Some surgeons believe in strict adherence to a relatively short list of indications for total disc replacement (TDR). Singh et al.¹ proposed that if TDR use follows the pattern of pedicle screws and cylindrical cages, rapid growth will be followed by expanded indications, resulting in inconsistent clinical outcomes and increased complications, leading to curtailed use. If TDR is used in ‘properly indicated patients’, this rise and fall in use may be avoided. The appropriate indication for surgery is critical for success.² The first purpose of this review is to evaluate information regarding general indications for TDR. The second purpose is to develop an overview for determining specific indications for different TDRs.

Lumbar fusion is used to treat pain attributed to abnormal motion or mechanical insufficiency produced by degenerative change.³ Often compressed nerves are released simultaneously to reduce back and leg pain. In the last 40 years, indications for fusion have remained largely constant. Described indications include: degenerative disc disease (DDD), isthmic spondylolisthesis, unstable spinal stenosis, degenerative spondylolisthesis, degenerative scoliosis, segmental instability, disc-related back pain, failed previous surgery, and post-facetectomy syndrome.^{3, 4} Bambakidis et.al.⁴ define radiographic instability of a motion segment as translational motion of >3 mm in levels above L5-S1 or 5 mm at L5-S1, or motion segment angulation of >10° on lateral flexion/extension radiographs.

Some fusion indications have been found to be appropriate for lumbar TDR, as demonstrated in randomized Food and Drug Administration Investigational Device Exemption (FDA IDE) studies with up to five-year follow-up.^{5, 6} Wong et al.⁷ proposed that the ideal TDR patient is likely earlier in the Kirkaldy-Willis degenerative cascade than a fusion patient. The primary indication for lumbar TDR is symptomatic DDD.^7–29 In defining general indications, data needed includes patient history, pain and disability, clinical and image findings, diagnostic procedures, and psychosocial factors. TDR indications/contraindications literature is summarized below.

Siepe et al.²³ found that ProDisc at L4-L5 resulted in better outcomes and higher patient satisfaction than use at L5-S1; however, L5-S1 patients had lower complication and reoperation rates. Some studies evaluated single-level TDR only at L4-L5 or L5-S1,^{9, 18, 21, 22, 28, 33, 35, 36} others included two-level TDR from L3 to S1,^{8, 10, 11, 13, 15, 17, 29, 31, 32, 37, 41} and some included three-level²⁶ and four-level^{12, 44} TDR. Tropiano et al.^{25, 26} found TDR could be used successfully at 2 or 3 contiguous levels. Zigler et al.⁶⁴ found no differences in 1-vs. 2-level outcomes, with both groups improving significantly. Siepe et al.²³ reported better outcomes for single-level than for 2-level procedures, associated with greater complication and reoperation rates. Siepe et al.²⁴ found bisegmental TDR results deteriorated at 12 and 24 months compared with the monosegmental outcomes. Patient satisfaction rates were 85.7% for mono- and 64.3% for bi-segmental TDR. Chin³¹ found more favorable results in patients with isolated disc disease compared with multi-level disease.

In studies reviewing large series of fusion cases to determine how many would have been TDR candidates, the figure was <10%.^{7, 31} Numerous contraindications to TDR are cited in the literature (Table 1). Chin et al.³¹ and Huang et al.⁶⁵ considered contraindications to fall under two broad categories: 1) Painful conditions not caused by the disc, and 2) Conditions that may compromise long-term device functionality.

There are several widely described general TDR indications including painful DDD unresponsive to >6 months of nonoperative care, no significant facet joint degeneration, no osteopenia/osteoporosis, and lack of conditions that may compromise outcome and/or interfere with proper TDR functioning such as severe instability. They are mainly base on FDA-study-related parameters and surgeons’ experiences. To date, there is nothing published about potentially different indications for different TDR types with regard to design or materials.

Lemaire et al.² proposed that “disc prosthesis is indicated particularly in situations where restoration of a center of rotation and redefinition of segmental kinematics are required.” But TDR designs simulate different disc functions. Inherent in current designs are disc height restoration, intervertebral angle, and varying degrees of motion and stability.

