The factors of deterioration in long-term clinical course of lumbar spinal canal stenosis after successful conservative treatment

Lumbar spinal canal stenosis (LSS) is a disease characterized by one or several symptoms such as low back pain, pain and/or numbness in the lower extremities, and neurogenic claudication that occurs secondary to spinal canal narrowing [1]. The treatment of LSS depends on symptom severity. The major indications for surgery are as follows: intolerable pain in daily activities despite adequate conservative treatment, progressively limited walking distance or standing endurance, and major neural deficits or evolving/progressive neural damage such as cauda equina syndrome, drop foot, and bowel and/or urinary dysfunction [2‐6]. On the other hand, in the absence of severe symptoms such as severe motor disturbances or bowel and/or urinary dysfunction, conservative treatment is generally the first choice for the treatment of LSS [7‐10]. Components of conservative treatment vary but generally include bed rest, medication or injection of anti-inflammatory analgesics, physical therapy or reduction of postural lordosis, and muscle weakness treatment with reportedly good outcomes [9, 11‐13]. However, we experienced cases of worsening symptoms even after successful conservative treatment. Although there have been several reports regarding prognostic factors for the conservative treatment of LSS, studies with long-term follow-up have rarely been reported [9, 14, 15]. In addition, these reports did not describe the time course of the progress of symptom deterioration: when and how the symptom deteriorated in the patients with negative factors after successful conservative treatment in spite of its importance for decision-making around treatment. This study aimed to investigate the clinical course and prognosis of LSS after successful conservative treatment and identify the prognostic factors for predicting aggravation risk.

Thirty-four patients (56.7%) preserved the conditions at discharge until the latest follow-up, while 26 patients (43.3%) showed deterioration during the follow-up period. The average age of the patients who showed deterioration during the follow-up period (67.8 ± 6.9 years) was significantly higher than that of the patients whose conditions at discharge were preserved (62.5 ± 8.8 years). Furthermore, the patients with severe intermittent claudication, muscle weakness, and redundant nerve root tend to be significantly worse of their symptoms. On the other hand, there were no significant differences in sex, clinical LSS classification and the presence of vertebral body slip, scoliosis, and block of myelography between the preservation group and deterioration group (Table 1). In the preservation group, 3 patients had no symptoms, 20 patients had minor symptoms such as slight low back pain or numbness of the lower extremities, and 11 patients needed only medication therapy at the latest follow-up. On the other hand, of the patients with deterioration, 16 patients had a decreased intermittent claudication distance more than 50 m and 10 patients had newly developed or progressive paralysis. Subsequently, 19 patients received surgical intervention. The patients who got worse of their symptoms after discharge and then were performed conservative treatment were 3 patients of 26 patients with deterioration. The inpatient conservative treatment also consisted of pelvic traction, the application of a body cast, and an epidural steroid injection such as an epidural or selective nerve root block as described before. All these 3 patients received conservative treatment only once; however, 2 patients of them received surgical intervention after the second conservative treatment. Although the probability of symptom preservation after the successful conservative treatment of LSS gradually decreased over time, it remained 68.3% at 5 years post-discharge and 57.2% at 7 years post-discharge (Fig. 1). The number of patients who were able to follow up was 39 to 7 years after discharge. The outcomes of conservative treatment were worse in patients with severe intermittent claudication at the initial visit (≤ 100 m), cauda equina or mixed-type LSS, lower limb muscle weakness, and block on myelography than those in patients with mild intermittent claudication (between 100 and 500 m) (p = 0.004), radicular-type LSS (p = 0.039), no muscle weakness (p = 0.049), and no block on myelography (p = 0.045), respectively (Fig. 2a–d). The probability of symptom preservation at 5 years and median preservation expectancy in patients with severe intermittent claudication (≤ 100 m) were 38.9% and 4.1 years, respectively, whereas those in patients with mild intermittent claudication (between 100 and 500 m) were 81.1% and > 10 years. Similarly, in patients with the cauda equina and mixed types, the 5-year preservation probability and preservation expectancy were 52.4% and 5.8 years, respectively, whereas those in patients with the radicular type were 76.9% and 9.5 years. On the other hand, the 5-year preservation probabilities in patients with and without muscle weakness were 63.6% and 71.1%, although preservation expectancies were 5.8 years and > 10 years. In patients with and without block on myelography, the 5-year preservation probabilities were 60.0% and 74.3%, whereas preservation expectancies were 5.8 years and 9.5 years (Table 2). Age, vertebral body slip (≥ 3 mm), scoliosis (Cobb angle ≥ 10°), and redundant nerve roots of the cauda equina did not significantly influence symptom deterioration (Fig. 3a–d). Logistic regression analysis demonstrated that only severe intermittent claudication at the initial visit (≤ 100 m) was a significant risk factor incurring a poor outcome even after successful conservative treatment for LSS (p = 0.005, odds ratio = 6.665, correlation coefficient = 0.492) (Table 3). A post hoc power analysis demonstrated that the present study was enough powered with a power of 0.861 to detect differences between the patients with poor prognosis and good prognosis and that required sample size was 16/38 patients for the patients with poor/good prognosis, respectively.

