A Comprehensive Review of Spinal Arthroplasty Authors: David D Liu, Evan Dennis, Advait Patil, Emily H. Song, Joshua I. Chalif, Rohaid Ali, John H. Chi Published: 2026 ABSTRACT Spinal arthroplasty aims to preserve or reconstruct the normal biomechanical functions of the spine with motion-preserving implants. Although fusion is a proven technique to stabilize the spine, it is nonphysiologic with known limitations such as loss of mobility and risk of adjacent segment disease. This paper focuses on the main areas of spinal arthroplasty including cervical disc replacement, lumbar disc replacement, and lumbar facet arthroplasty. We review the biomechanics, history, outcomes, and future directions for each of these over the last 30 years. Although today spinal arthroplasty is only used in very specific degenerative settings, as innovation in spinal arthroplasty progresses, this will lead to wider adoption and a future where spine surgery is truly reconstructive and motion-preserving. INTRODUCTION Spinal arthroplasty aims to preserve or reconstruct the normal biomechanical functions of the spine with motion-preserving implants. The notion of the spine as a complex biomechanical object with 24 motion segments has been appreciated since antiquity [1]. Spine surgery began with limited decompressions to respect this mobility but has evolved to become more reconstructive over time. Indeed, one of the first dorsal spinal implants recorded, the Luque wire, sought to preserve the integrity of the dorsal elements and limit flexion while largely preserving other physiologic motion [2,3]. Since that time there has been significant progress in our ability to adequately decompress, stabilize, and realign the spine, often accompanied by arthrodesis (Fig. 1). With widespread adoption of well-proven tools such as polyaxial pedicle screw fixation, interbody cages, and biologic agents, spinal fusion has become routine in spine surgery and improves outcomes [4,5]. However, spinal fusion is not physiologic and has known limitations such as loss of mobility, pseudoarthrosis, and risk of adjacent segment disease (ASD). Over the last 30 years, there has been considerable development of motion-preserving implants that seek to replace damaged spinal structures while recapitulating the motion of the cervical disc, lumbar disc, and lumbar facet joints. These devices have developed along a spectrum, ranging from constrained designs (mechanical limitations and a fixed center of rotation), to unconstrained designs (no mechanical limitations, motion guided by soft tissues). Semiconstrained devices bridge the two by allowing for movement partially limited by mechanical stops. The degree of constraint determines how closely an implant replicates natural spinal movement [6]. The field of spinal arthroplasty continues to evolve, with advancements in clinical evidence, device innovation, and biomechanics; spinal arthrodesis remains a reliable mode of treatment and remains the gold standard for many pathologies. The relatively slow pace of adoption for spinal arthroplasty is noticeable in comparison to other orthopedic arthroplasty surgeries such as hip or knee replacement which have enjoyed considerable growth and adoption over the past several decades [7,8]. This article is a comprehensive evidence-based review of spinal arthroplasty devices, their varying degrees of success, outcomes, in vivo translational challenges, regulatory and payor barriers, and future directions. CERVICAL ARTHROPLASTY 1. Biomechanics of the Cervical Spine as Relevant to Motion Preservation The cervical spine exhibits significant motion at each segment in multiple planes: flexion/extension, axial rotation, lateral bending, and a small degree of translation [9]. The spinal column bears increasing weight from a cranial to caudal direction, and therefore the cervical spine bears the least amount of load. The main purpose of the cervical spine, especially the anterior column, is to support the weight of the cranium and maintain gaze [10,11]. The intervertebral disc constitutes 20%–33% of the entire height of the vertebral column. It is an amphiarthrotic joint, meaning that it allows for limited movement while providing stability. An increasingly large body of literature aims to understand how the nucleus pulposus, annulus fibrosus, and cartilaginous endplates create an organic environment that dynamically maintains load [12]. The disc is flexible at low loads but becomes stiffer and more stable at higher loads [13]. Therefore, in a cervical spine with physiologic alignment, an ideal arthroplasty device should be able to preserve motion in all directions while bearing the load of the cranium. These biomechanical factors have made cervical disc replacement (CDR) more accessible from an engineering perspective. As a result, there have been several cervical arthroplasty devices created, with good outcomes and clinical adoption. 