This study examined whether sagittal alignment, preexisting adjacent level degeneration, and smoking predispose patients to adjacent segment degeneration following lumbar fusion. Fifty-one patients with adjacent segment degeneration were identified and matched with control patients based on age, sex, level, and date of index surgery. Preexisting adjacent level degeneration and sagittal alignment through the fusion and from L1-S1 were determined before and after initial surgery. Patients with adjacent segment degeneration had significantly less lordosis through the fusion and lumbar spine following their initial surgery. There was no significant difference in the amount of preexisting adjacent level degeneration and smoking between the adjacent segment degeneration and control groups. Fusion of the lumbar spine in abnormal sagittal alignment, with loss of lumbar lordosis, predisposes patients to the development of adjacent segment degeneration. Adjacent segment degeneration does not appear to be just a progression of preexisting degenerative changes at the adjacent level.
Degeneration of a mobile segment adjacent to a lumbar or lumbosacral fusion is a common long-term complication of fusion surgery. As the number of fusion operations in the United States continues to increase, with an estimated 220,000 fusion operations performed in 2007, this complication will be seen with increasing frequency. Long-term studies demonstrate radiographic adjacent segment degeneration in almost 100% of patients,1 and clinically significant symptomatic adjacent segment degeneration occurs in 5% to 27.5%.2-10
In symptomatic patients, adjacent segment degeneration sometimes requires revision surgery, usually an extension of the decompression with fusion of the adjacent segment. Because the clinical results of these revision surgeries generally have been inferior to those of primary surgery, prevention of adjacent segment degeneration is an important area of focus.
Several risk factors for the development of adjacent segment degeneration have been proposed in the literature. These include age, gender, diagnosis, fusion length, interbody fusion, sagittal alignment, the use of rigid pedicle screw instrumentation, and preexisting degenerative changes already present at the adjacent level.9
The role of sagittal alignment in causing or accelerating adjacent level degeneration has been debated. Kumar et al,11 in a retrospective study, noted an increased incidence of adjacent segment degeneration in patients who demonstrated abnormalities in the location of the C7 plumb line or sacral inclination. In another retrospective study, Rahm and Hall7 reported a minimal difference that did not reach statistical significance in sagittal alignment in patients with adjacent segment degeneration. Many surgeons intuitively believe fusing the lumbar spine in relative kyphosis will lead to an increased likelihood of adjacent segment degeneration.
This case-control study compared sagittal alignment, the level of preexisting degeneration at the adjacent segment prior to index surgery, and smoking status in patients who developed adjacent segment degeneration to matched controls.
Materials and Methods
A review of the surgical database at our institution revealed 51 patients who underwent revision surgery for symptomatic adjacent segment degeneration between January 1997 and December 2003. All patients underwent instrumented posterolateral fusion for a degenerative lumbar spine condition for their initial surgery. The technique used for the initial posterolateral fusion was consistent among the different surgeons in the study.
Instrumentation consisted of a rigid pedicle screw-rod construct in all cases, with autologous iliac crest bone graft used as the graft material. Pedicle screws were placed under direct vision using anatomic landmarks. In all cases, careful attention was paid to not violate the capsule of the unfused facet joint adjacent to the most rostral pedicle screw. All spondylolisthesis and scoliosis cases were fused in situ.
Patients were positioned on a Wilson frame, which was initially raised during the case, then lowered at the conclusion of the procedure prior to locking down the rod with set screws, to extend the hips. Patients undergoing instrumented fusion for trauma, tumor, infection, or major deformity were excluded.
In all patients, both the index and revision surgery were performed at the same institution, and all had radiographically proven adjacent level stenosis by computed tomography (CT)-myelogram prior to their revision surgery. Indications for revision surgery for adjacent level degeneration included recurrent back and leg pain unresponsive to conservative treatments such as exercise therapy, medication, and epidural injection with CT-myelogram proven stenosis at the level directly adjacent to the prior fusion. The same group of 5 senior surgeons from our center made the decision to operate on patients in both the adjacent segment degeneration and the control groups. Patients were seen annually as part of their standard of care.
