- Email: [email protected]
The natural history of lumbar degenerative spinal stenosis
Joint Bone Spine 2002 ; 69 : 450-7 © 2002 Éditions scientifiques et médicales Elsevier SAS. All rights reserved S1297319X02004293/REV
REVIEW
The natural history of lumbar degenerative spinal stenosis Michel Benoist* Orthopedic surgery department, consultation in rheumatology, Hôpital Beaujon 92110 Clichy, France; Rheumatology department, Hôpital Bichat, 75018 Paris, France (Submitted for publication January 24, 2001; accepted in revised form July 25, 2001)
Summary – The purpose of this review is to present current information on the natural course of lumbar spinal stenosis. As the population becomes older this condition is encountered more frequently. The diagnosis accuracy has improved and the number of cases detected is increasing internationally. Because of the relative unpredictability of surgical treatment, good knowledge of natural evolution and of the predictive factors influencing the course of the disease is crucial. Unfortunately, and in contrast with numerous surgical series few studies have dealt with natural evolution. In addition to anecdotal reports, a few non randomized studies will be reviewed. Only one randomized study has compared short and long term results of medical versus surgical treatment. Most of these studies are retrospective, with methodological flaws and are difficult to compare.. At the present time no scientifically based recommandations can be made to LSS. patients at diagnosis. Similarly predictors of success of medical and surgical treatment still need to be identified. However results of the studies analyzed in this review suggest that a substantial proportion of patients do not automatically deteriorate and will remain unchanged or even improved by medical means. They also suggest that patients with severe baseline symptoms, block stenosis and degenerative spondylolisthesis tend to require surgical decompression. Randomized studies with the necessary ethical precautions are needed to obtain clear-cut conclusions. Joint Bone Spine 2002 ; 69 : 450-7. © 2002 Éditions scientifiques et médicales Elsevier SAS lumbodynia / lumbar spinal stenosis / spinal surgery
The purpose of this review is to present current information on the natural course of lumbar spinal stenosis (LSS). Numerous studies have described the clinical presentation and evaluated the diagnostic usefulness of the various symptoms and signs. Much of the literature has focused on the diagnostic accuracy of imaging studies including myelography, computed tomography (CT) and magnetic resonance imaging (MRI). Cor-
* Correspondence and reprints.
roboration of the clinical findings by the results of imaging studies has benefited the diagnosis of LSS. As the general population ages, this condition is becoming more common, although its epidemiology remains largely unknown. Rates of surgical treatment are increasing internationally. However, the effectiveness of surgery for LSS is still a matter of controversy. According to a recent meta-analysis, surgical outcomes vary greatly across studies [1]. Some authors have reported deterioration of initial good results with time [2, 3]. Complications or symptom exacerbation can occur after
Lumbar degenerative spinal stenosis
surgery. At present, there is no scientific evidence that surgery is effective in LSS. Because of the relative unpredictability of surgical outcomes in the individual patient, a good knowledge of the natural history of LSS is crucial. It is also important to identify the factors that influence the course of the disease, such as age, sex, duration and type of baseline clinical symptoms, and site and severity of the stenosis. Unfortunately, although numerous surgical series have been published, studies of the natural history are scarce. A single randomized study compared short- and long-term results of medical and surgical treatment [4]. Such studies are obviously difficult to conduct in an older population with a high rate of comorbidity and mortality and with a variety of lifestyles and socioeconomic conditions. There are also ethical problems. For example, it would not be ethically acceptable to randomize an older patient with mild symptoms to surgery or a patient with severe pain and functional disability to medical therapy. To appraise current information on this topic, relevant data from the literature must be identified. Acceptable studies are few in number and difficult to compare given the absence of a severity index evaluating the clinical and imaging data. Moreover there is no consensus regarding the criteria for evaluating the results. ANECDOTAL REPORTS A few reports of successful nonsurgical treatment of LSS have been published. Jones and Thompson [5] studied 13 cases. Three patients were treated conservatively. Two of these were unchanged or improved, and the third was lost to follow-up [5]. Blau and Logue reported two cases treated without surgery; one patient was unchanged after 10 years and the other had a slight deterioration after 7 years [6]. Among 70 patients reviewed by Tile, two were considered for surgery but had a favorable outcome with medical treatment [3]. Rosomoff and Rosomoff reported the case of a 70-yearold man who had a complete L3-L4 block by myelography but recovered fully without surgery [7]. Postachini [8] studied 12 untreated patients with central LSS followed up for 3 to 12 years. All patients underwent myelography, CT, and MRI. The stenosis involved one level in seven cases and two or more levels in five cases. Compression was severe in seven cases and moderate in five. The radicular symptoms worsened in two patients with severe stenosis, remained unchanged in six, and resolved in four patients of whom three had
451
moderate stenosis. Finally, I was able to conduct telephone interviews of 22 patients treated conservatively, among many others, in the past 10 years. Seven had surgery. The others were either unchanged or improved. These anecdotal observations have no scientific value but indicate that some patients do not deteriorate with time and are able to tolerate their disability without surgical decompression. Moreover, there are a few nonrandomized long-term patient-based studies that describe the natural history of LSS and the efficacy of medical treatment [1-4]. STUDIES ON THE NATURAL HISTORY OF LSS – REVIEW OF THE LITERATURE I will now review the few available studies, in chronological order. The first attempt to study the natural history of LSS was made by Porter and associates in 1984 [9]. Entrapment of the lumbar root within the root canal was recognized using four criteria; 1) severe, constant root pain to the lower leg; 2) pain unrelieved by bed rest; 3) minimal tension signs; and 4) patients over 40 years of age. Two hundred forty-nine patients fulfilled these criteria. Radiological degenerative changes were seen in 80% of the patients. No active treatment was given in the majority of the patients, with the exception of back school attendance in 22%. Fourteen percent received one or more steroid epidural injections. Surgical decompression of the root canal was done in 24 (10%) patients, among whom only three had recovered fully at the 1-year evaluation, while 15 were moderately improved and six were unchanged. The course in the 225 (90%) patients treated without surgery was assessed by a mailed questionnaire, to which 75% replied. After 3 years, 78% still had some leg pain but most were not bothered enough to undergo surgery. The clinical criteria used in this study to identify root entrapment are now recognized as classical clinical signs of lateral stenosis. However, the diagnosis was based only on clinical symptoms and signs: in most cases confirmation of the diagnosis by imaging studies was not obtained. Myelography was performed only in the few patients who had surgery. The study was done before widespread use of CT and MRI became the rule. Johnson and associates [10] compared the clinical course of central LSS in 44 surgically treated patients and 19 conservatively treated patients. All patients underwent lumbar myelography with water-soluble contrast. The 19 nonsurgical patients were advised to have surgery but either declined or were excluded by
452
M. Benoist
