João Luiz Pinheiro-Franco, Alexander R. Vaccaro, Edward C. Benzel and H. Michael Mayer (eds.)Advanced Concepts in Lumbar Degenerative Disk Disease10.1007/978-3-662-47756-4_23

23. Adjacent Segment Disease: Natural History of Lumbar Degeneration or Consequence of Fusion?

Mark P. Coseo¹, Nelson S. Saldua², Eric B. Harris² and Alan Hilibrand³

(1)

Department of Orthopaedic Surgery, Naval Medical Center Portsmouth, 34800 Bob Wilson Drive, San Diego, CA 92134-5000, USA

(2)

Department of Orthopaedic Surgery, Naval Medical Center San Diego, San Diego, CA, USA

(3)

Thomas Jefferson University Medical College, 925 Chestnut Street 5th Floor, Philadelphia, PA 19107, USA

Mark P. Coseo

Email: mark.coseo@med.navy.mil

Nelson S. Saldua

Email: nelsonsaldua@gmail.com

Eric B. Harris (Corresponding author)

Email: ericbharris@gmail.com

Alan Hilibrand

Email: ahilibrand@gmail.com

Keywords

ArthrodesisAdjacent segment diseaseFusionArthroplastyMotion-sparing surgery

23.1 Introduction

Spinal arthrodesis is the surgical gold standard treatment for symptomatic lumbar degenerative disk disease. Over the last decade, the number of lumbar fusion procedures in the United States has risen dramatically [1, 2]. However, there is significant concern over the potential affects lumbar fusion procedures have on the health of adjacent segments. Loss of a lumbar motion segment has been shown to alter lumbar mechanics and increase disk pressures and loading of endplates and facet joints [3–5]. Long-term follow-up studies have shown rates of adjacent segment degeneration after lumbar fusion procedures over 80 % [6–8]. The question remains as to whether adjacent segment degeneration and subsequent disease are a result of lumbar fusion procedures or the result of natural history of progression of lumbar disk disease.

23.2 Adjacent Segment Disease: Definitions

Adjacent segment disease is symptomatic degenerative changes of an adjacent level following surgical intervention, whereas adjacent segment degeneration is the presence of abnormal imaging findings in the adjacent level following a surgical intervention at the index level that is asymptomatic. While there is no classification system for adjacent segment disease, it constitutes a myriad of signs and symptoms including axial low back pain, instability, stenosis, and radiculopathy of the cranial or caudal adjacent segment [9, 10].

23.3 Natural History of Lumbar Degeneration

The lumbar spine is susceptible to arthritic changes with age. Under pure compression, healthy lumbar disks demonstrate near uniform stress distribution. Physiologic forces are generally distributed equally along the endplate [11]. As a disk degenerates, stress concentrations shift to the annulus, progressing to disk bulges and protrusions. In addition, the posterior elements encounter greater transmission of forces. The affected segment becomes more susceptible to shear forces and less able to transfer load appropriately [12]. With moderate disk degeneration, there is greater translation under compression as the instantaneous axis of rotation of the spinal functional unit moves inferiorly and has a wider range toward the periphery [13, 14]. This results in an increase in motion of the spinal functional unit with moderate degeneration. However, as degenerative changes progress, motion decreases [12]. These degenerative changes occur most frequently at the L5–S1 level, followed by L4–L5, and then L3–L4 [15] and occur in asymptomatic patients. Multiple studies have shown a disconnect between imaging findings of degeneration and symptom profile. MRI studies of asymptomatic patients have demonstrated abnormalities in the lumbar spine in greater than 50 % of patients over 60 [15]. Even in asymptomatic patients, lumbar degeneration has been shown to progress over time. In a 2002 study, Elfering et al. examined 41 asymptomatic individuals with MRI studies over a 5-year period. The investigators found 41 % of patients had evidence of progression of lumbar disk degeneration over the 5-year span [16]. However, there was no statistically significant correlation of progression of degenerative change to development of symptoms.

If the progression of lumbar degeneration is a result of natural history as these studies indicate, are certain patient populations at higher risk? Multiple studies have shown there is a likely genetic predisposition to lumbar degenerative disease [17, 18–22]. Much of this evidence has been the result of twin studies. Battie et al. published on the results of MRI studies of 116 pairs of monozygotic Finnish twins. The investigators found familial and genetic influences accounted for variance in disk degeneration for 61 % in the upper lumbar region and 34 % in the lower lumbar region [19]. Another classic twin study, from Sambrook et al., reviewed lumbar MRI studies of 172 monozygotic and 154 dizygotic twins. They showed 64 % heritability for severe disease in the lumbar spine and overall heritability of lumbar degenerative disease of 74 % (95 % CI 64–81 %) [17]. A follow-on twin study of monozygotic and dizygotic Finnish twins by Battie et al. supported the high rate of genetic influence in lumbar degenerative disk disease. They estimated heritability ranging from 29 to 54 % depending on lumbar level and phenotype [23]. The higher concordance rate seen in the monozygotic twin populations has clearly shown there is a genetic predisposition to the development of lumbar degenerative disk disease.