Fifty-one matched controls were identified from the database prior to evaluation of their chart or radiographs. Controls were defined as patients who underwent a posterolateral instrumented fusion using pedicle screws for lumbar degenerative disease and had not developed clinically symptomatic adjacent level degeneration at their latest follow-up. Each patient with adjacent segment degeneration was assigned a control patient who was matched for age (within 10 years), gender, fusion levels, and date of index surgery (within 6 months). Control patients also were matched for date of surgery so that both cases and controls had a similar time period after surgery in which to develop adjacent level degeneration. This yielded 2 distinct patient populations: a group of 51 patients who developed symptomatic adjacent level stenosis requiring surgical treatment and a group of 51 control patients who did not develop adjacent level problems during the same time period. The 2 groups then were compared on 3 parameters: sagittal alignment, the level of degeneration preexisting at the adjacent level, and smoking status.
Sagittal Cobb measurements were made on standing lateral radiographs, both from L1 to S1 (total lumbar lordosis) and through the levels of the index fusion. Measurements were made at 2 time points. Preoperative measurements were made on the radiographs just prior to the index surgery, and postoperative measurements were made on standing radiographs at the patients first postoperative visit.
Because patients differed in the levels fused at their index surgery and different levels of the lumbar spine vary in the amount of normal lordosis expected at that level, we standardized these measurements by calculating a derived measure that we termed relative hypolordosis. This was calculated by the difference of the patients sagittal Cobb measurement and the expected lordosis through those particular levels, as defined by Jackson and McManus12 in normal volunteers. This derived measure (relative hypolordosis) essentially was a quantitative measure of how much that patients segmental sagittal alignment deviated from normal.
We also wanted to determine whether patients who developed adjacent segment degeneration already had a greater level of degeneration preexisting at the adjacent level prior to the index surgery. Therefore, the degree of degenerative changes present on plain radiographs at the adjacent level prior to the index surgery were graded for both the cases and controls, and then prior to revision surgery in the cases or at latest follow-up in the control patients. Degenerative changes were graded according to the UCLA grading system as no disease (grade I), mild disease (grade II), moderate disease (grade III), or severe disease (grade IV).13
Finally, smoking status and other demographic information were obtained from patients charts. For the purposes of this study, a nonsmoker was defined as someone who had never smoked or quit at least 1 month prior to surgery and did not resume smoking.
Statistical analysis of continuous variables, including sagittal Cobb measurements, was performed using the t test. Cases were compared to controls for their preoperative and postoperative sagittal Cobb measurements. P values <.05 were considered statistically significant. Categorical variables such as the presence or absence of smoking, or the progression of UCLA grade were analyzed using the chi-square test, again comparing cases to controls.
Because matching was carried out for age, sex, and level of fusion, these factors were identical in the cases and controls. Both groups consisted of 23 men and 28 women with a mean age of 54 years (range, 32-80 years).
Diagnoses at the time of index surgery were similar in the 2 groups. Diagnoses at the time of index surgery in the adjacent segment degeneration group included spinal stenosis (34 patients), spondylolisthesis (12 patients), postdiskectomy instability (7 patients), herniated disk and concomitant degenerative disk disease (9 patients), nonunion (2 patients), and degenerative scoliosis (4 patients). Diagnoses in the control group included spinal stenosis (28 patients), spondylolisthesis (16 patients), postdiskectomy instability (15 patients), nonunion (1 patient), herniated disk with concomitant degenerative disk disease (9 patients), osteomyelitis (1 patient), and degenerative scoliosis (1 patient).
The mean number of levels fused in each group was 1.68 levels, with a median of 2 levels in both groups. There were 20 single-level fusions, 27 two-level fusions, and 4 three-level fusions in each group. The most common levels fused at the index surgery were L4-S1 (14 patients), followed by L4-5 (11 patients), L3-5 (10 patients), L3-4 (5 patients), L5-S1 (4 patients), L2-4 (3 patients), L2-5 (3 patients), and L3-S1 (1 patient).