the anesthesiologist. The patients were divided into 3 groups according to the treatment and to the severity of the stenosis at myelography: the 19 nonsurgical patients had moderate stenosis, and among the 44 surgical patients, 30 had moderate stenosis and 14 a complete block. The midsagittal diameter of the contrast column was similar in the nonsurgical group and in the surgical group with moderate stenosis. These two groups were, therefore, comparable. Mean follow-up was approximately 3 years in the nonsurgical group and 4 years in the surgical group. The visual analog scale scores are reported in table I. Most nonsurgical patients were unchanged or improved; only 10% got worse, as compared to 20% in the surgical group with moderate stenosis. Table II shows the results evaluated by a nonindependent surgeon. Interestingly, the proportion of improved patients (approximately 60%) was similar in the three groups. At last follow-up, no significant differences were found for pain severity or analgesic use. However, there was a significantly greater increase in walking capacity in the surgical group than in the nonsurgical group. Among the nonsurgical patients, 30% improved and 60% remained unchanged, whereas after surgery 60% of patients improved and 25% deteriorated. However, as indicated by the authors, these findings must be interpreted with a critical mind. First, this was not a randomized study. Second, baseline pain severity is not reported. Third, the baseline walking capacity was lower and the imaging study findings more severe in the surgical patients. Specifically, there were significantly more patients with degenerative spondylolisthesis in the surgical group with moderate stenosis Table I. Three-year visual analog scale assessment of 19 conservatively treated and 44 surgically treated patients (from [10]). n
Worse %
No surgery 19 Surgery, moderate stenosis 30 Surgery, severe stenosis 14
10 20 36
Unchanged Improved % % 58 7 0
32 57 64
(n = 12) than in the nonsurgical group (n = 5). Moreover, the outcome assessment was made by a nonindependent observer and was based on subjective opinion. In another article, Johnson et al. [11] reported the natural course of 32 patients studied prospectively. Mean follow-up after the myelogram was 4 years, and mean age was 60 years. Surgical decompression was recommended to all 32 patients, but 30 declined and two were excluded by the anesthesiologist. About 75% of the patients had neurogenic claudication, whereas the others had radicular pain or mixed symptoms. Outcome measures, completed by the patient, included two visual analog scales, one for the overall course and the other for pain severity. Results were also assessed by clinical examination. Walking capacity was evaluated by a visual analog scale and by measuring the symptomfree walking distance on a level surface. Five patients were lost to follow-up (of whom three died). Overall results at last follow-up are shown in table III. Four patients (15%) worsened and the 23 remaining patients were unchanged or improved. As shown in table IV, all patients still complained of back and/or leg pain, which was severe in three patients. Although the walking distance of the whole group was unchanged overall, it was worse in 30% of the patients. Overall, at last follow-up, 70% of the patients were unchanged, 15% were better, and 15% were worse. When the course was studied according to the midsagittal diameter of the contrast column, the narrowest diameters were found to be in the four patients whose symptoms worsened (table V). The natural course of 91 nonsurgically treated patients was reported at the 1996 annual meeting of the International Society for the Study of the Lumbar Spine, by Herno and associates [12]. This was a retroTable III. Four-year visual-analog scale results in 32 conservatively treated patients (from [11]). n patients 27
Worse %
Unchanged %
Improved %
15% 4
70% 19
15% 4
Table II. Three-year clinical examination assessment of 19 conservatively treated and 44 surgically treated patients (from [10]).
No surgery Surgery, moderate stenosis Surgery, severe stenosis
Worse %
Unchanged %
Improved %
1 0 1
37 36 29
58 63 64
Table IV. Pain severity after 4 years in 32 conservatively treated patients (from [11]). • Mild • Moderate • Severe
19% 70% 11%
Lumbar degenerative spinal stenosis Table V. Correlation between anterior-posterior diameter and clinical outcome in 27 conservatively treated patients (from [11]). VAS result
n patients
AP diameter (mm)
Worse Unchanged Improved
4 19 4 27
4.7 6.8 8.2
VAS: visual analog scale; AP diameter: anterior-posterior diameter.
spective study with a mean follow-up of 8 ± 3 years. The patients were divided into four groups according to the radiological findings: 11 had a complete block, 40 had moderate stenosis (anterior-posterior diameter < 10 mm), 18 had mild stenosis (10–12 mm), and 22 had lateral stenosis. Outcomes were evaluated based on the Oswestry questionnaire and on the treadmill walking capacity. The severity of pain in the back and legs before and after walking on the treadmill was assessed using a visual analog scale. At follow-up, the overall patients’ subjective evaluation showed that 27 patients were unchanged, 41 improved, and 23 worse. The mean Oswestry score was 31 ± 16. The score did not differ significantly according to the radiological findings or follow-up duration. Pain severity before and after the treadmill test did not differ between the radiological groups. No difference in walking capacity was noted among the four groups. The final conclusion of the study was that the natural course was benign in all 91 patients and that the subjective and physical manifestations were remarkably stable. In another study, Herno and associates used a matched-pair design to compare outcomes in surgically treated and nonsurgically treated patients [13]. They were able to form 54 matched pairs. The matching criteria included sex, age, symptom duration, and myelographic findings. The main manifestations were neurogenic claudication, leg pain, and mixed symptoms. The myelographic findings were graded in four categories: total block, subtotal block (anterior-posterior diameter < 10 mm), moderate stenosis (10–12 mm), and lateral stenosis. Outcomes were evaluated based on the Oswestry score and on an estimation of the functional status by clinical examination. The mean Oswestry score was similar in the overall population and in the women with and without surgery but was significantly better in the surgically treated men. Functional status was good in both groups. Outcomes were not correlated with age, site of the stenosis (central, n = 35; or lateral, n = 19), or presence or absence of coexisting
453
illness. This study has major limitations. As indicated by the authors, it is retrospective and the baseline levels of pain and disability are unknown. Moreover, in 37 nonsurgically treated patients, the pain was not severe enough to require an operation. This casts some doubt on the reliability of the matching. At the 1997 annual meeting of the International Society for the Study of the Lumbar Spine, the same authors [14] presented a longitudinal analysis of 38 nonsurgically treated patients with LSS. Nineteen women and 19 men were examined in 1989 and 1995. Most had one or several coexisting diseases. The radiological findings were graded according to the severity and site of the stenosis (central, n = 27; or lateral, n = 11). The mean Oswestry score of all 38 patients was 34.3 in 1989 and 33.4 in 1995. The overall alteration in the patients’ condition was evaluated based on four criteria: physical condition, ability to perform daily activities, walking capacity on a treadmill, and pain on a visual analog scale after the treadmill test. Taken together, these parameters show that the overall condition remained almost unchanged. However, the patients with a complete block at myelography worsened. Herno et al. concluded that patients with block stenosis require surgical decompression. Saal et al. [15] studied 52 patients with a follow-up ranging from 2 to 8 years. All patients had radiculopathy, walking capacity restriction, and evidence of central stenosis with or without lateral stenosis on the imaging studies. Treatment included analgesics, epidural injection, and physiotherapy. An independent observer used a questionnaire and telephone interview to assess outcomes. Thirty-three of the 52 patients had good pain control with non-narcotic analgesics and no or minimal walking restriction. The remaining patients received epidural and physical therapy. There was no neurologic deterioration. Four patients were treated surgically. Saal et al. concluded that surgery can be avoided and that medical and physical treatments are associated with better results than the natural history as reported by other authors. Unfortunately, no correlation was made in this study between the outcome and the clinical symptoms at diagnosis or the site and severity of the stenosis. Although nonrandomized, the Maine lumbar spine study is informative [16, 17]. One hundred and fortyeight patients were enrolled in this prospective cohort study performed by a group of surgeons in the state of Maine in the U.S.A. Choice of treatment was decided after discussion between the patient and surgeon.