However, it is also clear there is no signal gene responsible for this increased risk. Progress has been made on delineating possible candidate genes yielding the greatest influence on lumbar degeneration. Polymorphisms of the genes coding proinflammatory cytokines and matrix metalloproteinases leading to overexpression in the intervertebral disk have been hypothesized to be a contributing factor. Of particular concern is overexpression of MMP-2, MMP-3, and IL-1 leading to increased inflammatory response and increased susceptibility to lumbar disk degeneration [24–27]. In addition, healthy disks contain high concentrations of aggrecan and collagen-9. Certain polymorphisms in the genes encoding both these proteins have also been shown to be associated with increased risk of lumbar degenerative change [28–31]. Asporin is a leucine-rich repeat protein found in extracellular matrix. Overrepresentation of the D-14 allele of the asporin gene has been associated with increased risk of osteoarthritis and lumbar degeneration [32–34].

While progress has been made and certain gene alleles have been identified that are associated with lumbar degeneration, there is still much that is unknown. As gene sequencing becomes faster and computing power increases, genome-wide studies across multiple populations will become more feasible. This will not only help delineate the genotypes most at risk for developing lumbar degenerative disease but also may identify genetic profiles that are more resistant to disease progression. While the entire genomic influence in lumbar degenerative disease is unclear, it is evident from available studies there are populations that are more susceptible. This lends credence to the belief adjacent segment disease may be progression of natural history, rather than as a result of surgical intervention in the lumbar spine.

23.4 Patient Factors in Adjacent Segment Disease

In addition to genetic predisposition, certain patient factors have been hypothesized to increase risk of adjacent segment disease after fusion. There is evidence to suggest facet tropism and laminar inclination could predispose toward disease [35–38]. Okuda et al. reviewed 87 patients who underwent PLIF for L4 degenerative spondylolisthesis and found that facet tropism at L3–4 and horizontal lamina at L3 correlated with increased rate of degenerative change at L3–4 after L4–5 posterior lumbar interbody fusion [36]. The investigators hypothesized that facet joint asymmetry and increased lamina inclination angle resulted in abnormal motion of the spinal unit with increased intradiscal pressures and subsequent degradation [36]. In a follow-on study of this population, investigators found patients undergoing reoperation rates for adjacent segment disease had higher rates of facet tropism and lamina horizontalization in the cranial level to the fused segment [39].

Patient age is also thought to be a risk factor for adjacent segment disease [6, 9]. In a retrospective review of 3188 patients undergoing lumbar fusion procedures, Ahn et al. found a correlation between increasing age at index procedure and need for a reoperation on adjacent segments [6]. Interestingly, the study also found males were more at risk for reoperation [6]. Harrop et al. found similar conclusions in another retrospective review. Older age was associated with development of adjacent segment disease [9].

Finally, it has been hypothesized that obesity and history of smoking play a role in adjacent segment degeneration, while some studies reported no evidence of increased risk with higher BMI or nicotine intake [6, 9, 36, 40, 41]. Cho et al. reviewed 154 patients in a retrospective study and found age, BMI, and preoperative existing degenerative changes at the cranial level had the most significant risk for requiring repeat operative procedures for adjacent segment disease [42].

23.5 Prevalence of Adjacent Segment Disease

Many long-term studies have shown the progression of radiographic evidence of degenerative changes on adjacent segments following surgical intervention. While adjacent segment degeneration is common, rates of adjacent segment disease are much lower. A recent meta-analysis reviewed 94 studies with 34,716 patients included. In subgroup analysis, investigators found pooled prevalence rates for adjacent segment degeneration from 21.8 to 37.4 % and rates of 3.2–12.1 % of adjacent segment disease [43]. Harrop et al. in a meta-analysis found similar rates of adjacent segment degeneration and disease. The investigators reported rates of adjacent segment degeneration of 34 % (314/926) and disease rate of 14 % (173/1216) [9]. There is good evidence showing adjacent segment disease is more prevalent in the levels proximal to surgical intervention [9, 43–45]. Celestre et al. found in patients undergoing L4–5 fusion, 75 % developed adjacent segment degeneration at L3–4, while only 25 % had similar progression at L5–S1 [44]. The study found in all cases of adjacent segment disease the cranial levels were affected 90 % of the time [44].

The conclusions from these meta-analyses must be interpreted with caution due to their retrospective nature, differences in symptom evaluation, follow-up time, and imaging studies. However, our best available evidence does show a significant rate of adjacent segment degeneration and disease following lumbar fusion. The literature also shows cranial levels are at highest risk of development.