2. History of CDR Devices Compared to the first anterior cervical discectomy and fusion (ACDF) in 1955 by Robinson and Smith, CDR is a relatively young field [14,15]. In a 1990s European trial, the Prestige ST (Medtronic Inc., USA) device, a metal-on-metal ball-and-trough style disc, was the first CDR device first implanted. It was subsequently approved in a U.S. Food and Drug Administration trial in the mid-2000s [16,17]. This was followed closely by the Pro-Disc-C (Synthes GmbH, Switzerland) in late 2007, making these two the earliest cervical discs available [18]. Both were constrained devices, as their metal core resulted in a fixed center of rotation with higher intrinsic restriction to motion. These devices have recognizable ball-and-socket or metal-on-metal dome-in-trough configuration which use their own geometry to stabilize the segment and prevent excessive displacement. The introduction of the Bryan disc (Medtronic Inc.) in the late 1990s, and its subsequent approval in 2009, marked a shift in biomechanical principles [19]. Comprised of a polyurethane viscoelastic nucleus encased between 2 titanium alloy shells, it is an unconstrained device without mechanical stops that attempts to replicate the kinematics of the healthy functional spine unit. From a biomechanical perspective, this emulates physiologic cervical motion by allowing for multidirectional rotation and translation. This differs from the constrained metal core devices, since the elastomeric core results in a center of rotation that can migrate dynamically rather than being fixed [20]. These devices established the safety of motion preservation in the neck. Subsequent devices have largely followed these 2 different philosophies (constrained metal-on-metal vs. unconstrained viscoelastic) as well as an intermediate semiconstrained approach. Unconstrained devices have the most physiologic motion pattern guided primarily by soft tissues, not the implant’s geometry; however, they are less stable under shear and more sensitive to surgical placement. Constrained devices have mechanical limits on motion dictated by the implant which make them more predictable and stable; however, they may induce more facet strain and metal implant wear, especially if not aligned with that motion segment’s natural center-of-rotation. Semiconstrained devices (e.g., Mobi-C, Simplify) aim to capture the advantages of both approaches by having a limited-translation core that allows for a variable center of rotation; however, given more complex mechanics, they are not as predictable and also face implant wear. Other differentiating factors among CDR devices include compressible cores (M6-C), and materials compatible with magnetic resonance imaging (MRI) imaging (Simplify). Devices may also differ in their implantation techniques, utilizing either keels or spikes for fixation. Table 1 summarizes the key U.S. Food and Drug Administration (FDA)-approved CDR devices, their manufacturers, and features. 3. Indications for Cervical Arthroplasty CDR devices have FDA approval for use at 1 or 2 (adjacent) levels from C3 to C7 in patients with myeloradiculopathy due to disc degeneration or herniation who have failed conservative treatment [21,22]. Contraindications include significant cervical instability, deformity, or facet joint arthrosis. These factors may alter the biomechanics of the physiologic cervical disc. For example, facet degeneration is thought to constrain motion to a smaller range, which would cause abnormal shear or instability on an implant leading to early failure or heterotopic ossification (HO) [23,24]. Significant pain with neck extension may be an indicator or facet mediated pain and is also a relative contraindication for cervical disc arthroplasty. Other contraindications to CDR include osteoporosis/osteopenia, other metabolic bone diseases, autoimmune spondyloarthropathies, systemic disease, and morbid obesity [25]. In practical terms, cervical arthroplasty is used for single or double-level disc herniations in younger adults, where there is usually no pathologic motion present preoperatively. There are increasing reports of “hybrid constructs” which combine CDR with ACDF at an adjacent level [26] (Fig. 2), mainly being used when adjacent level disease occurs following a previous ACDF. However, it should be noted that hybrid constructs are considered off-label use in the United States. 