454
M. Benoist
Eighty-one patients underwent laminectomy, accompanied in three cases with instrumented fusion. The 67 remaining patients received medical treatment by analgesics, epidural steroids, back-school exercises, and physiotherapy. Results were evaluated at months 3, 6, 12, 24, 36 and 48. The evaluation criteria included a symptom severity index based on frequency and severity of low back and leg pain (score from 0 to 24). Other outcome measures included the SF36, the Roland modified functional score, and a health questionnaire with five outcome categories (excellent, 1; bad, 5). Satisfaction with care was evaluated also. Tables VI and VII summarize the 1-year outcomes in the two groups. Overall, approximately 75% of the surgically treated patients were improved and satisfied, and 40% of the medically treated patients were improved. These data indicate a clear superiority of surgical treatment. However, the two groups are difficult to compare because, at baseline, the surgical patients had more severe pain, worse functional incapacity, and more extensive stenosis on imaging studies. The 119 (80%) patients evaluated after 4 years included 52 (78%) of the 67 nonsurgically treated patients and 81 (80%) surgically treated patients [16, 17]. Again, outcomes were better after surgery, but the difference was less marked, as only 70% of the operated patients were still improved and 63% satisfied. At the same follow-up evaluation, 52% of the medically treated patients were improved and 42% were satisfied with their care. Some additional information must be emphasized. First, 22% of the medically treated patients underwent surgery during the 4-year follow-up period, with less satisfactory results than in the patients who had surgery at baseline. Second, 51 of the 119 patients evaluated after 4 years had moderate symptoms at baseline. In this group, the 29 surgical patients had significantly better outcomes than did the 22 nonsur-
gical patients. The high rate of follow-up at 4 years and the good quality of the outcome measures are important strengths of this nonrandomized study. However, as pointed out by Nachemson [18], the imaging data are not reported in detail, so that it cannot be determined whether the severity of the stenosis on imaging studies was correlated with an obvious superiority of surgery over medical treatment. A multivariate analysis [16, 17] identified interesting predictive factors. Unilateral leg pain and an interval of less than 1 year from pain onset to the first physician visit predicted a good outcome after surgery. In contrast, predominance of low-back over radicular pain and severe deterioration in functional capacity were associated with worse outcomes. Hurri et al. [19] focused on the long-term prognosis. Seventy-five patients, out of a total of 134 patients with LSS diagnosed by myelography between 1978 and 1982, were evaluated in 1993 by telephone interview, using a structured questionnaire. Forty-eight had died and 11 could not be traced. Disability was assessed according to the Oswestry score and the subjective assessment by the patient (improved, unchanged, or worse). The stenosis was severe (7.0 mm) in 32 patients (26 surgically treated) and moderate (7–10.5 mm) in 43 patients (31 surgically treated). Tables VIII and IX summarize the subjective assessments in the two groups. Most medically treated patients were unchanged or improved. Worsening was usually correlated with a severe stenosis. Shortcomings of this study are the large number of patients lost to follow-up and the nature of the outcome measurements. Simotas et al. used conservative treatment in 49 patients for 16 to 55 months (mean, 33 months) [20]. Treatment was not standardized and included analgesics, epidural steroids, and physical therapy. Nine of the 49 patients had surgery, with a mean interval of 13
Table VI. Outcome of the non-operated group (from [2]).
Table VIII. Twelve-year subjective assessment of 18 conservatively treated patients (from [9]).
Low back pain Leg pain
Better %
Same %
Worse %
41 44
38 42
20 12
Table VII. Outcome of the operated group (from [2]). Low back pain Leg pain
Better %
Same %
Worse %
77 78
17 15
5 6
Worse
Unchanged
Improved
11%
45%
44%
Table IX. Twelve-year subjective assessment of 57 operated patients (from [9]). Worse
Unchanged
Improved
18%
19%
63%
Lumbar degenerative spinal stenosis
months after initiation of medical treatment. Five of the 40 remaining patients worsened, 17 were unchanged, and 26 were moderately improved. The outcome measures used in this study were excellent and included a condition-specific health status index, inspired from that of Stucki and well adapted to accurate evaluation of LSS. An effort was made to correlate the treatment results with the imaging study data, including severity and extent of stenosis and presence of degenerative spondylolisthesis or scoliosis. Most surgically treated patients had severe or multilevel narrowing, spondylolisthesis, or scoliosis. However, the small number of study patients studied precluded a valid statistical analysis. Furthermore, the surgical patients had more severe symptoms at baseline. PARTIALLY RANDOMIZED STUDY In a recent study by Amundsen et al. [4], a small group of patients were randomized to surgery or conservative therapy. One hundred patients with back and leg pain and LSS were included. Those patients with severe pain and disability were immediately directed to surgery (n = 19). On the other hand, the patients with mild symptoms underwent medical treatment including rest, bracing, back school, and physiotherapy and were encouraged to resume a normal lifestyle with their usual physical activity (n = 50). The 31 remaining patients had moderate clinical symptoms and were randomized, after informed consent, to surgery (n = 13) or conservative therapy (n = 18). Surgery consisted of laminectomy without fusion or instrumentation. An independent observer reevaluated the patients after 6 months and after 1, 4, and 10 years. Evaluation of results was based on the overall evaluation of pain on a visual analog scale, claudication distance, level of daily activities, and consumption of analgesics. The patients were asked to classify the results as better, unchanged, or worse. Together with the examining physician, overall results were considered good if there was total restitution or partial restitution with clear improvement over the initial condition and bad if there was either deterioration or no change. Among the patients randomized to surgery, one died and 11 of the remaining 12 patients had a good result after 6 months and 4 years. However, after 1 year, only nine patients had a satisfactory outcome. Only seven (39%) of the 18 patients randomized to conservative treatment had good results after 6 months and eight
455
after 4 years. Six patients underwent surgery 3 to 26 months after study inclusion. In the group of nonrandomized medically treated patients who had mild symptoms at baseline, 64% of 50 patients had a good result after 1 year and 57% after 4 years. Ten patients required surgery. In the group of 19 nonrandomized surgically treated patients, 89% had good results after 1 year and 84% after 4 years. There were no complications, and only one patient required repeat surgery. Surgery was performed eventually in 20 of the 68 medically treated patients (18 randomized and 50 nonrandomized), i.e., slightly more than one-third. Overall, the treatment result was significantly better in the patients randomized to surgery than in those randomized to medical treatment. However, it should be noted that adequate conservative therapy maintained a favorable outcome in approximately two-thirds of the nonrandomized patients with mild or moderate symptoms at baseline. In this series, the results of surgery after failure of conservative treatment were similar to those of first-line surgery. Fourteen patients died. All the other patients were reevaluated after 10 years and found to have no significant worsening of symptoms. The treatment results [4] were not correlated with clinical parameters such as age, sex, symptom duration, socioeconomic status, or physical findings. Similarly, no associations were found between clinical results and various imaging parameters such as spondylolisthesis, site of stenosis (central and/or lateral), severity of the narrowing, and whether one or more levels were affected. The small number of study patients does not allow definitive conclusions. CONCLUSIONS Data in the literature and our own experience suggest a number of conclusions, although several questions remain without clear answers. First, the overall results of surgery are better than those of medical treatment. This appears clearly in studies comparing the two treatments [4, 16, 17, 10]. The proportion of good results (70 to 75%) after surgery in these studies is similar to that already disclosed in other published series [21]. These results usually remain stable over time, with no major complications. Second, a substantial proportion of conservatively treated patients have good results that are sustained for up to 10 years. Consequently, and despite the apparent superiority of surgical treatment, in many cases surgery is not indispensable. Further-