Several reports have compared conservative and surgical treatments for LSS [14, 17, 18]. Amundsen et al. reported that the long-term results were better for surgery in a prospective 10-year follow-up study. However, more than half of the conservatively treated patients had a satisfactory outcome, and delaying surgery for some months did not worsen the prognosis. Therefore, a primarily conservative approach is recommended [14]. In their prospective 8- to 10-year follow-up study, Atlas et al. reported that low back pain relief, predominant symptom (either back or leg pain) improvement, and satisfaction were similar among patients who initially underwent surgery and those treated conservatively [17]. The randomized controlled trial study by Malmivaara et al. reported that patients treated both surgically and conservatively showed improved clinical outcomes in a 2-year follow-up [18]. Therefore, an initial conservative treatment approach seems advisable for appropriately selected patients.

Actually, we showed 56.7% of LSS patients treated by inpatient intensive conservative treatment maintained their condition at discharge for at least 5 years, average 7.3 years follow-up. Although there have been differences in evaluation methods or primal symptom severity, the reported outcomes of conservative treatment for LSS are generally good [7‐10]. Johnson et al. reported 32 LSS patients who had conservative treatment, and the mean follow-up period was 4.1 years. Fifteen percent improved, 70% remained unchanged, and 15% deteriorated as assessed by the evaluation with the visual analog scale while 47% improved, 37% remained unchanged, and 16% deteriorated as assessed by the evaluation with their clinical examination [7]. Swezey reported 36 LSS patients who had conservative treatment by intermittent traction and/or epidural steroid injection, and the mean follow-up period was 5 years. Forty-three percent improved and 30% remained unchanged as assessed by a subjective assessment by using a telephone interview [10]. Miyamoto et al. reported 120 LSS patients who had conservative treatment by pelvic traction, application of body cast, and/or epidural steroid injection, and the mean follow-up period was 7.9 years. Forty-three percent improved, 17% remained unchanged, and 40% deteriorated as assessed by using a general evaluation with a modified JOA score [8]. Simotas et al. reported 49 LSS patients who had conservative treatment by exercise and analgesics and/or epidural steroid injection, and the mean follow-up period was 2.8 years. Forty-seven percent improved, 24% remained unchanged, and 29% deteriorated as assessed by using the pain/function score [9]. Our inpatient intensive conservative treatment was applied to the patients who failed to get satisfied with symptom relief by outpatient conservative treatment and demonstrated to be effective even for such persistent cases. Therefore, we believe that such inpatient conservative treatment should be positioned between outpatient conservative treatment and surgery.

The predictive factors for the prognosis of conservative treatment for LSS have been investigated in several studies [8, 9, 14, 15]. Although Herno et al. [15] and Amundsen et al. [14] failed to identify any predictive factors, Simotas et al. referred to the presence of degenerative scoliosis and advanced age as negative factors for the outcomes of conservative treatment [9], and Miyamoto et al. cited degenerative scoliosis as a negative factor in outcomes and reported that radicular LSS was a positive factor for outcomes [8]. However, these reports described only the prognostic factors at final follow-up and lacked the information regarding the time course and risk factors of symptom deterioration. It is indispensable to know when and how many patients experienced symptom deterioration after successful conservative treatment for making a decision whether or not to have surgery without trying intensive conservative treatment.

Here, we investigated the prognosis of patients with LSS after successful conservative treatment. Cauda equina or mixed type, severe intermittent claudication at the initial visit (≤ 100 m), lower limb muscle weakness, and the block on myelography were the prognostic factors predicting the risk of aggravation. Surgery should be considered for patients with severe intermittent claudication.