23.6 Is Fusion a Risk Factor?

Lumbar spinal motion is complex, as the functional spinal unit does not have a fixed center of rotation. It is inherently difficult to study given lumbar kinematic behavior is nonlinear. The lumbar spine is viscoelastic and motion coupling occurs in vivo [46]. This makes in vitro studies of lumbar spinal motion following arthrodesis difficult. However, multiple in vitro models have been developed to study the effect fusion has on adjacent levels [5, 47–49]. In a cadaveric study, Weinhoffer et al. measured intradiscal pressures in uninstrumented, single-level bilateral instrumented L5–S1 and multiple-level bilateral instrumented (L4–S1) models [48]. The investigators found intradiscal pressures were increased in the levels above the instrumentation and more significantly increased as more levels were instrumented [48]. The findings of this study were corroborated by a similar cadaveric study by Chow et al., in 1996. Single-level instrumentation at L4–5 increased intradiscal pressures in the cranial segment. Double-level instrumentation from L4–S1 resulted in marked increase in intradiscal pressure of the cranial segments [5]. This study also measured mobility of the cadaveric spines prior to and following instrumentation. They found one- and two-level instrumentation resulted in increased motion of the unfused segments closer to the extremes of their functional range [5]. Similar results have also been seen in in vitro studies of calf lumbosacral spine specimens. Shono et al. used this model to study the differences in motion following one-, two-, and three-level instrumentation in flexion, extension, lateral bending, and rotation [47]. As levels of instrumentation increased, the motion of the segments proximal to the construct had greater increases in motion [47]. This finding has also been seen in vivo. Axelsson et al. utilized roentgen stereophotogrammetric analysis of six patients undergoing fusion for low-grade spondylolisthesis to study the mobility of pre-fused and fused lumbar segments. Mobility of L4–5 and L5–S1 segments was measured preoperatively and after fusion. They found increased mobility of the proximal segment following fusion procedure [49]. While the results of these studies show concerning effects of loss of a motion segment, they must be interpreted critically. Current in vitro studies do not fully recreate the complexities of in vivo spinal motion, and available in vivo studies are limited by sample size. Improvements are needed in in vitro studies, and progress has been made to develop models that more accurately replicate spinal motion in vivo for bench studies [50].

It is very plausible biomechanical changes to the lumbar spine secondary to fusion result in abnormal forces on the adjacent segments and subsequent disease. However, does surgical intervention alone without fusion increase this risk? Some hypothesize decompressive laminectomy alone without fusion may alter lumbar mechanics and predispose to disease [51]. Biomechanical studies have shown posterior element excision in laminectomy produces increased motion of the surgical segment in flexion, extension, and rotation [52]. Using comprehensive national data from Sweden, Jansson et al. examined a 10-year follow-up on 9664 patients who underwent decompressive laminectomy for spinal stenosis [53]. They found progression of symptoms and need for reoperation in 11 % of patients on 10-year follow-up [53]. The results of this study and those like it show that fusion alone is not the sole risk for developing adjacent segment disease. The process of degeneration proceeds with non-fusion procedures as well.

While available data does show there is an increase in intradiscal pressures and motion in the levels surrounding a fusion mass, the question still remains if this change increases risk of adjacent segment disease. Models that more accurately represent spinal motion may provide insight if changes following loss of motion segments in the lumbar spine are well compensated in vivo or if they lead to symptomatic degeneration. Adjacent segment disease occurs after spinal fusion; however, there is no definitive evidence in the literature to date proving this phenomenon occurs as a result of the fusion.