4. Outcomes Clinical outcomes for CDR have been extensively studied through multiple prospective randomized trials and follow-up studies. Overall, for the indications listed above, the literature demonstrates that cervical arthroplasty is at least as effective as ACDF both in terms of clinical and biomechanical outcomes. The safety and efficacy of CDR have not yet been validated in high-risk populations such as those with severe deformities or instability. Recent studies have reported CDR to be equally as effective for those over 65 years old and those under 40 years in terms of clinical outcomes, recovery, and complication rates, but longer-term follow-up for the elderly population is required and CDR continues to be performed at a greater volume in younger patients [27]. Each device underwent large-scale trials comparing it to ACDF for single-level cervical disc disease. Follow-up in these cohorts has been reported at regular intervals. Results were first reported at 2 years which established clinical efficacy and noninferiority to ACDF, which led to trial extensions. At 5-, 7-, and 10-year follow-ups, CDR has shown to be superior to ACDF in several areas including patient-reported outcome measures (PROMs), revision rates, and radiological outcomes. For example, the M6-C device demonstrated short-term improvement in visual analogue scale (VAS), Neck Disability Index (NDI), and Pain Catastrophizing Scale (PCS) at around 3 months that was maintained over the entire long-term course (Fig. 3). This pattern of initial improvement at 6 weeks to 3 months that is sustained at each longer-term follow-up is common to CDR devices, although the magnitude of improvement varies [28-30]. Table 2 summarizes outcomes for CDR devices from latest studies. In addition to superior clinical outcomes, several studies have demonstrated biomechanical superiority of CDR devices through reduced operations for ASD and preserved range of motion [31,32]. Ten-year follow-up for the Prestige, ProDisc, Bryan, and Mobi-C have demonstrated lower rates of secondary surgery for adjacent segment disease compared to ACDF [33-36]. Radiological studies have been undertaken with flexion-extension films which have demonstrated preserved motion of at least 90% in all physiologic motions (flexion, extension, rotation), compared to fusion which has a motion of 0 degrees [37]. Several complications unique to CDR include HO, implant failure or debris, and in the worst case, removal of the implant [38]. HO is the formation of bone around the implant, which constrains motion and therefore nullifies the original intent of motion preservation. HO is graded from I–IV, with III and IV representing fusion. The incidence of HO ranges but has been reported to be between 10%–25% by 10 years postoperatively. Repetitive motion can cause wear of load-bearing surfaces and eventual failure (device migration, subsidence, loosening, breakage). If the disc replacement fails or symptoms recur, revision surgery to remove the implant and perform fusion can be performed, which is rare, between 0%–3% of cases [39]. The material debris produced by metal-on-metal and metal-on-polymer devices can promote a proosteoclastic inflammatory cascade and resultant osteolysis. Recent studies have suggested that depending on implant, the rate of osteolysis can range from 2.4% to 36.2%, although it is often asymptomatic [40,41]. However, there is concern that reporting bias (e.g., Australian experience and recent MAUDE database) may partially explain the significant variability in these rates; further study is required to determine true long-term osteolysis rates. Notably the M6 device, which is designed to mimic the natural movement of a disc, is semiconstrained and has no metal-on-metal or metal-on-polymer interface which would lead to material debris. It has been reported to have the highest rate of osteolysis for unclear reasons. Thus, if osteolysis rates are related to a physiologic biomechanical design, further research on mechanisms of device failure is needed to improve future development. In addition to osteolysis, wear debris can lead to inflammatory hypersensitivity reactions, pseudotumor formation, or tissue damage [42]. Finally, a meta-analysis characterizing CDR complications reported that the pooled prevalence of complications such as dysphagia, recurrent laryngeal nerve palsy, hematoma formation, perioperative hemorrhage, and weakness and paresthesia of the upper limbs was low, ranging from 0.8% for vascular events to 4.7% for dysphagia [43]. 