Villavicencio, Alan T. MD^*,†; Nelson, E. Lee MD^*; Rajpal, Sharad MD^*,*; Burneikiene, Sigita MD^*,†

Success rates of surgical interventions for lumbar disorders vary significantly depending on multiple factors and among them the duration of symptoms. A fairly large body of literature has been published that examines the impact of the symptom duration on clinical outcomes in patients undergoing conservative or surgical treatment for different conditions of low back pain. To summarize, patients undergoing microdiscectomy for lumbar disk herniation should have surgery performed as early as clinically possible to have improved clinical outcomes, including resolution of sciatica symptoms. The same criteria cannot be applied to patients undergoing fusion surgery, because it is not reasonable to expect that fusion surgeries will be performed or even recommended within the same time frame. Further, there is no consensus whether the duration of symptoms adversely impact outcomes in patients undergoing decompression and fusion surgeries for radicular pain symptoms due to degenerative disk disease. To some degree these conclusions are explained by the fact that nerve compression develops more gradually in case of, for example, foraminal narrowing versus an acute onset in disk herniations. To our knowledge, the duration of symptoms in regards to resolution of radicular symptoms in patients undergoing lumbar decompression and fusion has not been studied.

The amount of connective tissue varies between the nerves. The lack of perineurium and poorly developed epineurium makes spinal nerve roots especially susceptible to mechanical compression. The nerve compression histopathologic changes progress from segmental demyelination to axonotmetic injury, which are still fully reversible followed by a gradual development of a variable degree of scar tissue in the connective tissue sheath. The irreversible nerve damage with associated impaired function depends on the amount of scar tissue. Eliminating mechanical compression also diminishes the effect of nerve root sensitization caused by inflammatory mediators. Because nerve injury correlates with the degree and duration of nerve compression, the duration of symptoms should also play a role in resolution of radicular symptoms. It is not clear whether there is a “cutoff” time when decompression and fusion surgery becomes less effective in the conditions with chronic nerve root compression symptomatology.

The main objective of this prospective observational study was to analyze whether duration of symptoms has any effect on clinical outcomes, including resolution of radicular pain symptoms due to degenerative disk disease and stenosis with spondylolisthesis in patients undergoing transforaminal lumbar interbody fusion (TLIF).

METHODS

A total of 84 patients were enrolled into a prospective observational study. The patients with predominantly radicular pain symptoms due to degenerative disk disease and stenosis with spondylolisthesis, with no previous fusion surgeries and undergoing 1- to 3-level TLIF surgeries, were included. All patients met the criteria for fusion due to the presence of instability, loss of disk space or foraminal height causing structural impingement from the pedicles requiring stabilization. Some patients also needed fusion due to iatrogenic instability as a result of an aggressive decompression/facetectomy if stenosis was present. Fifteen patients (18%) were lost to follow-up and were excluded from this analysis leaving a total of 69 patients.

Standardized questionnaires were used to analyze clinical outcomes and were administered preoperatively within 3 months of scheduled surgery, and postoperatively at 3, 6, 12, and 24 months. Two separate scales (11-point numerical rating scale) were used to evaluate the severity of low back and leg pain. Patient functional outcomes were assessed using the Physical Component Summary (PCS) scale of the 36-Item Short-Form (SF-36) questionnaire and Oswestry Disability Index (ODI) was used to evaluate chronic disability and activities of daily living. The patients also completed a self-reported patient satisfaction with results survey (Fig. 1). Answers were scored on a scale from 0 to 100: 100=very satisfied/much better/definitely yes; 75=somewhat satisfied/better/probably yes; 50=don’t know/same/don’t know; 25=somewhat dissatisfied/worse/probably no; 0=dissatisfied/much worse/definitely no. A total score was calculated for each patient by averaging the scores from all 3 responses.

Figure 1

Clinical outcomes are generally evaluated and compared using average scores calculated from individual scores. Although often the results are statistically significant, they may not fully reflect clinical importance and significance to the patients undergoing surgery. The minimal clinically important difference (MCID) values represent the patients’ perception of adequate improvement in health status and pain control. In a paper by Copay et al,¹ the MCID for ODI, SF-36/PCS, and back/leg pain scale was defined for lumbar spine surgery patients. The following reference MCID values were used to evaluate clinical outcomes: 12.8 points for ODI, 4.9 points for SF-36 PCS, 1.2 points for back pain, and 1.6 points for leg pain. Because MCIDs could be dependent on the baseline scores,^1,2 we used the change between the postoperative and baseline scores as a starting point in our study. The change scores were calculated by subtracting the postoperative scores (Post-) from the baseline scores (Pre-). To emphasize the change in clinical outcome scores, the relevant scores were calculated as the ratio of MCID values and change scores (eg, MCID/PostODI−PreODI). Such reporting should allow more accurate assessment of different interventions on a patient level.

Statistical Analyses

Multiple regression analyses were conducted to examine the relationship of the duration of symptoms and relevant MCID ratio values while controlling for independent variables. The proportions of patients that reached MCID for each of the measured clinical outcome parameter were also calculated. Further, the 2 groups of patients were compared to analyze the changes of clinical outcomes for the patients who underwent fusion surgeries within <24 months versus ≥24 months. The comparison between groups was made using the Student t tests for all independent continuous quantitative variables. Categorical values were compared using the χ² analysis.