23.7 Surgical Intervention: Does Method of Fusion Matter?

It is clear there are multiple methods to achieve fusion of a functional unit in the lumbar spine. However, it is still unclear if approach and fusion method utilized affects risk of adjacent segment disease. Some believe posterior approaches and pedicle screw-based instrumentation may predispose adjacent segment disease over anterior-based techniques [52, 54, 55]. Chen et al., in 2008, utilized CT scans following pedicle screw fixation for lumbar fusions to assess the rate of superior segment facet joint violation. Of this study cohort, 47 % of patients had evidence of superior facet joint violation after instrumentation [54]. It is possible that violation of the facets at the fusion interface may increase the risk of degeneration. If posterior approaches and instrumentation increase the risk of adjacent segment disease, it would follow then that anterior lumbar interbody fusion procedures would have lower rates of reoperation for adjacent disease. There is some level IV data to support this hypothesis. While not examining adjacent segment disease directly, Strube et al. found a significant difference in patient satisfaction, Oswestry disability index, and visual analog scale in patients undergoing anterior lumbar interbody fusion (ALIF) alone versus anteroposterior lumbar fusion [56]. Wai et al. performed MRI studies on 39 patients after ALIF with a minimum of 20-year follow-up. Only three patients in this group required additional surgery. The investigators also found 30.7 % of patients had advanced degeneration, but of that group, 17.9 % had preservation of the adjacent segment. They concluded the rates in their study were similar to unfused patients [8]. A study by Min et al. compared 48 patients who had undergone either L4–5 ALIF or PLIF for degenerative spondylolisthesis. They reported rates of adjacent segment degeneration of 44 % in the ALIF group and 82.6 % in the PLIF group. However, there was no significant difference in adjacent segment disease rate [57]. It is difficult to ascertain the clinical significance of the differential finding in adjacent segment degeneration in these two groups. Though likely, it is more consistent with progression of natural history of the degenerative spine rather than fusion related. Finally, while these studies may show a benefit to stand-alone anterior lumbar interbody fusion in association with development of adjacent segment disease, there is also evidence fusion method has no effect. Analyzing outcomes from the SPORT trial, Abdu et al. found no significant difference in 4-year outcomes comparing posterolateral in situ fusion, posterolateral instrumented fusion, and 360° fusion [58]. Given the disparity in the literature, there is no definitive answer as to the effect the method of fusion has on the development of adjacent segment disease.

Finally, potentially avoidable surgical results have been associated with increased risk of adjacent segment disease. Over distraction at time of fusion is hypothesized to pathologically load the posterior elements resulting in earlier degeneration [59]. In a study of 85 patients after posterior lumbar interbody fusion at L4–5, Kaito et al. found 13 out of 85 patients developed adjacent segment disease at 2-year follow-up. In the group developing adjacent segment disease, there was an average distraction of 6.1 mm at L4–5, while in the asymptomatic group, there was an average distraction of 3.1 mm [59]. There is also evidence to suggest the importance of maintaining near anatomic sagittal balance following fusion [60–63]. Keorochana et al. reviewed MRI findings of 430 patients with low back pain and found the patients with hyperlordosis (>50°) and hypolordosis (<20°) had increased disk degeneration compared to those with normal lordosis [63]. Djurasovic et al. corroborated these findings, where they found hypolordosis at the fused lumbar segment correlated with adjacent segment degeneration [62]. These results clearly show the importance of good surgical technique when performing lumbar fusion procedures. Particular attention must be paid in avoiding over-distraction and changes in postoperative sagittal alignment.

23.8 Motion-Sparing Surgery: Is Arthroplasty the Answer?

Concern of adjacent segment disease has helped lead to the development of disk and spinal functional unit arthroplasties of the lumbar spine. The goals of motion-preserving surgical interventions are to replace the diseased segment and remove the pain generator while maintaining as close to normal spinal kinematics and motion. It is thought this would help limit the risk fusion has on the development of adjacent segment disease. There is literature to support this claim. In a recent meta-analysis, Harrop et al. included 27 retrospective studies to compare the development of adjacent segment degeneration and disease in patients who had undergone lumbar arthrodesis versus arthroplasty. They found rate of adjacent segment disease in the arthrodesis cohort to be 14 % (173/1216) and a rate in the arthroplasty group to be 1 % (7/595) [9]. Another recent meta-analysis of 1584 patients at 2-year follow-up found a significant improvement in ODI and VAS scores of patients following lumbar disk arthroplasty compared to those after fusion [64]. These results appear promising in support of motion-sparing technology; however, they must be interpreted with caution given the nature of studies included in the meta-analysis. In addition, there is relatively little data on the long-term survivability of lumbar disk arthroplasty. The longest follow-up data is available from case series out of Europe [65, 66]. David reported on 106 patients with mean 13.2-year follow-up after one-level lumbar disk arthroplasty. In this case series, 82.1 % reported good to excellent outcomes, 90.6 % were still mobile with mean flexion and extension range of 10.1°, and only three cases (2.8 %) of adjacent segment disease [66]. Huang et al. reported on 42 patients at a mean of 8.7 years following lumbar disk arthroplasty. They found an adjacent segment degeneration rate of 24 % but noted no adjacent segment degeneration developed in patients with motion greater than 5° at their arthroplasty site [67]. Siepe et al. performed MRI studies on 93 patients at a mean of 53.4 months following lumbar disk arthroplasty and found an incidence of adjacent level degeneration of 10.2 % but characterized the level of degeneration as mild in all cases [68]. Comparing motion at adjacent segments following fusion versus arthroplasty suggests there is less change in postoperative range of motion following disk replacement [69, 70]. Berg et al. compared 72 patients following arthrodesis and 80 patients after total disk arthroplasty at 2-year follow-up and found significant difference in preservation of preoperative mobility of the adjacent segments in the arthroplasty group [70].

Only gold members can continue reading. Log In or Register to continue