Mean interval between the index and revision surgery in the adjacent segment degeneration group was 58 months (range, 24 to 116 months). Mean follow-up in the control group was 55 months (range, 24 to 118 months). In the adjacent segment degeneration group, 43 patients developed adjacent segment degeneration at the level proximal to the index fusion, and 7 patients developed adjacent segment degeneration distal to the index fusion. One patient, with an index fusion from L2-4, developed adjacent segment degeneration both proximal (L1-2) and distal (L4-5) to the index fusion.
The results for sagittal alignment through the fusion levels are summarized in Tables 1 and 2. Mean lordosis expected through these levels, according to measurements by Jackson and McManus12 in normal volunteers, is -27.8°. Sagittal alignment through the levels of the fusion showed patients who developed adjacent segment degeneration had mean lordosis of -15.3° through the levels of fusion immediately postoperatively, whereas control patients had mean lordosis of -23.4°. Thus, patients who developed adjacent segment degeneration had less lordosis through the levels of their index fusion than control patients (P<.01). Patients with adjacent segment degeneration had 55% of normal lordosis through the fusion levels, and control patients had 84% of normal lordosis. Mean relative hypolordosis (or relative kyphosis) through the fusion levels was 12.2° for patients who developed adjacent segment degeneration and 4.1° in control patients. Differences in sagittal alignment between patients with adjacent segment degeneration and control patients prior to the index operation were not statistically significant.
The results for sagittal alignment from L1-S1 (total lumbar lordosis) are summarized in Tables 3 and 4. After the index surgery, mean total lumbar lordosis was 46.3° in patients who developed adjacent segment degeneration and 51.9° in control patients (P<.05). Mean total lumbar lordosis reported by Jackson and McManus12 was 60.9°, thus patients with adjacent segment degeneration had 76% of expected normal lumbar lordosis, whereas controls had 85% of expected normal lumbar lordosis. Again, there was a nonsignificant trend toward less total lumbar lordosis prior to the index surgery in patients who developed adjacent segment degeneration.
Level of Preexisting Degeneration at Adjacent Segment
Table 5 summarizes the UCLA grade of radiographic degeneration at the adjacent level prior to the index surgery and then just prior to revision in the adjacent segment degeneration patients and at latest follow-up in the control patients. There was no significant difference in the degree of degeneration at the adjacent level prior to the index surgery between the 2 groups. Mean UCLA grade of the adjacent level in patients with adjacent segment degeneration increased from 1.2 to 2.4, whereas mean UCLA grade of the adjacent level in control patients increased minimally from 1.5 to 1.6. Mean UCLA grade increased in 36 of 51 (71%) patients with adjacent segment degeneration compared with 5 of 51 control patients (9%) (P<.05). The final UCLA grade of the adjacent level in patients with adjacent segment degeneration was higher than in control patients (P<.01).
Table 6 shows the number of smokers in the adjacent segment degeneration and control groups before and after surgery. There was an early trend toward increased smoking rates in patients who developed adjacent segment degeneration. The odds ratio for smokers was 2.10 for developing adjacent segment degeneration compared to control patients; however, this did not reach the level of statistical significance (95% confidence interval, 0.48, 9.13) given the limited number of patients in the study.
Degeneration of the unfused motion segment adjacent to a spinal fusion remains a significant clinical problem. A study by Ghiselli et al10 demonstrated a 27.5% rate of revision surgery at 6.7 years of follow-up for symptomatic adjacent level degeneration. In the same study, Kaplan-Meier survival analysis predicted only a 63.9% disease-free survival at 10 years following the index operation.
Moreover, the results of surgical treatment for adjacent level degeneration generally have proven inferior to those obtained with primary lumbar and lumbosacral fusion performed for well-accepted indications.14-18 Thus, prevention of adjacent segment degeneration remains a primary goal, and prevention necessitates a clearer understanding of the risk factors that may predispose patients to the development of adjacent segment degeneration following lumbar fusion.