RESULTS

The mean age of patients was 59.8 years (range, 37–77 y); there were 39 (56%) females and 30 (44%) males. Clinical diagnoses included radiculopathy (100%) and spinal stenosis with spondylolisthesis (52%). The selected demographic, clinical, and surgical patient characteristics for the patient groups that underwent fusion surgery within 2 or >2 years of becoming symptomatic are presented in Table 1. There were no statistically significant differences detected in the demographic, clinical, or surgical characteristics between the patient groups.

Table 1

Overall, a total of 55 (80%) patients had reached MCID and had back pain improvement by at least 1.2 points (mean ratio 4.1; range, 1.7–8.3); 9 (13%) patients did not reach MCID (mean ratio 0.4; range, 0–0.8) and 5 (7%) patients rated their back as worse than before surgery (mean ratio −1.7; range, −3.3 to −0.8).

In regards to radicular symptoms, 54 (78%) patients had reached MCID and had leg pain improvement by at least 1.6 points (mean ratio 3.4; range, 1.3–6.3); 10 (15%) patients did not reach MCID (mean ratio 0.25; range, 0–0.6) and 5 (7%) patients rated their radicular pain as worse than before surgery (mean ratio −0.7; range, −1.3 to −10.6).

Improvement of at least 12.8 points on ODI scale was noted in 46 (67%) patients (mean ratio 2.2; range, 1–5.2); 15 (22%) patients did not reach MCID (mean ratio 0.5; range, 0–0.9) and 8 (11%) patients had increased disability scores (mean ratio −0.6; range, −1.6 to −0.2).

A total of 52 patients reported at least 4.9 points improvement of SF-36 PCS scores (mean ratio 3.2; range, 1–8.3); 8 (12%) patients did not reach MCID (mean ratio 0.4; range, 0.04–0.7) and 9 (13%) patients had reduction in their physical scores (mean ratio −1.1; range, −2 to −0.1).

The association between duration of symptoms and clinical outcomes was further analyzed by comparing 28 patients who were treated within <24 (mean 7.8; range, 1.5–18) months of becoming symptomatic and 41 patients who were treated at ≥24 (mean 124.4; range, 24–480) months of becoming symptomatic (Table 2). The patients with shorter duration of symptoms had significantly better radicular symptom resolution (Students t test, P=0.032) compared with patients who waited at least 24 months or longer to undergo fusion. The percentage of patients who reached the MCID in leg pain resolution was also higher (89%) in the patient group who had surgery in <24 months.

Table 2

Multiple regression analyses were performed to examine the relationship of the duration of symptoms and relevant MCID ratio values controlling for independent quantitative variables (age, sex, previous surgeries, TLIF levels, complications, reoperations, and fusion status) to determine correlations with duration of symptoms. This analysis showed that the duration of symptoms was a significant predictor of better radiculopathy symptom resolution (P=0.018), but not for back pain resolution (P=0.27), or improvement in ODI (P=0.10) and SF-36 PCS scores (P=0.19).

DISCUSSION

TLIF is an effective treatment for lumbar degenerative disorders; however, surgical outcome success rates vary significantly depending on multiple factors. The most important finding of this study was that the shorter duration of symptoms was a statistically significant predictor for resolution of predominantly radicular pain symptoms in patients undergoing TLIF for radiculopathy due to degenerative disk disease and stenosis with spondylolisthesis.

Although the shorter duration of symptoms was previously identified as one of the factors associated with improved clinical outcomes and better resolution of radicular symptoms, the majority of studies analyzed patients undergoing microdiscectomy for lumbar disk herniations.^3–7 The optimal timing for microdiscectomy was suggested to be <2 months,^6,8 and no longer than 12 months,^3,4,7,9 which generally aligns with clinical practice.

A retrospective “as-treated” analysis of the Spine Outcomes Research Trial (SPORT) demonstrated that the patients with <6 months’ symptom duration and undergoing microdiscectomy for lumbar disk herniation had improved clinical outcomes following surgical or conservative treatment, but surgery was more effective than nonsurgical treatment at all follow-up time intervals.⁵

Silverplats et al⁷ demonstrated that patients who had <6 months of leg pain before undergoing microdiscectomy had better clinical outcomes based on Macnab classification and patient satisfaction compared with the patients with symptom duration between 6 and 12 months or longer than 12 months. The authors thought that microdiscectomy should be performed after a relatively short waiting period if conservative treatment is ineffective. Similar conclusions were reached and even a lower threshold established in a study performed by Nygaard et al.⁴ They thought that patients undergoing microdiscectomy after 8 months of having sciatica symptoms might be at risk of developing chronic pain and recommended 2–8 months as an optimal timing. Ng and Sell³ found the association between >12 months of sciatica symptoms and less satisfactory outcomes. Although the authors did not find any correlations between the postoperative pain scores and duration of sciatica, the 0.6% change of ODI scores was noted for an increase over 1 month in the duration of symptoms. Rothoerl et al⁶ recommended conservative treatment for up to 2 months before considering surgery based on the results of their study, which demonstrated that better clinical outcomes were achieved for shorter duration of radicular symptoms or sensory deficits, but not for motor deficits.