456
M. Benoist
more, if the symptoms worsen, surgery performed after failure of medical treatment seems to yield similar results to those obtained in patients who have surgery earlier. However, and as mentioned above, the conclusions of the studies by Atlas et al. [16, 17] are less optimistic. Additional prospective studies are needed to conclude. For ethical reasons, such studies could proceed by stages, with interim analyses every 6 to 12 months. Patients with disabling pain and functional status would be randomized to surgery or no surgery. Predictors of success or failure of medical or surgical treatment need to be identified. The results of the studies discussed above suggest that patients with severe baseline symptoms, block stenosis, and degenerative spondylolisthesis tend to deteriorate over time and require surgical decompression. However, this has not yet been scientifically demonstrated. The main limitation of the studies analyzed in this review lies in the outcome evaluation. LSS is a specific subgroup of low back pain, and outcome measures adapted to its characteristics must be used to obtain an accurate evaluation of treatment efficacy and to allow a comparison with the natural course. I believe that future studies should use: – two visual analog scales, one for evaluate low-back pain and the other for radicular pain; – a condition-specific health status index, such as that proposed by Stucki [22]; – a generic health status index, such as the SF36, to evaluate both physical and nonphysical impairments [23];; – a psychological evaluation using a self-administered depression scale; and – a cumulative illness rating scale to detect potential comorbidities [24]. At present, it seems reasonable to follow the recommendations made by Amundsen. Treatment should start with well-organized and closely monitored conservative therapy. If the pain remains mild or moderate with no major disability, conservative care can be continued with fair chances of stability or even improvement. If the low-back and leg pain remains or becomes severe, surgery offers a good chance of stable improvement. Considering the present absence of predictive imaging factors, the indication for surgery must be based on pain and disability rather than on the severity of the stenosis. Ultimately, the treatment is chosen by the patient, who should be fully informed of the advantages and disadvantages of each treatment option. The choice also depends on the patient’s objectives, comorbidity,
cardiovascular status, general health condition, lifestyle, and socioeconomic status. Furthermore, medical treatment must be properly organized. Epidural steroids seem efficacious. In a recent randomized study, caudal epidural steroid injections showed efficacy after 3 months, although the beneficial effect faded between 6 and 12 months after the injection [25]. Further studies are needed to confirm these data and to determine the role for other medical treatment modalities such as orthotics and physical therapy. REFERENCES 1 Tuite GF, Stern JD, Doran SE, et al. Outcome after laminectomy for lumbar spinal stenosis. Part I : Clinical correlations. J Neurosurg 1994 ; 81 : 699-706. 2 Katz JN, Lipson SJ, Chang LC, et al. Seven to 10 year outcome of decompressive surgery for degenerative lumbar spinal stenosis. Spine 1996 ; 21 : 92-8. 3 Tile M, Mcneil SR, Zains RK, et al. Spinal stenosis: results of treatment. Clin Orthop 1976 ; 115 : 104-8. 4 Amundsen T, Weber H, Nordal H, et al. Lumbar spinal stenosis: conservative or surgical management. A prospective 10 year study. Spine 2000 ; 25 : 1424-35. 5 Jones RAC, Thompson JLP. The narrow lumbar canal. A clinical and radiological review. J Bone Joint Surg 1968 ; 50 B : 595-605. 6 Blau JN, Logue V. The natural history of intermittent claudication of the cauda equina. Brain 1978 ; 101 : 211-22. 7 Rosomoff H, Rosomoff R. Nonsurgical aggressive treatment of lumbar spinal stenosis. Spine 1987 ; 3 : 383 p. 8 Postacchini F. Lumbar spinal stenosis. Long-term results. In: Wiesel SW, Weinstein JN, Herkowitz H, Dvorak J, Bell G, Eds. The lumbar spine, Volume 2. Philadelphia: WB Saunders; 1990. p. 766-81. 9 Porter RW, Hibbert C, Evans C. The natural history of root entrapment syndrome. Spine 1984 ; 9 : 418-21. 10 Johnsson Ke, Uden A, Rosen I. The effect of decompression on the natural course of spinal stenosis. A comparison of surgically treated and untreated patients. Spine 1991 ; 16 : 615-9. 11 Johnsson Ke, Rosen I, Uden A. The natural course of lumbar spinal stenosis. Clin Orthop 1992 ; 279 : 82-6. 12 Herno A, Nevalainen S, Saari T, et al. The natural course of 91 nonoperated patients with lumbar spinal stenosis. Presented at the annual meeting of the International Society for the Study of the Lumbar Spine. Burlington 1996; June : 25-9. 13 Herno A, Airaksinen O, Saari T, et al. Lumbar spinal stenosis: a matched pair study of operated and nonoperated patients. Br J Neurosurg 1996 ; 10 : 461-5. 14 Herno A, Nevalainen S, Saari T, et al. The longitudinal analysis of 38 monoperated patients with lumbar spinal stenosis. Presented at the annual meeting of the International Society for the Study of the Lumbar Spine. Singapore 1997; June : 2-6. 15 Saal JS, Saal JA, Parthasarathy R, et al. The natural history of lumbar spinal stenosis. The results of nonoperative treatment. Presented at the annual meeting of the International Society for the Study of the Lumbar Spine. Singapore 1997; June : 2-6. 16 Atlas SJ, Deyo RA, Keller RB, et al. The Maine Lumbar Spine study. Part III. Year outcomes of surgical and nonsurgical management of lumbar spinal stenosis. Spine 1996 ; 21 : 1787-95. 17 Atlas SJ, Keller RB, Robson D, et al. Surgical and non surgical
Lumbar degenerative spinal stenosis
18 19 20 21
management of Lumbar Spinal Stenosis. Four year outcomes from the Maine Lumbar Spine Study. Spine 2000 ; 25 : 556-62. Nachemson A. Point of view. Spine 1996 ; 21 : 1794-5. Hurri H, Slatis P, Soini J, et al. Lumbar spinal stenosis: assessment of long-term outcome 12 years after operative and conservative treatment. J Spinal Disord 1998 ; 11 : 110-5. Simotas A, Dorey F, Hansray K. Non operative treatment for lumbar spinal stenosis. Spine 2000 ; 25 : 197-204. Turner JA, Ersek K, Herron L, Deyo R. Surgery for lumbar spinal stenosis. Attempted meta-analysis of the literature. Spine 1992 ; 17 : 1-8.
457
22 Stucki G, Daltroy L, Liand M, et al. Measurement properties of a self-administered outcome measure. Spine 1996 ; 21 : 796803. 23 Linn BS, Linn M, Currel L. Cumulative illness rating scale. J Am Geriatr Soc 1968 ; 16 : 622-6. 24 Ware JE, Sherbourne CD. The MOS 36 item short form health survey (SF36). I. Conceptual frame work and item selection. Med Care 1992 ; 30 : 473-83. 25 Fortin L, Bergeron Y, Lambert R, et al. Double blind randomized clinical trial of epidural injections in spinal stenosis. Arthritis Rheum 2000 ; 43 (9 Suppl).