5. Future Directions The adoption of CDR continues to increase rapidly—as evidence for its efficacy accumulates, insurers and government payors have followed. For example, in the United States, at this time, more than 95% of commercial insurers cover CDR in appropriate patients [44]. However, Medicare currently does not cover CDR for patients over the age of 60 due to limited evidence for CDR compared to ACDF in this patient population [45]. More evidence will be required to evaluate the efficacy of CDR in the elderly population, where age-related bone quality changes and comorbidities may favor the predictable, long-term stability offered by spinal fusion. Further research will focus on long-term data, improvements in device design, and expanding surgical indications [46]. More patients will reach the 10- to 15-year mark, which will give clearer insight into true long-term clinical and biomechanical durability. Next-generation cervical discs are also exploring more physiologic designs such as elastomeric cores and fiber inlays that provide dynamic load bearing. Collaborations with materials scientists may yield devices made from materials that do not produce MRI artifact and minimize HO, wear debris, and osteolysis. Finally, as surgeons become more adept with CDR, there may be a movement toward using cervical arthroplasty for 3+ levels and in hybrid constructs, as well as eventually deformity. LUMBAR DISC ARTHROPLASTY 1. Biomechanics of the Lumbar Spine as Relevant to Motion Preservation The lumbar spine differs from the cervical spine in both magnitude of load and the degree of permissible motion. In vivo intradiscal pressure on lumbar discs in healthy volunteers [47] was shown to be more than 3 times the weight of the trunk, and these loads are even greater with weight-bearing activities. The lumbar spine also has significantly constrained motion compared to the cervical spine [48]. Normal lumbar segments allow flexion-extension, less axial rotation (especially limited at L5–S1 due to sagittal facet orientation), and lateral bending [49,50]. Importantly, lumbar segments also undergo translational movements (anterior-posterior shear) coupled with rotation [51]. Thus, there are more engineering constraints for LDR compared to CDR, due to high compressive load and more complicated motion constraint. Abnormal mechanics can accelerate facet and adjacent degeneration, or implant failure. Early lumbar disc replacements were typically ball-and-socket designs (e.g., Charité, Maverick, ProDisc-L) that primarily allow rotation and some bending, but they inherently constrain translation [52]. This can potentially create nonphysiologic kinematics or increased facet joint loading, because the natural lumbar segment’s center of rotation is not fixed but translates during motion. Newer designs like the activL incorporate a mobile polyethylene core that permits a few millimeters of translational movement in addition to rotation, more closely mimicking normal motion [53]. 2. History of LDR Devices LDR began earlier than CDR, with some of the earliest attempts in Germany in the 1980s with the Charité disc prosthesis [52]. The Charité I and II models were implanted in the mid-1980s in Europe, evolving into the Charité III, which became the first widely used lumbar disc device internationally. The Charité III features 2 cobalt-chrome endplates with a sliding ultra-high-molecular-weight polyethylene (UHMWPE) core in between. It gained significant usage in Europe in the 1990s. This gained traction based on a 2004 noninferiority randomized controlled trial comparing Charité to anterior lumbar interbody fusion for single-level degenerative disc disease at L4–5 or L5–S1 [54]. The Maverick (Medtronic Inc.) was another early-generation LDR consisting of a metal-on-metal constrained ball-and-socket design. While it demonstrated safety, efficacy, and superiority to fusion in a large U.S. FDA investigational device exemption (IDE) clinical trial in 2004, it ultimately did not reach the market [52]. The second LDR to be approved was the ProDisc-L in 2006, which is a fixed-center ball-and-socket design with metal (cobalt-chrome) endplates and a UHMWPE inlay [55]. ProDisc-L remains in use with over 15 years of follow-up data (the longest continuous use) [56] (Fig. 4). Notably, Depuy Spine voluntarily withdrew the Charité device after release of the improved Pro-Disc-L, as it owns both devices (subsequently acquired by Centinel Spine). Several other devices have been developed, but the activL was the third disc approved by the FDA, in 2015 [53]. No lumbar discs have been approved since that time. Table 3 summarizes the key lumbar arthroplasty devices and their status. 3. Indications for Lumbar Arthroplasty Currently, the ProDisc-L and activL devices are approved for single-level use at L4–5 or L5–S1 in skeletally mature patients. As of 2020, ProDisc-L also has approval for two-level use (e.g., both L4–5 and L5–S1), though relatively few 2-level LDRs are performed and long-term data on two-level cases is limited [57]. Other devices like M6-L (the lumbar version of the compressible core design) are in clinical trials but not yet approved in the United States [58]. The indications for LDR are more stringent than CDR, given the multifactorial nature of low back pain and more complex biomechanics of the lumbar region. It is typically performed in younger patients with single-level degenerative disc disease at L3–S1 and resultant pure discogenic back pain with or without radiculopathy, on imaging with clear disc degeneration but without foraminal stenosis, flavum hypertrophy, facet arthropathy, or instability [59,60]. Other contraindications include osteoporosis, deformity, prior fusion at adjacent segments, and obesity. Nonetheless, lumbar fusion remains the gold standard for more complex pathologies such as severe instability, deformities, or multilevel disease. 