Park et al9 reviewed the literature regarding adjacent segment degeneration and identified several proposed risk factors for the development of adjacent segment degeneration. These include the level of preexisting degeneration at the adjacent segment, the use of pedicle screw instrumentation, posterior lumbar interbody fusion, facet joint injury, fusion length, diagnosis, sagittal alignment, age, female gender, postmenopausal state, and osteoporosis. Impingement on the adjacent nonfused facet by the cephalad pedicle screw also may predispose this facet to accelerated degeneration. However, the association of many of these risk factors remains unproven.
Our study focused on whether fusion of the lumbar spine in abnormal sagittal alignment predisposed patients to the development of adjacent segment degeneration. We used a retrospective case-control study design with a matching process that allowed us to control for a number of confounding variables believed to contribute to the development of adjacent segment degeneration.
The role of sagittal alignment in causing adjacent segment degeneration has been studied previously by Kumar et al.11 They reviewed 83 patients who had undergone lumbar fusion and divided their patient population into 4 groups based on the position of the C7 plumb line and sacral inclination. At 5 years of follow-up, they found a 53% incidence of radiographic adjacent segment degeneration in the subgroup that had abnormalities in both measures, whereas the group that had a normal C7 plumb line and sacral inclination demonstrated only an 8% incidence of adjacent segment degeneration.
Rahm and Hall,7 in a retrospective study on 49 instrumented fusions with 2 years of follow-up, reported patients who developed adjacent segment degeneration had 31° less lordosis compared with standard values of lordosis published by Bernhardt and Bridwell19; in comparison, patients who did not develop adjacent segment degeneration had 24° less lordosis. They could not demonstrate a significant association between sagittal alignment and the development of adjacent segment degeneration with the number of patients available in their study. It should be noted the normative values determined by Bernhardt and Bridwell19 were obtained from a pediatric population screened for scoliosis, with a mean age of 13 years, whereas the normal values from Jackson and McManus19 used in our study were from adults with a mean age of 39 years.
Our study demonstrated an association between abnormal sagittal alignment and the development of adjacent segment degeneration. The patients in our study who developed adjacent segment degeneration had significantly less lordosis both through their fusion levels and through the entire lumbar spine following their index surgery compared with matched control patients who did not develop adjacent segment degeneration. As with any observational, retrospective study, this study cannot definitively prove causation, but we have shown a definite association between sagittal malalignment and adjacent segment degeneration. It appears that fusing the lumbar spine in relative kyphosis may predispose patients to the development of adjacent segment degeneration.
The mechanism by which abnormal sagittal alignment causes adjacent level problems remains open to speculation. Relative hypolordosis through the fusion may shift the weight-bearing line anteriorly. In the adjacent unfused motion segment, this may lead to increased compressive load transmission through the disk space anteriorly, accelerating its degeneration. In vitro studies have shown lumbar fusion will cause increased segmental motion and intradiskal pressure in the adjacent segment20-24; thus, alteration of the normal compressive load distribution may accelerate these changes.
Our study also examined the amount of preexisting degeneration at the adjacent segment. Some have theorized that patients who develop adjacent level degeneration simply have a greater degree of degeneration already present at the adjacent level and that adjacent segment degeneration is simply a continuation or progression of these changes. We demonstrated there was no significant difference in the amount of preexisting degeneration at the adjacent segment in patients who developed adjacent segment degeneration and control patients who did not. Thus, in our study, patients who developed adjacent segment degeneration did not appear to have a greater degree of adjacent level degeneration already present prior to index surgery. This was based on a plain radiographic grading system, namely the UCLA grading system, which has been used by others in studies of adjacent segment degeneration.10,13
Although magnetic resonance imaging (MRI) may have been a more sensitive method of detecting and following adjacent segment degeneration, the majority of patients in our study had CT-myelography but not MRI. Plain radiographs appeared to be moderately sensitive to the development of adjacent segment degeneration as 70% of patients with degeneration demonstrated progression of their UCLA grade in the follow-up period compared to only 9% of control patients. It is instructive to note that 30% of the patients with adjacent segment degeneration in our study did not show any radiographic progression, thus confirming the need for advanced imaging studies (eg, MRI or CT-myelography) in the evaluation of patients for adjacent segment degeneration.