REVIEW
The natural history of lumbar degenerative spinal stenosis Michel Benoist* Orthopedic surgery department, consultation in rheumatology, Hôpital Beaujon 92110 Clichy, France; Rheumatology department, Hôpital Bichat, 75018 Paris, France (Submitted for publication January 24, 2001; accepted in revised form July 25, 2001)
Summary – The purpose of this review is to present current information on the natural course of lumbar spinal stenosis. As the population becomes older this condition is encountered more frequently. The diagnosis accuracy has improved and the number of cases detected is increasing internationally. Because of the relative unpredictability of surgical treatment, good knowledge of natural evolution and of the predictive factors influencing the course of the disease is crucial. Unfortunately, and in contrast with numerous surgical series few studies have dealt with natural evolution. In addition to anecdotal reports, a few non randomized studies will be reviewed. Only one randomized study has compared short and long term results of medical versus surgical treatment. Most of these studies are retrospective, with methodological flaws and are difficult to compare.. At the present time no scientifically based recommandations can be made to LSS. patients at diagnosis. Similarly predictors of success of medical and surgical treatment still need to be identified. However results of the studies analyzed in this review suggest that a substantial proportion of patients do not automatically deteriorate and will remain unchanged or even improved by medical means. They also suggest that patients with severe baseline symptoms, block stenosis and degenerative spondylolisthesis tend to require surgical decompression. Randomized studies with the necessary ethical precautions are needed to obtain clear-cut conclusions. Joint Bone Spine 2002 ; 69 : 450-7. © 2002 Éditions scientifiques et médicales Elsevier SAS lumbodynia / lumbar spinal stenosis / spinal surgery
The purpose of this review is to present current information on the natural course of lumbar spinal stenosis (LSS). Numerous studies have described the clinical presentation and evaluated the diagnostic usefulness of the various symptoms and signs. Much of the literature has focused on the diagnostic accuracy of imaging studies including myelography, computed tomography (CT) and magnetic resonance imaging (MRI). Cor-
* Correspondence and reprints.
roboration of the clinical findings by the results of imaging studies has benefited the diagnosis of LSS. As the general population ages, this condition is becoming more common, although its epidemiology remains largely unknown. Rates of surgical treatment are increasing internationally. However, the effectiveness of surgery for LSS is still a matter of controversy. According to a recent meta-analysis, surgical outcomes vary greatly across studies [1]. Some authors have reported deterioration of initial good results with time [2, 3]. Complications or symptom exacerbation can occur after
Lumbar degenerative spinal stenosis
surgery. At present, there is no scientific evidence that surgery is effective in LSS. Because of the relative unpredictability of surgical outcomes in the individual patient, a good knowledge of the natural history of LSS is crucial. It is also important to identify the factors that influence the course of the disease, such as age, sex, duration and type of baseline clinical symptoms, and site and severity of the stenosis. Unfortunately, although numerous surgical series have been published, studies of the natural history are scarce. A single randomized study compared short- and long-term results of medical and surgical treatment [4]. Such studies are obviously difficult to conduct in an older population with a high rate of comorbidity and mortality and with a variety of lifestyles and socioeconomic conditions. There are also ethical problems. For example, it would not be ethically acceptable to randomize an older patient with mild symptoms to surgery or a patient with severe pain and functional disability to medical therapy. To appraise current information on this topic, relevant data from the literature must be identified. Acceptable studies are few in number and difficult to compare given the absence of a severity index evaluating the clinical and imaging data. Moreover there is no consensus regarding the criteria for evaluating the results. ANECDOTAL REPORTS A few reports of successful nonsurgical treatment of LSS have been published. Jones and Thompson [5] studied 13 cases. Three patients were treated conservatively. Two of these were unchanged or improved, and the third was lost to follow-up [5]. Blau and Logue reported two cases treated without surgery; one patient was unchanged after 10 years and the other had a slight deterioration after 7 years [6]. Among 70 patients reviewed by Tile, two were considered for surgery but had a favorable outcome with medical treatment [3]. Rosomoff and Rosomoff reported the case of a 70-yearold man who had a complete L3-L4 block by myelography but recovered fully without surgery [7]. Postachini [8] studied 12 untreated patients with central LSS followed up for 3 to 12 years. All patients underwent myelography, CT, and MRI. The stenosis involved one level in seven cases and two or more levels in five cases. Compression was severe in seven cases and moderate in five. The radicular symptoms worsened in two patients with severe stenosis, remained unchanged in six, and resolved in four patients of whom three had
451
moderate stenosis. Finally, I was able to conduct telephone interviews of 22 patients treated conservatively, among many others, in the past 10 years. Seven had surgery. The others were either unchanged or improved. These anecdotal observations have no scientific value but indicate that some patients do not deteriorate with time and are able to tolerate their disability without surgical decompression. Moreover, there are a few nonrandomized long-term patient-based studies that describe the natural history of LSS and the efficacy of medical treatment [1-4]. STUDIES ON THE NATURAL HISTORY OF LSS – REVIEW OF THE LITERATURE I will now review the few available studies, in chronological order. The first attempt to study the natural history of LSS was made by Porter and associates in 1984 [9]. Entrapment of the lumbar root within the root canal was recognized using four criteria; 1) severe, constant root pain to the lower leg; 2) pain unrelieved by bed rest; 3) minimal tension signs; and 4) patients over 40 years of age. Two hundred forty-nine patients fulfilled these criteria. Radiological degenerative changes were seen in 80% of the patients. No active treatment was given in the majority of the patients, with the exception of back school attendance in 22%. Fourteen percent received one or more steroid epidural injections. Surgical decompression of the root canal was done in 24 (10%) patients, among whom only three had recovered fully at the 1-year evaluation, while 15 were moderately improved and six were unchanged. The course in the 225 (90%) patients treated without surgery was assessed by a mailed questionnaire, to which 75% replied. After 3 years, 78% still had some leg pain but most were not bothered enough to undergo surgery. The clinical criteria used in this study to identify root entrapment are now recognized as classical clinical signs of lateral stenosis. However, the diagnosis was based only on clinical symptoms and signs: in most cases confirmation of the diagnosis by imaging studies was not obtained. Myelography was performed only in the few patients who had surgery. The study was done before widespread use of CT and MRI became the rule. Johnson and associates [10] compared the clinical course of central LSS in 44 surgically treated patients and 19 conservatively treated patients. All patients underwent lumbar myelography with water-soluble contrast. The 19 nonsurgical patients were advised to have surgery but either declined or were excluded by
452
M. Benoist