TDR designs include functional two- and three-component ball and socket variations with gliding surfaces and 1-piece designs consisting of multiple components bonded together or one compact component (Table 2). These designs inherently have different biomechanical characteristics, leading to advantages and disadvantages. Spherical ball and socket designs provide axial rotational without limitation, but with risk of damaging surrounding anatomy, namely facet joints. This design does not provide load sharing in axial compression. Nevertheless spherical ball and socket discs with polyethylene have minimal elastic features with potential positive influence on axial load distribution at the vertebral endplates. Spherical ball and socket discs are generally metal-on-metal, metal-on-polyethylene, or PEEK-on-PEEK. Patients with metal-on-metal devices may have increased metallic ion levels from the implant,^66–68 a phenomenon much more documented in hip implants.^71–73 One TDR study found some of the serum ion levels to be greater than reported for hip replacements;⁶⁶ while other studies found the levels to be similar for hip replacement⁶⁷ and below levels determined high enough to merit monitoring of hip replacement patients68 as described by the Medicines and Healthcare Products Regulatory Agency⁶⁹ or in more recent literature.⁷⁰ Some metal-on-polyethylene designs may become impinged on one area of the core⁷⁴ or on an area of the metal plates.^{75, 76} Deformation and failure of polyethylene cores have occurred, often when devices were inaccurately placed or when inappropriate prosthetic components were used.

One-piece discs consist of combinations of metal and polymers or graduated modulus elastomers, where simultaneous injection of polymers with different moduli (stiffness) provides a dual modulus disc with a graduated modulus region. Designs with one or more components implanted as one piece can potentially provide performance characteristics most like those of a healthy disc because they are viscoelastic.⁷⁷ They can simulate a disc to a large extent, but with equal and reduced^{19, 77} ROM in every direction due to the homogeneous material. The natural intervertebral disc needs dampening properties for producing motion, because it does not have typical joint surfaces.

To date, there is no research addressing whether a particular TDR design may be better for some patients than for others. Kinematic studies may be needed to determine if motion pattern or other characteristics afforded by a particular design may best address specific needs of an individual patient.

There is no guide to “determine the right patient with the right indication for the right TDR.” More is known about the “right patient” than the “right total disc,” nearly nothing about the “right TDR type.” Bertagnoli et al.⁷⁸ correlated surgical outcome with indications and categorized prime, good, borderline, and poor indications based on combinations of the number of levels operated, disc space height, and the condition of facet joints and adjacent segments. The authors always used the same disc in these different patient groups, as no viscoelastic disc was marketed at the time.

Compared to discs with spherical ball and socket gliding surfaces (Table 2), viscoelastic discs may be judged on their design expanding TDR indications. But there are differences between various viscoelastic discs and between different ball and socket discs. Thus it is impossible to propose specific indications for all ball and socket discs and for all viscoelastic discs.

It is desired to determine the “right TDR” for each patient, but no clinical TDR guide exists. In comparison to spherical ball and socket discs (Table 3) advantages of viscoelastic discs are related to more stable biomechanical properties.

Devices providing suitable ROM and stability may be appropriate for patients with a wide set of TDR indications. Huang et al.⁶⁵ suggested contraindications may vary between different implant designs, specifically related to constraint, noting that constrained designs may be more suitable for patients with instability such as spondylolisthesis or post-facetectomy instability, or patients with mild facet arthrosis. Some 1-piece viscoelastic discs have limited ROM which may not protect adjacent levels.

Many questions must be answered before having specific indications for the “right total disc,” including:

Is there osteoporosis/osteopenia?

The literature review was assimilated into an overview on TDR patient selection. The target of this study was to improve the surgeon's decision making for the “right patient” with the “right indication,” and the “right TDR.” The proposals in Table 4 are not yet the final version, but a step toward being more precise in selection for TDR. Functional three- and two-component ball and socket discs have gliding areas not providing physiological ROM in all directions. In the future those TDRs will probably be replaced by implants with real physiological ROM, which 1-piece devices do not have. Table 4 is not intended to be comprehensive, but rather to identify what else should be evaluated when decision making is not clear.

The ideal TDR candidate may be an individual between 35 and 45 years old, with back pain severe enough to impact activities of daily living and/or work. Symptomatic DDD with or without radicular pain is the primary indication for TDR. Indications for TDR are based on patients’ clinical problems, on several image findings, and other information. Three examples:

Low back pain (DDD) caused by osteochondrosis

In summary, we believe TDR candidates should have failed sufficient nonoperative treatment, have no structural anatomic abnormalities, have BMD T score >-1.0, no significant psychological issues, and diagnostic studies confirming the disc as the pain generator. All major contraindications should be absent. There are few viscoelastic discs, but they may offer advantages over current ball and socket devices; more outcome data is needed to determine if they expand TDR indications. Three- and two-component discs with physiological ROM are not yet developed.