation, understand its inherent limitations and potential failure modes, and be aware of bailout options. Most instrumentation failures can be differentiated into categories of deviation from these established principles such as failure to recognize unique circumstances where assumed principles do not apply or underestimation of the degree of instability that the construct is meant to resist. Intrinsic patient factors may play a role in spinal implant failure as well, but this entity should probably be considered as a matter of last resort.
Cervical spinal instrumentation consists of three categories of concepts, which are discussed briefly in the following sections.
The first and most common anterior device is the anterior plate, which has been well studied clinically and biomechanically. Because of its location anterior to the spine, this type of plate is designed to function as a buttress against anterior graft extrusion, to function as a tension band to resist extension forces, and depending on plate design, may provide resistance to axial loads. Lateral bending forces are resisted to an intermediate extent, but the location close to the axis of rotation provides little resistance to rotational forces. As Koh and colleagues showed in a traumatic biomechanical model, flexion loading of the cervical spine is relatively poorly resisted in the absence of posterior stabilization, either from an intact osseoligamentous structure or from an additional posterior procedure. This concept has been reflected in the findings of Johnson et al., who have reported a 13% failure rate in unilateral flexion distraction injuries treated anteriorly only. Integrity of the anterior column bony elements has a significant impact on the ability of an anterior plate and screw-based implant to resist deforming forces. The degree of instability imparted by the injury or surgical destabilization is highly predictive of failure rates of anterior stand-alone reconstructions as well, with greater failure rates seen with more severe injuries, multisegmental reconstructions, or in the presence of circumferential destabilization. An important limitation at the bone-screw interface is the type of cancellous bone purchase, with unicortical fixation in cancellous bone predisposing to failure by cutout compared with bicortical fixation.
The use of wires or cables attached to various structures of the posterior elements of the cervical spine has been the spinal implant with perhaps the longest track record. This type of instrumentation relies on structural integrity of posterior elements and primarily can resist flexion moments. Sublaminar wire or cable passage is associated with risk for neural element injury in up to 7% of cases. Such constructs impart little to no stability against extension moments, and lateral bend or rotational forces are resisted depending on how far from the midline the specific wiring technique chosen is anchored. Restricting fusions to single or short motion segment fusions is not always possible with cable or wire passing techniques because of some degree of violation of adjacent segments during surgical implant passage. Because of its relative biomechanical shortcomings, patients treated with cable or wiring only usually receive some form of rigid supplemental external immobilization. Nonetheless, clinical series in appropriate indications have shown equivalent results and lower costs compared with segmental fixation.
The second category of posterior devices, posterior screw-based techniques, have largely supplanted wiring in most applications because of their superior biomechanical properties, namely, the ability to resist forces in more than a single direction. Screw-rod and screw-plate constructs may be used in combination with either of the earlier-mentioned techniques or on their own. The technique of achieving posterior screw implantation varies with each cervical spine level, clinical circumstances, patients’ anatomy, and surgeon experience. Most commonly, lateral mass screws or pedicle screws are used with posterior segmental systems. These constructs have been shown to be associated with high union rates, have excellent biomechanical stiffness superior to wiring, and are equal or superior to anterior plating, especially for multilevel reconstructions. Even in cases of multilevel posterior element deficiencies and cervical spine deformities, posterior segmental rod-based systems allow for a safe and stable reconstruction. As in the anterior spine, simple unconstrained screw-plate designs have been largely supplanted by constrained systems, which offer improved construct stiffness. This can be of particular relevance for patients who have impaired poor-quality bone or segmental bone loss.
CERVICAL ALIGNMENT AND DEFORMING FORCES
Restoration of physiologic cervical lordosis is a critical component of surgical reconstruction of an impaired neck. The goal of achieving physiologic axial load distribution of 36% through the anterior column and 32% through either facet joint complex is contingent on achieving 15 to 25 degrees of cervical lordosis. This is a major challenge associated with long anterior reconstructions. Deviation from this distribution model—for example, by instrumenting a spine into kyphotic malalignment—may well overwhelm otherwise appropriate spinal instrumentation applied to either column alone.
Ideally, implants are chosen that neutralize or counteract the deforming forces acting on the cervical spine because of trauma or non–injury-related conditions. For instance, flexion-type injuries or multilevel anterior column deficiencies incurred in decompression surgery will be prone to sag into kyphosis. Consequently, constructs ideally are chosen to specifically resist these initially displacing forces. For example, anterior plates provide biomechanically inferior construct stiffness compared with posterior segmental fixation for three-column flexion-type injuries, as well as multilevel arthrodeses, because they are applied to resist compressive forces rather than tensile forces and are close to the center of sagittal rotation in the presence of a flexion moment arm. Posterior constructs are better situated to resist these forces because they are counteracting distraction of posterior elements and are placed farther away from the sagittal center of rotation. In contrast, extension injuries and single or short segment instability induced by surgical decompression may be effectively resisted by an anterior plate, which can be applied through a relatively atraumatic anterior approach. Attempts at long anterior column reconstruction with an anterior approach, however, may simply overwhelm the vertebral body purchase offered by the screws placed into the cancellous vertebral bodies.
ROLE OF FUSION
In any discussion on spinal implant failures, the relevance of proper arthrodesis technique cannot be overemphasized. Attention to detail in graft bed preparation, selection, and fitting of graft material, and proper deployment are crucial factors in the attempt at achieving the end result of a solid fusion of a physiologically realigned cervical spine. Excessive reliance on bone graft substitutes may lead to disappointing healing rates with the possibility for implant failure, even in the presence of satisfactory implant placement.
The most common cause of implant breakage is a failed fusion, for which a number of causes exist. For a metal structure to fail by fatigue, the anchor points must have maintained their integrity in the face of repetitive loads for a period of time, or else they would have failed by earlier loosening at the bone-screw interface. Any type of internal fixation, even very large plates or rods used in the femur or tibia in orthopedic fracture surgery, is bound to fail, given enough force and enough cycles in the absence of a solid osseous union. Thus, the timing of failure can provide important clues as to the cause. Early failure can be defined as occurring within the first 3 weeks from implantation. In these cases, usually a clear mismatch occurs between host and forces acting on the spinal column and the implant placed. Intermediate failures (>3 weeks to <1 year) occur around the time of expected bone healing and can be the result of biomechanical forces or delays in graft incorporation. Deep surgical infection of obvious or occult nature should be high on the differential diagnoses in such cases. Destructive nonunions will usually manifest with implant failures in this period. Late failures (>1 year) may reflect an occult nonunion or bone remodeling, as well as fatigue failure of implant subjected to notch creation during implantation.