the anesthesiologist. The patients were divided into 3 groups according to the treatment and to the severity of the stenosis at myelography: the 19 nonsurgical patients had moderate stenosis, and among the 44 surgical patients, 30 had moderate stenosis and 14 a complete block. The midsagittal diameter of the contrast column was similar in the nonsurgical group and in the surgical group with moderate stenosis. These two groups were, therefore, comparable. Mean follow-up was approximately 3 years in the nonsurgical group and 4 years in the surgical group. The visual analog scale scores are reported in table I. Most nonsurgical patients were unchanged or improved; only 10% got worse, as compared to 20% in the surgical group with moderate stenosis. Table II shows the results evaluated by a nonindependent surgeon. Interestingly, the proportion of improved patients (approximately 60%) was similar in the three groups. At last follow-up, no significant differences were found for pain severity or analgesic use. However, there was a significantly greater increase in walking capacity in the surgical group than in the nonsurgical group. Among the nonsurgical patients, 30% improved and 60% remained unchanged, whereas after surgery 60% of patients improved and 25% deteriorated. However, as indicated by the authors, these findings must be interpreted with a critical mind. First, this was not a randomized study. Second, baseline pain severity is not reported. Third, the baseline walking capacity was lower and the imaging study findings more severe in the surgical patients. Specifically, there were significantly more patients with degenerative spondylolisthesis in the surgical group with moderate stenosis Table I. Three-year visual analog scale assessment of 19 conservatively treated and 44 surgically treated patients (from [10]). n
Worse %
No surgery 19 Surgery, moderate stenosis 30 Surgery, severe stenosis 14
10 20 36
Unchanged Improved % % 58 7 0
32 57 64
(n = 12) than in the nonsurgical group (n = 5). Moreover, the outcome assessment was made by a nonindependent observer and was based on subjective opinion. In another article, Johnson et al. [11] reported the natural course of 32 patients studied prospectively. Mean follow-up after the myelogram was 4 years, and mean age was 60 years. Surgical decompression was recommended to all 32 patients, but 30 declined and two were excluded by the anesthesiologist. About 75% of the patients had neurogenic claudication, whereas the others had radicular pain or mixed symptoms. Outcome measures, completed by the patient, included two visual analog scales, one for the overall course and the other for pain severity. Results were also assessed by clinical examination. Walking capacity was evaluated by a visual analog scale and by measuring the symptomfree walking distance on a level surface. Five patients were lost to follow-up (of whom three died). Overall results at last follow-up are shown in table III. Four patients (15%) worsened and the 23 remaining patients were unchanged or improved. As shown in table IV, all patients still complained of back and/or leg pain, which was severe in three patients. Although the walking distance of the whole group was unchanged overall, it was worse in 30% of the patients. Overall, at last follow-up, 70% of the patients were unchanged, 15% were better, and 15% were worse. When the course was studied according to the midsagittal diameter of the contrast column, the narrowest diameters were found to be in the four patients whose symptoms worsened (table V). The natural course of 91 nonsurgically treated patients was reported at the 1996 annual meeting of the International Society for the Study of the Lumbar Spine, by Herno and associates [12]. This was a retroTable III. Four-year visual-analog scale results in 32 conservatively treated patients (from [11]). n patients 27
Worse %
Unchanged %
Improved %
15% 4
70% 19
15% 4
Table II. Three-year clinical examination assessment of 19 conservatively treated and 44 surgically treated patients (from [10]).
No surgery Surgery, moderate stenosis Surgery, severe stenosis
Worse %
Unchanged %
Improved %
1 0 1
37 36 29
58 63 64
Table IV. Pain severity after 4 years in 32 conservatively treated patients (from [11]). • Mild • Moderate • Severe
19% 70% 11%
Lumbar degenerative spinal stenosis Table V. Correlation between anterior-posterior diameter and clinical outcome in 27 conservatively treated patients (from [11]). VAS result
n patients
AP diameter (mm)
Worse Unchanged Improved
4 19 4 27
4.7 6.8 8.2
VAS: visual analog scale; AP diameter: anterior-posterior diameter.
spective study with a mean follow-up of 8 ± 3 years. The patients were divided into four groups according to the radiological findings: 11 had a complete block, 40 had moderate stenosis (anterior-posterior diameter < 10 mm), 18 had mild stenosis (10–12 mm), and 22 had lateral stenosis. Outcomes were evaluated based on the Oswestry questionnaire and on the treadmill walking capacity. The severity of pain in the back and legs before and after walking on the treadmill was assessed using a visual analog scale. At follow-up, the overall patients’ subjective evaluation showed that 27 patients were unchanged, 41 improved, and 23 worse. The mean Oswestry score was 31 ± 16. The score did not differ significantly according to the radiological findings or follow-up duration. Pain severity before and after the treadmill test did not differ between the radiological groups. No difference in walking capacity was noted among the four groups. The final conclusion of the study was that the natural course was benign in all 91 patients and that the subjective and physical manifestations were remarkably stable. In another study, Herno and associates used a matched-pair design to compare outcomes in surgically treated and nonsurgically treated patients [13]. They were able to form 54 matched pairs. The matching criteria included sex, age, symptom duration, and myelographic findings. The main manifestations were neurogenic claudication, leg pain, and mixed symptoms. The myelographic findings were graded in four categories: total block, subtotal block (anterior-posterior diameter < 10 mm), moderate stenosis (10–12 mm), and lateral stenosis. Outcomes were evaluated based on the Oswestry score and on an estimation of the functional status by clinical examination. The mean Oswestry score was similar in the overall population and in the women with and without surgery but was significantly better in the surgically treated men. Functional status was good in both groups. Outcomes were not correlated with age, site of the stenosis (central, n = 35; or lateral, n = 19), or presence or absence of coexisting
453
illness. This study has major limitations. As indicated by the authors, it is retrospective and the baseline levels of pain and disability are unknown. Moreover, in 37 nonsurgically treated patients, the pain was not severe enough to require an operation. This casts some doubt on the reliability of the matching. At the 1997 annual meeting of the International Society for the Study of the Lumbar Spine, the same authors [14] presented a longitudinal analysis of 38 nonsurgically treated patients with LSS. Nineteen women and 19 men were examined in 1989 and 1995. Most had one or several coexisting diseases. The radiological findings were graded according to the severity and site of the stenosis (central, n = 27; or lateral, n = 11). The mean Oswestry score of all 38 patients was 34.3 in 1989 and 33.4 in 1995. The overall alteration in the patients’ condition was evaluated based on four criteria: physical condition, ability to perform daily activities, walking capacity on a treadmill, and pain on a visual analog scale after the treadmill test. Taken together, these parameters show that the overall condition remained almost unchanged. However, the patients with a complete block at myelography worsened. Herno et al. concluded that patients with block stenosis require surgical decompression. Saal et al. [15] studied 52 patients with a follow-up ranging from 2 to 8 years. All patients had radiculopathy, walking capacity restriction, and evidence of central stenosis with or without lateral stenosis on the imaging studies. Treatment included analgesics, epidural injection, and physiotherapy. An independent observer used a questionnaire and telephone interview to assess outcomes. Thirty-three of the 52 patients had good pain control with non-narcotic analgesics and no or minimal walking restriction. The remaining patients received epidural and physical therapy. There was no neurologic deterioration. Four patients were treated surgically. Saal et al. concluded that surgery can be avoided and that medical and physical treatments are associated with better results than the natural history as reported by other authors. Unfortunately, no correlation was made in this study between the outcome and the clinical symptoms at diagnosis or the site and severity of the stenosis. Although nonrandomized, the Maine lumbar spine study is informative [16, 17]. One hundred and fortyeight patients were enrolled in this prospective cohort study performed by a group of surgeons in the state of Maine in the U.S.A. Choice of treatment was decided after discussion between the patient and surgeon.