Finally, we decided to evaluate whether tobacco use was a significant risk factor for the development of adjacent segment degeneration. Although tobacco use has not been mentioned specifically previously as a risk factor, multiple studies have shown an association of cigarette smoking and disk degeneration in the unoperated spine.25-27 It makes sense intuitively that if nicotine products negatively alter normal disk nutrition and physiology, thus predisposing the disk to degeneration, then this effect would be accentuated when the disk is exposed to the greater biomechanical stresses associated with a juxta-fusion environment. There was a nonsignificant trend toward increased smoking rates in our adjacent segment degeneration group compared to our control group, but with the numbers in this study, we were unable to demonstrate a definite association. This question perhaps could be studied further in a larger prospective or retrospective cohort study design.
Our study demonstrated an association between loss of lumbar lordosis during lumbar fusion and the development of adjacent segment degeneration. It does not appear that adjacent segment degeneration is simply a progression of previous degenerative changes at the adjacent segment. Proper attention to sagittal alignment is one risk factor for adjacent segment degeneration that remains largely within the surgeons control via techniques such as proper patient positioning, correction with instrumentation, interbody grafting, and other techniques.
In the future, the maintenance or improvement of sagittal alignment during primary lumbar fusion may help to lower the incidence of adjacent segment degeneration. In addition, further understanding and clarification of the risk factors for adjacent segment degeneration will help clinicians in counseling patients with respect to their risk of requiring additional surgery in the future.
- Miyakoshi N, Abe E, Shimada Y, Okuyama K, Suzuki T, Sato K. Outcome of one-level posterior lumbar interbody fusion for spondylolisthesis and postoperative intervertebral disc degeneration adjacent to the fusion. Spine. 2000; 25(14):1837-1842.
- Frymoyer JW, Hanley EN Jr, Howe J, Kuhlmann D, Matteri RE. A comparison of radiographic findings in fusion and nonfusion patients ten or more years following lumbar disc surgery. Spine. 1979; 4(5):435-440.
- Booth KC, Bridwell KH, Eisenberg BA, Baldus CR, Lenke LG. Minimum 5-year results of degenerative spondylolisthesis treated with decompression and instrumented posterior fusion. Spine. 1999; 24(16):1721-1727.
- Kumar MN, Jacquot F, Hall H. Long-term follow-up of functional outcomes and radiographic changes at adjacent levels following lumbar spine fusion for degenerative disc disease. Eur Spine J. 2001; 10(4):309-313.
- Etebar S, Cahill DW. Risk factors for adjacent-segment failure following lumbar fixation with rigid instrumentation for degenerative instability. J Neurosurg. 1999; 90(2 suppl):163-169.
- Lehmann TR, Spratt KF, Tozzi JE, et al. Long-term follow-up of lower lumbar fusion patients. Spine. 1987; 12(2):97-104.
- Rahm MD, Hall BB. Adjacent-segment degeneration after lumbar fusion with instrumentation; a retrospective study. J Spinal Disord. 1996; 9(5):392-400.
- Kuslich SD, Danielson G, Dowdle JD, et al. Four-year follow-up results of lumbar spine arthrodesis using the Bagby and Kuslich lumbar fusion cage. Spine. 2000; 25(20):2656-2662.
- Park P, Garton HJ, Garton VC, Hoff JT, McGillicuddy JE. Adjacent segment disease after lumbar or lumbosacral fusion: review of the literature. Spine. 2004; 29(17):1938-1944.
- Ghiselli G, Wang JC, Bhatia NN, Hsu WK, Dawson EG. Adjacent segment degeneration in the lumbar spine. J Bone Joint Surg Am. 2004; 86(7):1497-1503.
- Kumar MN, Baklanov A, Chopin D. Correlation between sagittal plane changes and adjacent segment degeneration following lumbar spine fusion. Eur Spine J. 2001; 10(4):314-319.