There were also studies that assessed optimal timing of surgery and compared clinical outcomes in patients undergoing microdiscectomy versus conservative treatment. A prospective trial analyzed data according to an “intention-to-treat” principle and did not find that patients undergoing microdiscectomy for sciatica symptoms at 6–12 weeks had better functional recovery rates than patients randomized to conservative treatment.¹⁰ However, 39% patients in the conservative treatment group had surgery after randomization. The crossover rates were not accounted for in the original study design and most likely the study lacked the effective power to detect any potential differences that may have existed between the treatment groups.¹¹ At the 5-year follow-up, the authors noted that later surgery may have been less effective in a small group of patients compared with early surgery, which they attributed to “more chronic changes around disk protrusion or sequester, causing more difficulty in freeing the nerve from compression.”¹²

About 40% of patients (out of 380) had suboptimal clinical outcomes at 2-year follow-up in a prospective multicenter observational study of patients with disk herniation and sciatica symptoms.¹³ Although discectomy was performed in only 32% of these patients, the ones who had the duration of back pain longer than 1 year and sciatica longer than 3 months were about twice as likely to report nonsuccess.

There is less agreement in the studies analyzing clinical outcomes in regards to duration of symptoms for spinal stenosis and degenerative spondylolisthesis. A retrospective “as-treated” analysis of the SPORT demonstrated that there was improvement of clinical outcomes noted in patients with <12 months compared with >12 months of symptom duration who were treated conservatively or surgically for spinal stenosis, but not for the patients with degenerative spondylolisthesis.¹⁴ The authors suggested that the dynamic nature of nerve compression in spondylolisthesis resulted in less long-term nerve ischemia and demyelination. In our opinion, these groups were not quite homogenous, both with regards to procedures performed (the majority of patients in the degenerative spondylolisthesis group had fusion vs. decompression surgery in the spinal stenosis group) and surgical versus conservative treatment modes (about 1/3 of the patients in each group had conservative treatment only). Although the analysis comparing treatment effect between surgical and nonsurgical groups did not reveal any statistically significant differences, the study essentially analyzed conservative and 2 different surgical treatment methods for 2 slightly different clinical syndromes. The sample size in our study was too small, so we could not completely avoid this second limitation. However, all our patients underwent TLIF surgeries and we were able to demonstrate that shorter duration of symptoms predicted better resolution of radicular pain symptoms for radiculopathy due to degenerative disk disease and stenosis with spondylolisthesis. Another very significant finding of the SPORT study was that duration of symptoms >12 months was the only specific risk predictor for reoperation.¹⁵ The incidence rate of reoperation was 13% at 4 years and the majority of patients underwent repeated surgeries due to recurrent stenosis or progressive spondylolisthesis. These patients also had less improvement in outcome and permanent neuropathy or incorrect diagnoses were suggested as a cause for this.

Franklin et al¹⁶ evaluated lumbar fusion outcomes in 388 injured Washington state workers’ compensation patients and, although clinical outcomes are known to be generally worse in this patient population, shorter time from injury to surgery predicted better functional outcomes as measured by disability status 2 years after fusion surgery. The majority of these patients had disk herniation, spinal stenosis, and instability symptoms. These conclusions were in agreement with the study results by Sigmundsson et al¹⁷; however, in the majority of the studies that previously analyzed surgical treatment of stenosis, decompression without fusion was performed.

Other studies examining clinical outcomes after decompression surgeries for spinal stenosis did not find any correlations with duration of symptoms,^18,19 but a disagreement exists as to what is considered an early surgery. Amundsen et al²⁰ concluded that outcomes of delayed surgery (up to 3 y) were not significantly inferior and a meta-analysis performed by Niggemeyer et al²¹ determined that better results were achieved if decompression was performed in the first 7 years and instrumented fusion had better outcomes if patients were symptomatic for ≥15 years.

This study was designed as prospective observational, but it was not randomized and potentially was susceptible to various biases. One of them being that there is a possibility that by waiting longer the patients with more advanced disease are selected, therefore they have inferior clinical outcomes. There were no statistically significant intergroup differences detected for patient demographic, clinical, or surgical parameters, but we did not categorize and evaluate any other morphologic or pathophysiological changes beyond these variables. Although such correlations may exist, but this would not diminish the importance of the study conclusion. In addition to the fact that compressed nerves undergo irreversible structural changes due to more prolonged compression, the study results support the theory that the longer duration of pain results in physical deconditioning, psychological distress, and other psychosocial factors that are playing role in pain centralization mechanisms. This makes any surgical intervention less successful and adversely impacts clinical outcomes.

CONCLUSIONS

A shorter duration of symptoms was found to be a statistically significant predictor for better resolution of radicular symptoms in patients undergoing TLIF for radiculopathy due to degenerative disk disease and stenosis with spondylolisthesis.

REFERENCES