4. Outcomes Early European experiences in the 1980s–1990s with Charité showed that surgery was feasible and could relieve pain for some patients [61]. By the mid-2000s, more extensive U.S. trials (Charité vs. fusion, ProDisc-L vs. fusion) demonstrated that in appropriately selected patients, lumbar arthroplasty was noninferior to fusion at 2 years, with similar improvements in pain and function and slightly faster recovery [56,62]. The early enthusiasm in the mid-2000s was tempered by uncertainty in the 2010s regarding long-term variable outcomes and challenging revisions. Notably, no trials have shown clear clinical superiority of TDR over fusion, although several have demonstrated noninferiority [63]. Studies with Charité, ProDisc-L, and activL have all demonstrated slight and sustained improvement in PROMs such as ODI, 36-Item Short Form Health Survey, and VAS back pain, with meta-analyses showing a slightly greater improvement in back pain scores on average (p=0.04) [63]. It should be noted that overall clinical success as measured by patient satisfaction is high for TDR, which may reflect baseline high satisfaction with lumbar fusion as well, both over 90% [64]. Table 4 summarizes outcomes for TDR devices from latest available studies. From a biomechanical perspective, there is some evidence that TDR lowers ASD. In a cohort of 186 patients who received activL or ProDisc-L, 9.7% of patients had radiologic ASD and only 2.3% required surgery for this; after propensity matching, TDR had a significant lower likelihood of ASD compared to fusion (odds ratio, 0.32) [65]. A 2014 meta-analysis which pooled 1,270 TDR patients found lower rates of radiological and clinical ASD compared to fusion at both short (<5 years) and long-term (>5 years) follow-up. Specifically, there was 4.6% clinically significant ASD in the TDR group compared to 13.5% in fusion, with 1.0% and 7.8% operation rates respectively. Radiological studies with lumbar flexion-extension films have found preserved motion at the TDR segment (8° or more), but this did not correlate to differences in PROMs [66]. The literature suggests that some of these findings may be due to patient selection. For example, some subgroup analyses have shown that patients younger than 50 have greater improvements in PROMs, as opposed to patients over 60 [55,60]. There are concerns that implants in older patients may be more susceptible to implant migration, dislocation, or loss of durability [67]. Procedures such as minimally invasive spinal fusion have demonstrated good clinical outcomes in the elderly due to decreased surgical morbidity and more predictable bone-implant behavior [68]. In addition to younger age, multiple studies suggest that optimal candidates for LDR are nonobese, nonsmoking patients without significant spinal deformity. Patients with minimal to no facet arthropathy who exhibit preserved disc height and less comorbidities also tend to experience better outcomes [69]. LDR can cause facet joint degradation, which occurs due to shifts in the center of rotation that increase ligament and facet forces. Such requirements can limit eligibility for LDR to as little as 5%, contributing to its limited adoption across all patient demographics [70]. Similarly, given the importance of implant location relative to the biomechanics of the lumbar spine, surgeon experience has shown to influence outcomes [71]. Identifying an optimal patient cohort for LDR remains an active area of research. Complications of LDR are similar to CDR. HO can also occur, with a similar grading scale and incidence [72]. A European study with average 17-year follow-up of the Charité disc found a 60% incidence of spontaneous ankylosis, which highlights the increased strain in the lumbar region [73]. There is a theoretically higher risk of implant wear given the higher loads in the lumbar spine, although revision rates remain low, below 5% [57]. However, a longer-term study of the Charité disc with median follow-up 12.3 years showed that revision with fusion occurred in nearly 20% of patients at max 21-year follow-up and was associated with intraoperative challenges such as substantial bleeding [74]. A recent review of the U.S. National Inpatient Sample found revision rates for LDR to exceed that of fusion [70]. Surgeons often cite challenging revision scenarios due to the anterior or lateral approach as a concern for performing LDR. 5. Future Directions The adoption of LDR has been limited and the evidence base for LDR continues to grow. Compared to