454
M. Benoist
Eighty-one patients underwent laminectomy, accompanied in three cases with instrumented fusion. The 67 remaining patients received medical treatment by analgesics, epidural steroids, back-school exercises, and physiotherapy. Results were evaluated at months 3, 6, 12, 24, 36 and 48. The evaluation criteria included a symptom severity index based on frequency and severity of low back and leg pain (score from 0 to 24). Other outcome measures included the SF36, the Roland modified functional score, and a health questionnaire with five outcome categories (excellent, 1; bad, 5). Satisfaction with care was evaluated also. Tables VI and VII summarize the 1-year outcomes in the two groups. Overall, approximately 75% of the surgically treated patients were improved and satisfied, and 40% of the medically treated patients were improved. These data indicate a clear superiority of surgical treatment. However, the two groups are difficult to compare because, at baseline, the surgical patients had more severe pain, worse functional incapacity, and more extensive stenosis on imaging studies. The 119 (80%) patients evaluated after 4 years included 52 (78%) of the 67 nonsurgically treated patients and 81 (80%) surgically treated patients [16, 17]. Again, outcomes were better after surgery, but the difference was less marked, as only 70% of the operated patients were still improved and 63% satisfied. At the same follow-up evaluation, 52% of the medically treated patients were improved and 42% were satisfied with their care. Some additional information must be emphasized. First, 22% of the medically treated patients underwent surgery during the 4-year follow-up period, with less satisfactory results than in the patients who had surgery at baseline. Second, 51 of the 119 patients evaluated after 4 years had moderate symptoms at baseline. In this group, the 29 surgical patients had significantly better outcomes than did the 22 nonsur-
gical patients. The high rate of follow-up at 4 years and the good quality of the outcome measures are important strengths of this nonrandomized study. However, as pointed out by Nachemson [18], the imaging data are not reported in detail, so that it cannot be determined whether the severity of the stenosis on imaging studies was correlated with an obvious superiority of surgery over medical treatment. A multivariate analysis [16, 17] identified interesting predictive factors. Unilateral leg pain and an interval of less than 1 year from pain onset to the first physician visit predicted a good outcome after surgery. In contrast, predominance of low-back over radicular pain and severe deterioration in functional capacity were associated with worse outcomes. Hurri et al. [19] focused on the long-term prognosis. Seventy-five patients, out of a total of 134 patients with LSS diagnosed by myelography between 1978 and 1982, were evaluated in 1993 by telephone interview, using a structured questionnaire. Forty-eight had died and 11 could not be traced. Disability was assessed according to the Oswestry score and the subjective assessment by the patient (improved, unchanged, or worse). The stenosis was severe (7.0 mm) in 32 patients (26 surgically treated) and moderate (7–10.5 mm) in 43 patients (31 surgically treated). Tables VIII and IX summarize the subjective assessments in the two groups. Most medically treated patients were unchanged or improved. Worsening was usually correlated with a severe stenosis. Shortcomings of this study are the large number of patients lost to follow-up and the nature of the outcome measurements. Simotas et al. used conservative treatment in 49 patients for 16 to 55 months (mean, 33 months) [20]. Treatment was not standardized and included analgesics, epidural steroids, and physical therapy. Nine of the 49 patients had surgery, with a mean interval of 13
Table VI. Outcome of the non-operated group (from [2]).
Table VIII. Twelve-year subjective assessment of 18 conservatively treated patients (from [9]).
Low back pain Leg pain
Better %
Same %
Worse %
41 44
38 42
20 12
Table VII. Outcome of the operated group (from [2]). Low back pain Leg pain
Better %
Same %
Worse %
77 78
17 15
5 6
Worse
Unchanged
Improved
11%
45%
44%
Table IX. Twelve-year subjective assessment of 57 operated patients (from [9]). Worse
Unchanged
Improved
18%
19%
63%
Lumbar degenerative spinal stenosis
months after initiation of medical treatment. Five of the 40 remaining patients worsened, 17 were unchanged, and 26 were moderately improved. The outcome measures used in this study were excellent and included a condition-specific health status index, inspired from that of Stucki and well adapted to accurate evaluation of LSS. An effort was made to correlate the treatment results with the imaging study data, including severity and extent of stenosis and presence of degenerative spondylolisthesis or scoliosis. Most surgically treated patients had severe or multilevel narrowing, spondylolisthesis, or scoliosis. However, the small number of study patients studied precluded a valid statistical analysis. Furthermore, the surgical patients had more severe symptoms at baseline. PARTIALLY RANDOMIZED STUDY In a recent study by Amundsen et al. [4], a small group of patients were randomized to surgery or conservative therapy. One hundred patients with back and leg pain and LSS were included. Those patients with severe pain and disability were immediately directed to surgery (n = 19). On the other hand, the patients with mild symptoms underwent medical treatment including rest, bracing, back school, and physiotherapy and were encouraged to resume a normal lifestyle with their usual physical activity (n = 50). The 31 remaining patients had moderate clinical symptoms and were randomized, after informed consent, to surgery (n = 13) or conservative therapy (n = 18). Surgery consisted of laminectomy without fusion or instrumentation. An independent observer reevaluated the patients after 6 months and after 1, 4, and 10 years. Evaluation of results was based on the overall evaluation of pain on a visual analog scale, claudication distance, level of daily activities, and consumption of analgesics. The patients were asked to classify the results as better, unchanged, or worse. Together with the examining physician, overall results were considered good if there was total restitution or partial restitution with clear improvement over the initial condition and bad if there was either deterioration or no change. Among the patients randomized to surgery, one died and 11 of the remaining 12 patients had a good result after 6 months and 4 years. However, after 1 year, only nine patients had a satisfactory outcome. Only seven (39%) of the 18 patients randomized to conservative treatment had good results after 6 months and eight
455
after 4 years. Six patients underwent surgery 3 to 26 months after study inclusion. In the group of nonrandomized medically treated patients who had mild symptoms at baseline, 64% of 50 patients had a good result after 1 year and 57% after 4 years. Ten patients required surgery. In the group of 19 nonrandomized surgically treated patients, 89% had good results after 1 year and 84% after 4 years. There were no complications, and only one patient required repeat surgery. Surgery was performed eventually in 20 of the 68 medically treated patients (18 randomized and 50 nonrandomized), i.e., slightly more than one-third. Overall, the treatment result was significantly better in the patients randomized to surgery than in those randomized to medical treatment. However, it should be noted that adequate conservative therapy maintained a favorable outcome in approximately two-thirds of the nonrandomized patients with mild or moderate symptoms at baseline. In this series, the results of surgery after failure of conservative treatment were similar to those of first-line surgery. Fourteen patients died. All the other patients were reevaluated after 10 years and found to have no significant worsening of symptoms. The treatment results [4] were not correlated with clinical parameters such as age, sex, symptom duration, socioeconomic status, or physical findings. Similarly, no associations were found between clinical results and various imaging parameters such as spondylolisthesis, site of stenosis (central and/or lateral), severity of the narrowing, and whether one or more levels were affected. The small number of study patients does not allow definitive conclusions. CONCLUSIONS Data in the literature and our own experience suggest a number of conclusions, although several questions remain without clear answers. First, the overall results of surgery are better than those of medical treatment. This appears clearly in studies comparing the two treatments [4, 16, 17, 10]. The proportion of good results (70 to 75%) after surgery in these studies is similar to that already disclosed in other published series [21]. These results usually remain stable over time, with no major complications. Second, a substantial proportion of conservatively treated patients have good results that are sustained for up to 10 years. Consequently, and despite the apparent superiority of surgical treatment, in many cases surgery is not indispensable. Further-
