Surgery Eases Pain of Spinal Stenoses and Slipped Vertebra, but Sciatica Outcomes Similar

LEBANON, N.H., May 31 -- Surgery for degenerative spondylolisthesis with spinal stenosis is significantly more effective against pain than usual nonsurgical care, confirmed one large study.

But, another study found early surgery for sciatica accelerated pain relief but there were similar outcomes at one year compared with conservative treatment.

"Surgery is widely used, but its effectiveness in comparison with that of nonsurgical treatment has not been demonstrated in controlled trials," wrote James N. Weinstein, D.O., of Dartmouth here, and colleagues, in their study of spinal stenosis and slipped vertebra treatment.

These studies are helping to refine practice in the face of steadily rising spine surgery rates, commented Richard A. Deyo, M.D., M.P.H., of the University of Washington in Seattle. His editorial accompanied both studies in the May 31 issue of the New England Journal of Medicine.

Back surgery may not be necessary to preserve life or function for these patients without major neurologic deficits, but appropriate surgery may provide valuable pain relief, Dr. Deyo wrote.

Whether this justifies surgery has to be decided jointly by well-informed patients and physicians, he added.

Dr. Weinstein's analysis was part of the larger Spine Patient Outcomes Research Trial (SPORT), which aims to provide evidence for appropriate treatment of the most common causes of back pain.

The current study included a cohort of 304 patients randomized to standard decompressive surgery (95% with bilateral single-level fusion) or usual care and an observational cohort of 303 patients who chose treatment with their physician.

Usual care included physical therapy (for 42% of patients), epidural steroid injections (45%), opioids (34%), and nonsteroidal anti-inflammatory agents (51%).

However, extensive crossover between surgical and nonsurgical groups in both cohorts hampered the study. About half of participants assigned to usual care underwent surgery within two years of the start of the study; only 64% of those assigned to surgery actually had it. Of patients who initially chose surgery, just 25% had surgery within two years.

So the researchers pooled both cohorts and did an as-treated analysis with 372 patients having undergone surgery within two years after enrollment and 235 having received only nonsurgical treatment.

All 607 patients had spinal stenosis (60% severe) and degenerative spondylolisthesis, proven by imaging, as well as neurogenic claudication (85%) or radicular leg pain (77%). Symptoms had to have persisted for at least 12 weeks. All participants were surgical candidates.

While the observational and randomized cohorts were fairly similar, there were important differences between treatment groups. So the researchers controlled for covariates, including age, sex, work status, duration of current symptoms, reflex deficit, levels of disease, and baseline scores for each of the measured indices.

The intent-to-treat analysis showed no significant differences between treatment groups in the randomized cohort.

In contrast, the as-treated effects significantly favored surgery for all primary and secondary outcomes at all time points (P<0.001). Treatment effects for surgery versus nonsurgical therapy at two years were:

• 18.1 for bodily pain on the Short Form-36 questionnaire (95% confidence interval 14.5 to 21.7).
• 18.3 for physical function on the Short Form-36 (95% CI 14.6 to 21.9).
• ?16.7 for disability on the Oswestry Disability Index (95% CI ?19.5 to ?13.9).
• ?4.9 for Stenosis Bothersomeness Index scores (95% CI ?6.0 to ?3.8).
• ?1.5 for Leg Pain Bothersomeness Scale scores (95% CI ?1.8 to ?1.1).
• ?1.2 for Low Back Pain Bothersomeness Scale scores (95% CI ?1.3 to ?0.7).
• 36.6% reported they were "very" or "somewhat" satisfied with their symptoms (95% CI 28.0 to 45.1).
• 50.0% reported "major improvement" in progress (95% CI 42.2 to 57.9).
• Nearly identical results in the randomized and the observational cohorts.

Importantly, the researchers said, there appeared to be little evidence of harm for surgical or nonsurgical treatment. No patient in either group had cauda equina syndrome as a consequence of nerve root compression. Most surgical patients had no operative complications (89%).

"Often patients fear they will get worse without surgery, but the patients receiving nonsurgical treatment, on average, showed moderate improvement in all outcomes," they wrote.

They cautioned that their findings cannot be used to make direct comparisons between surgery and any specific nonsurgical treatment and vice versa.

In his editorial, Dr. Deyo likewise noted that "caution is in order," because combining the groups in an as-treated analysis essentially turned the study into a single large cohort study, which even when well designed may contradicted randomized, controlled trial results.

The "true magnitude of the surgical advantage probably lies somewhere between the two estimates" of as-treated and intent-to-treat effects, Dr. Deyo added.

As reported in November 2006, a separate analysis of the same trial found little difference between surgical and nonsurgical treatment for disk herniation with sciatica.

But because that analysis likewise suffered crossover problems that could have confounded the results, researchers in the Netherlands conducted a separate randomized, controlled trial in 283 patients with severe sciatica for six to 12 weeks.

Wilco C. Peul, M.D., of Leiden University Medical Center in Holland, and colleagues, randomized 141 patients to early microdiskectomy and 142 to prolonged conservative treatment with surgery later if needed.

Conservative treatment was provided by family doctors guided by research nurses who participated in pain management. Therapy was targeted to enable patients to resume daily activities. It included guidelines-directed prescription of pain medication and referral to a physiotherapist for patients afraid to move.

For those with sciatica persisting at six months or increasing, refractory leg pain or neurologic deficits at any point, microdiskectomy was offered. Thirty-nine percent underwent surgery at a mean of 18.7 weeks.

Most patients underwent early surgery as assigned -- 89% at a mean of 2.2 weeks. The treatment groups had similar baseline characteristics.

Among the findings, the researchers reported:

• No significant difference in Roland Disability Questionnaire scores in the first year (P=0.13).
• Twice as fast leg pain relief with early surgery as with conservative treatment (P<0.001).
• Faster perceived recovery -- complete or nearly complete disappearance of symptoms -- for patients in the early surgery group (4.0 versus 12.1 weeks, hazard ratio 1.97, 95% CI 1.72 to 2.22, P<0.001).
• Identical likelihood of perceived recovery at one-year follow-up (about 95% for both groups).
• No difference in visual analog pain scores between groups at one year.
• Similar pain, recovery, and disability score improvements for patients with early and delayed surgery at one year (P=NS).

"Thus, the major advantage of early surgical treatment is faster relief of sciatica," the researchers concluded.

They cautioned, though, that the additional support of research nurses in the conservative therapy group did not reflect usual care and may limit the generalizability of the findings.

"So who needs back surgery?" Dr. Deyo asked.

"Absent major neurologic deficits, patients with herniated disks, degenerative spondylolisthesis, or spinal stenosis do not need surgery," he concluded, "but the appropriate surgical procedures may provide valuable pain relief."

"In such situations, decisions should be made jointly by well-informed patients and their physicians," he added.

Dr. Weinstein reported receiving consulting fees from United Healthcare and an honorarium as editor-in-chief of Spine. Other researchers reported financial ties to the Pacific Business Group on Health, Blue Cross/Blue Shield of Michigan, Medtronic Sofamor Danek, HealthPoint Capital Partners, K2M, Mazor Surgical Technologies, Spinal Kinetics, DePuy Spine, Zimmer, Synthes, Biotek, Custom Spine, Stryker, Fastenetix, and AOSpine. Dr. Peul and colleagues reported no potential conflicts of interest. Dr. Deyo provided no information on conflicts of interest.