CDR, clinical outcomes have been equivocal, and as discussed previously, the complex multifactorial nature of low back pain makes the optimal group smaller and harder to identify. Currently, in carefully selected young patients with degenerative disc disease and no significant spinal deformities, instability, or osteoporosis or osteopenia, LDR may serve as an alternative to fusion. Larger sample sizes, longer-term follow-up, better understanding of implant behavior, and clearer indications will improve adoption. Lumbar fusion has consistently demonstrated clinical and patient-reported improvements across a very wide range of indications, making it the default choice for many spine surgeons [75]. In the United States, rates of LDR have decreased by 85% from 2005 to 2017, while rates of CDR have increased by 800% over the same time frame [70]. Given evolving evidence, LDR is still considered “experimental” in most settings [76]. Cost analyses with 2-year follow-up have shown that LDR is comparable or less expensive than fusion, largely due to decreased operative times and hospital stays. However, LDR is frequently denied insurance coverage due to concerns over delayed complications and revision surgery. Physicians and healthcare system receive less reimbursement payout for LDR versus fusion and are therefore disincentivized from adopting LDR in their clinical practice [70]. Engineering efforts in artificial disc mechanics are ongoing. Similarly to CDR, the main innovations will be in improved load absorption and materials science [77]. There is considerable interest in biological regeneration with stem cells or nucleus pulposus replacement which would completely obviate the need for device implantation but currently lacks any rigorous scientific validation for efficacy [78,79]. Surgical technique also continues to evolve, with improved lateral and oblique approaches that are often utilized in other types of minimally invasive spine surgery [80,81]. With improved evidence at the single level, lumbar LDR could eventually be used in multilevel or hybrid constructs. The future of LDR is cautiously optimistic—as usage and evidence increase, this will create a positive feedback cycle that will drive progress. However, given its complications, limited indications, and financial barriers, ongoing work is needed if adoption among the spinal surgeon community is to approach that of CDR. LUMBAR FACET ARTHROPLASTY 1. Biomechanics of the Lumbar Facet Joint as Relevant to Motion Preservation The lumbar facet joints play a critical role in guiding and constraining motion of the lower spine. Each lumbar motion segment has 2 facet (zygapophyseal) joints which are comprised of the inferior articular process of the superior vertebra and the superior articular process of the inferior vertebra [82]. Together with the intervertebral disc, they form a three-joint complex of each motion segment. Each pair of facets bears approximately 15%–25% of axial load under normal conditions, with this proportion increasing substantially when disc height decreases [82,83]. The facet orientation transitions from more sagittal in the upper lumbar spine (facilitating flexion-extension) to more coronal at L5–S1 (resisting anterior shear). The facets resist anterior shear and compression during extension; they distract slightly and unload the posterior elements and transfer load to the disc during flexion; during rotation, they carry up to 40% of the load. The facets thus act as bony guide rails that stabilize the spine while allowing controlled movement. Thus, the goal of lumbar facet arthroplasty is to constrain excessive motion—especially sagittal translation—and share axial and rotational loads with the anterior column. This is especially the case after a decompression procedure that entails removal of the lumbar facet, either due to facet hypertrophy, or as part of an interbody fusion that requires a more lateral exposure for cage insertion [84]. Theoretically, lumbar facet arthroplasty should reduce ASD while preserving physiologic modes of flexion, extension, and rotation [85]. 2. History of Lumbar Facet Arthroplasty Devices The Total Facet Arthroplasty System (TFAS) was the first complete posterior joint replacement, introduced in the mid-2000s [86]. It consisted of bilateral mechanical joints anchored into the pedicles, allowed controlled flexion-extension and limited axial rotation. Early feasibility trials in the United States demonstrated comparable pain and function outcomes to fusion and stable segmental motion (~3°–5°) at 2 years [87], but the device was never submitted for full FDA approval after the company dissolved in 2010. The ACADIA Facet Replacement System (Globus Medical Inc., USA) completed enrollment for an FDA IDE trial in 2009, with promising 1-year results, but the device has not been commercially released [88]. Most