- Jackson RP, McManus AC. Radiographic analysis of sagittal plane alignment and balance in standing volunteers and patients with low back pain matched for age, sex, and size: a prospective controlled clinical study. Spine. 1994; 19(14):1611-1618.
- Ghiselli G, Wang JC, Hsu WK, Dawson EG. L5-S1 segment survivorship and clinical outcome analysis after L4-L5 isolated fusion. Spine. 2003; 28(12):1275-1280.
- Chen WJ, Lai PL, Niu CC, Chen LH, Fu TS, Wong CB. Surgical treatment of adjacent instability after lumbar spine fusion. Spine. 2001; 26(22):E519-E524.
- Glassman SD, Pugh K, Johnson JR, Dimar JR II. Surgical management of adjacent level degeneration following lumbar spine fusion. Orthopedics. 2002; 25(10):1051-1055.
- Phillips FM, Carlson GD, Bohlman HH, Hughes SS. Results of surgery for spinal stenosis adjacent to previous lumbar fusion. J Spinal Disord. 2000; 13(5):432-437.
- Schlegel JD, Smith JA, Schleusener RL. Lumbar motion segment pathology adjacent to thoracolumbar, lumbar, and lumbosacral fusion. Spine. 1996; 21(8):970-981.
- Whitecloud TS III, Davis JM, Olive PM. Operative treatment of the degenerated segment adjacent to a lumbar fusion. Spine. 1994; 19(5):531-536.
- Bernhardt M, Bridwell KH. Segmental analysis of the sagittal plane alignment of the normal thoracic and lumbar spines and thoracolumbar junction. Spine. 1989; 14(7):717-721.
- Stokes IA, Wilder DG, Frymoyer JW, Pope MH. 1980 Volvo Award in Clinical Sciences: assessment of patients with low-back pain by biplanar radiographic measurement of intervertebral motion. Spine. 1981; 6(3):233-240.
- Axelsson P, Johnsson R, Stromqvist B. The spondylolytic vertebra and its adjacent segment. Mobility measured before and after posterolateral fusion. Spine. 1997; 22(4):414-417.
- Chen CS, Cheng CK, Liu CL, Lo WH. Stress analysis of the disc adjacent to interbody fusion in lumbar spine. Med Eng Phys. 2001; 23(7):485-493.
- Weinhoffer SL, Guyer RD, Herbert M, Griffith SL. Intradiscal pressure measurements above an instrumented fusion: a cadaveric study. Spine. 1995; 20(5):526-531.
- Cunningham BW, Kotani Y, McNulty PS, Cappucino A, McAfee PC. The effect of spinal destabilization and instrumentation on lumbar intradiscal pressure: an in vitro biomechanical analysis. Spine. 1997; 22(22):2655-2663.
- An HS, Silveri CP, Simpson JM, et al. Comparison of smoking habits between patients with surgically confirmed herniated lumbar and cervical disc disease and controls. J Spinal Disord. 1994; 7(5):369-373.
- Battie MC, Videman T, Gill K, et al. 1991 Volvo Award in Clinical Sciences: smoking and lumbar intervertebral disc degeneration: an MRI study of identical twins. Spine. 1991; 16(9):1015-1021.
- Frymoyer JW, Pope MH, Clements JH, Wilder DG, MacPherson B, Ashikaga T. Risk factors in low-back pain: an epidemiological survey. J Bone Joint Surg Am. 1983; 65(2):213-218.
Drs Djurasovic, Carreon, Glassman, Dimar, Puno, and Johnson are from the Kenton D. Leatherman Spine Center and the Department of Orthopedic Surgery, University of Louisville, Louisville, Kentucky.
Drs Djurasovic, Glassman, Dimar, Puno, and Johnson receive royalties, consulting fees, and research grant support from Medtronic Sofamor Danek. Dr Carreon has no relevant financial relationships to disclose.
Correspondence should be addressed to: Leah Y. Carreon, MD, MSc, 210 E Gray St, Ste 900, Louisville, KY 40202.