456
M. Benoist
more, if the symptoms worsen, surgery performed after failure of medical treatment seems to yield similar results to those obtained in patients who have surgery earlier. However, and as mentioned above, the conclusions of the studies by Atlas et al. [16, 17] are less optimistic. Additional prospective studies are needed to conclude. For ethical reasons, such studies could proceed by stages, with interim analyses every 6 to 12 months. Patients with disabling pain and functional status would be randomized to surgery or no surgery. Predictors of success or failure of medical or surgical treatment need to be identified. The results of the studies discussed above suggest that patients with severe baseline symptoms, block stenosis, and degenerative spondylolisthesis tend to deteriorate over time and require surgical decompression. However, this has not yet been scientifically demonstrated. The main limitation of the studies analyzed in this review lies in the outcome evaluation. LSS is a specific subgroup of low back pain, and outcome measures adapted to its characteristics must be used to obtain an accurate evaluation of treatment efficacy and to allow a comparison with the natural course. I believe that future studies should use: – two visual analog scales, one for evaluate low-back pain and the other for radicular pain; – a condition-specific health status index, such as that proposed by Stucki [22]; – a generic health status index, such as the SF36, to evaluate both physical and nonphysical impairments [23];; – a psychological evaluation using a self-administered depression scale; and – a cumulative illness rating scale to detect potential comorbidities [24]. At present, it seems reasonable to follow the recommendations made by Amundsen. Treatment should start with well-organized and closely monitored conservative therapy. If the pain remains mild or moderate with no major disability, conservative care can be continued with fair chances of stability or even improvement. If the low-back and leg pain remains or becomes severe, surgery offers a good chance of stable improvement. Considering the present absence of predictive imaging factors, the indication for surgery must be based on pain and disability rather than on the severity of the stenosis. Ultimately, the treatment is chosen by the patient, who should be fully informed of the advantages and disadvantages of each treatment option. The choice also depends on the patient’s objectives, comorbidity,
cardiovascular status, general health condition, lifestyle, and socioeconomic status. Furthermore, medical treatment must be properly organized. Epidural steroids seem efficacious. In a recent randomized study, caudal epidural steroid injections showed efficacy after 3 months, although the beneficial effect faded between 6 and 12 months after the injection [25]. Further studies are needed to confirm these data and to determine the role for other medical treatment modalities such as orthotics and physical therapy. REFERENCES 1 Tuite GF, Stern JD, Doran SE, et al. Outcome after laminectomy for lumbar spinal stenosis. Part I : Clinical correlations. J Neurosurg 1994 ; 81 : 699-706. 2 Katz JN, Lipson SJ, Chang LC, et al. Seven to 10 year outcome of decompressive surgery for degenerative lumbar spinal stenosis. Spine 1996 ; 21 : 92-8. 3 Tile M, Mcneil SR, Zains RK, et al. Spinal stenosis: results of treatment. Clin Orthop 1976 ; 115 : 104-8. 4 Amundsen T, Weber H, Nordal H, et al. Lumbar spinal stenosis: conservative or surgical management. A prospective 10 year study. Spine 2000 ; 25 : 1424-35. 5 Jones RAC, Thompson JLP. The narrow lumbar canal. A clinical and radiological review. J Bone Joint Surg 1968 ; 50 B : 595-605. 6 Blau JN, Logue V. The natural history of intermittent claudication of the cauda equina. Brain 1978 ; 101 : 211-22. 7 Rosomoff H, Rosomoff R. Nonsurgical aggressive treatment of lumbar spinal stenosis. Spine 1987 ; 3 : 383 p. 8 Postacchini F. Lumbar spinal stenosis. Long-term results. In: Wiesel SW, Weinstein JN, Herkowitz H, Dvorak J, Bell G, Eds. The lumbar spine, Volume 2. Philadelphia: WB Saunders; 1990. p. 766-81. 9 Porter RW, Hibbert C, Evans C. The natural history of root entrapment syndrome. Spine 1984 ; 9 : 418-21. 10 Johnsson Ke, Uden A, Rosen I. The effect of decompression on the natural course of spinal stenosis. A comparison of surgically treated and untreated patients. Spine 1991 ; 16 : 615-9. 11 Johnsson Ke, Rosen I, Uden A. The natural course of lumbar spinal stenosis. Clin Orthop 1992 ; 279 : 82-6. 12 Herno A, Nevalainen S, Saari T, et al. The natural course of 91 nonoperated patients with lumbar spinal stenosis. Presented at the annual meeting of the International Society for the Study of the Lumbar Spine. Burlington 1996; June : 25-9. 13 Herno A, Airaksinen O, Saari T, et al. Lumbar spinal stenosis: a matched pair study of operated and nonoperated patients. Br J Neurosurg 1996 ; 10 : 461-5. 14 Herno A, Nevalainen S, Saari T, et al. The longitudinal analysis of 38 monoperated patients with lumbar spinal stenosis. Presented at the annual meeting of the International Society for the Study of the Lumbar Spine. Singapore 1997; June : 2-6. 15 Saal JS, Saal JA, Parthasarathy R, et al. The natural history of lumbar spinal stenosis. The results of nonoperative treatment. Presented at the annual meeting of the International Society for the Study of the Lumbar Spine. Singapore 1997; June : 2-6. 16 Atlas SJ, Deyo RA, Keller RB, et al. The Maine Lumbar Spine study. Part III. Year outcomes of surgical and nonsurgical management of lumbar spinal stenosis. Spine 1996 ; 21 : 1787-95. 17 Atlas SJ, Keller RB, Robson D, et al. Surgical and non surgical
Lumbar degenerative spinal stenosis
18 19 20 21
management of Lumbar Spinal Stenosis. Four year outcomes from the Maine Lumbar Spine Study. Spine 2000 ; 25 : 556-62. Nachemson A. Point of view. Spine 1996 ; 21 : 1794-5. Hurri H, Slatis P, Soini J, et al. Lumbar spinal stenosis: assessment of long-term outcome 12 years after operative and conservative treatment. J Spinal Disord 1998 ; 11 : 110-5. Simotas A, Dorey F, Hansray K. Non operative treatment for lumbar spinal stenosis. Spine 2000 ; 25 : 197-204. Turner JA, Ersek K, Herron L, Deyo R. Surgery for lumbar spinal stenosis. Attempted meta-analysis of the literature. Spine 1992 ; 17 : 1-8.
457
22 Stucki G, Daltroy L, Liand M, et al. Measurement properties of a self-administered outcome measure. Spine 1996 ; 21 : 796803. 23 Linn BS, Linn M, Currel L. Cumulative illness rating scale. J Am Geriatr Soc 1968 ; 16 : 622-6. 24 Ware JE, Sherbourne CD. The MOS 36 item short form health survey (SF36). I. Conceptual frame work and item selection. Med Care 1992 ; 30 : 473-83. 25 Fortin L, Bergeron Y, Lambert R, et al. Double blind randomized clinical trial of epidural injections in spinal stenosis. Arthritis Rheum 2000 ; 43 (9 Suppl).