recently, the TOPS (Total Posterior Spine System, Premia Spine, USA) evolved from a facet replacement concept into a semiconstrained dynamic stabilization system, replacing both facets and the lamina with a mechanical articulating core [89]. After a multicenter FDA IDE trial showed superiority over transforaminal lumbar interbody fusion (TLIF) in composite success at 2 years, it became the first posterior motion-preserving implant approved by the FDA (2023) [90,91]. The TOPS group primarily included patients with single level symptomatic lumbar stenosis with or without spondylolisthesis from L2–5 and demonstrated superiority compared to TLIF on patient-reported outcomes, hardware complications, and recurring symptoms at 2 years. The TOPS system remains the only lumbar facet arthroplasty device available at this time. Of note, the TOPS device is a dynamic stabilization implant. It does not seek to replicate facet anatomy directly and therefore it is not a true lumbar facet replacement (Fig. 5). It is comprised of a polycarbonate urethane core that acts as an articulating element with an internal mechanical linkage system within a metal frame housing that allows for specific planes of motion and resists others; this is then linked to pedicle screw fixation [90]. The TOPS device controls flexion-extension (10°–15°), lateral bending (4°–6°), and axial rotation (1°–2°), with minimal to no anterior-posterior translation, which is similar to a physiologic lumbar facet joint. 3. Indications for Lumbar Facet Arthroplasty Key inclusion criteria include single level pathology between L2–5, degenerative spondylolisthesis with moderate to severe stenosis, at least 40/100 baseline ODI, age 35–80 years, and predominant leg (vs. back) symptoms. Contraindications include body mass index >40 kg/m2, multilevel disease, <4-mm disc height at index level, lytic spondylolisthesis, or previous fusion [92]. Thus, the typical use case is that of routine degenerative lumbar spinal stenosis with relatively preserved disc and no instability or deformity. 4. Outcomes Early feasibility studies (TFAS, ACADIA) demonstrated noninferiority to fusion with similar improvements in ODI and VAS scores, higher proportions of composite clinical success, and fewer adjacent segment degenerative changes at short- to midterm follow-up [91]. The TOPS IDE trial showed superior 2-year composite success (74% vs. 26% TLIF), with superiority in several PROMs including the ODI, VAS back pain, and all Zurich Claudication Questionnaire components scores. No patients in the TOPS developed symptomatic ASD, compared to 5.4% of patients in the TLIF group. Motion was maintained in flexion-extension and lateral bending, with no device failures or instability through 2 years. Long-term outcomes are being followed in a 5-year extension study [90,91]. Overall, these findings demonstrate the technical feasibility and biomechanical plausibility of facet arthroplasty and posterior motion preservation, though the evidence base remains limited relative to lumbar disc replacement. 5. Future Directions Facet arthroplasty remains an emerging field, constrained by the technical challenges of reproducing the complex kinematics and load-sharing of the posterior column. More long-term clinical follow-up is required to evaluate the efficacy of the TOPS device. Additional engineering innovations will be required to create simpler, lower-profile facet systems and possible hybrid constructs that integrate disc and facet replacements. Broader adoption will require demonstration of durable long-term outcomes, reduced adjacent-level stress, and improved revision strategies given the novelty of the device. As seen with CDR, clinical superiority will ultimately determine the trajectory of posterior motion-preserving surgery in the coming decade. CONCLUSION The future of spine surgery will involve motion-preserving implants, to preserve and restore physiologic motion. In many cases, fusion is an unnatural state for the spine and iatrogenic fusion confers an abnormal condition despite addressing other problematic issues. In this paper, we have reviewed the development, success, and ongoing trials of CDR, lumbar disc replacement, and lumbar facet arthroplasty that one day will help mitigate the need for fusion and arthrodesis of a spinal motion segment in appropriately selected patients. Spinal arthroplasty has seen limited adoption—particularly in the lumbar spine—due to ongoing clinical evaluation. As our understanding of arthroplasty evolves, in the future, these devices will also be used to restore motion in motion segments with pathological biomechanics, for example in disc collapse, deformity, infection, or trauma. This would allow spine surgeons to perform truly reconstructive surgery, as opposed to relying on our repertoire of decompression and fusion surgeries today.