Surgery for Low Back Pain

Youssef Ghabrial

Nikolai Bogduk

Fusion of the lumbar spine was initially developed to treat tuberculosis of the spine and to stabilize spinal deformities.¹^,²^,³ Later, it was used to treat the instability of spondylolisthesis and to prevent iatrogenic instability after disk excision for disk herniation. In due course, fusion was applied to treat chronic low back pain.

The procedures commonly used in the treatment of back pain are summarized in Figure 74.1. The procedures differ according to the direction of approach: whether the spine is fused or not, whether disks are excised or not, and if a bone graft or instruments or both are used to achieve fusion. Other, more avant-garde techniques include minimally invasive approaches through the psoas muscle.⁴

Posterolateral fusions were originally performed using only a bone graft. They were later supplemented by pedicle screws and plates when it was shown that instrumentation improved fusion rates. Likewise, posterior interbody fusions are now supplemented by pedicle screws in order to enhance arthrodesis. Anterior interbody fusions are classically performed using a bone dowel, but some surgeons add screws and plates or an interbody cage to promote bone union.

Rationale

It is difficult to find in the literature an explicit, stated rationale for fusion in the treatment of low back pain. Implicitly, the earliest rationale was that because back pain was aggravated by movement of the lumbar spine, instability must be causing the pain, and, therefore, stabilizing the spine should relieve the pain. According to this rationale, a specific diagnosis was not required; persistent pain was the sole indication for surgery. Some surgeons applied nominal diagnostic rubrics, such as spondylosis or degenerative disk disease, despite these conditions being no more than normal age changes,⁵ a paradox that some surgeons acknowledge.⁴

Later, some surgeons proposed that the intervertebral disk was the source of pain and adopted discography as the diagnostic test. Painful disks could be protected from aggravation by posterolateral fusion, or the disk could be excised and replaced by a bone graft.

FIGURE 74.1 A classification of types of surgery used for low back pain.

A contentious indication for fusion is spondylolisthesis. This condition is an attractive target for surgery because it constitutes a deformity and has been reputed to be unstable, both of which can be rectified by surgery. However, research has shown that spondylolisthesis is commonly asymptomatic, such that its presence on radiographs has no statistically significant or clinically significant association with back pain.⁵^,⁶ Furthermore, studies have shown no detectable instability in spondylolisthesis.⁷^,⁸^,⁹^,¹⁰ Other studies have found evidence of abnormal motion and paradoxical motion in patients with spondylolisthesis, but at the segment above, not at the affected segment.¹¹

Spondylolisthesis remains an indication for surgery in patients with radicular pain or radiculopathy, when it can be shown that the deformity is causing nerve root compression. In that event, surgery is remarkably successful at relieving radicular pain.¹²^,¹³ In patients with concurrent back pain, surgery is less often effective and, to a lesser degree, for the relief of back pain.¹²^,¹³

A literature review summarized the success rates of surgery for spondylolisthesis as reported in 34 original studies.¹⁴ For posterior fusion alone, the success rates ranged between 37.5% and 98%, with an average of 74.8%. For anterior fusion, the success rates ranged between 80% and 95%. These data, however, were based on irregular and sometimes ill-defined criteria for clinical success. Some studies measured success in terms of patients achieving complete relief of symptoms, whereas others reported only the proportion of patients achieving some degree of improvement. Few studies, however, reported the degree of impairment before surgery and rarely have outcomes been corroborated by quantitative measures of pain, disability, return to work, and use of other health care.

Better data are available from randomized controlled trials. Surgery for spondylolisthesis is patently more effective than conservative therapy,¹⁵^,¹⁶^,¹⁷ but surgery is neither universally nor completely effective. Only 56% of patients rate themselves as “much better,” whereas 11% consider themselves unchanged and consider their condition to be worse.¹⁶ Surgery is, therefore, an imperfect solution for spondylolisthesis. The implication is that fusion does not effectively target the cause of back pain in spondylolisthesis.

Effectiveness

Pivotal to the use of fusion for low back pain is the evidence for its effectiveness and efficacy. This evidence should be as much of interest to physicians who might refer patients for surgery as it is to surgeons who might perform the surgery. That evidence has evolved. Three epochs can be described, according to whether studies were performed before, during, or after the advent of the principles and demands of evidence-based medicine (EBM), which occurred around the turn of the 20th century.

BEFORE EVIDENCE-BASED MEDICINE

Before the advent of EBM, surgeons reported good outcomes from fusion in the treatment of chronic low back pain. These reports created the reputation of surgery of being a decisive option for patients for whom conservative therapy had failed to provide relief.

However, studies during that epoch used methods that nowadays would not be held in high regard. These include the following:

Surgeons evaluating their own outcomes, as opposed to having an independent, third party assessing outcomes
Reporting qualitative outcomes, such as “excellent” or “very good” results, without supporting quantitative data
Not providing baseline data
Reporting outcomes without stratifying for different presenting conditions or symptoms
Not using validated instruments for outcome measures
Not reporting on all variables of interest, such as pain, disability or function, and use of other health care for back pain

For example, the success rates reported during this epoch were summarized in terms such as “61% excellent and 31% good outcomes,” “88% excellent or good outcomes,” and “100% satisfactory outcomes” for posterior interbody fusion and “89% relieved,” “95% clinically favorable,” and “74% satisfactory outcomes” for anterior lumbar interbody fusion.¹⁸ For patients with persistent, disabling pain, or their treating physicians, such numbers are appealing.

Table 74.1 lists the outcomes reported by various studies for anterior lumbar interbody fusion in patients diagnosed by discography as having discogenic pain. Although the success rates are more modest than those of the earlier literature, many are still attractive. One in five patients, or better, could expect to achieve complete relief of their back pain.

TABLE 74.1 Outcomes Reported by Various Studies that Used Discography to Select Segments for Treatment by Fusion

Source	Criteria	Success Rate
Blumenthal et al.¹⁹	Normal activities and no medications or NSAIDs only	74%
Kozak and O’Brien²⁰	At least 75% relief of pain and return to work or Slight restriction of activities and no analgesics	74%
Kostuik et al.²¹	Absence of significant back or leg pain Occasional use of nonnarcotic analgesics	31%
Gill and Blumenthal²²	At least 75% relief of pain and return to work or Return to normal activities and no opioids	66%
Lee et al.²³	No pain Mild pain No medications No opioids No restricted activities Mild or moderate restriction Return to full work Return to restricted work	26% 61% 59% 30% 50% 30% 81% 11%
Wetzel et al.²⁴	Minimal symptoms and no analgesics, or marked improvement, and rare use of analgesics	46%
Parker et al.²⁵	Pain less than 4/10 No medications other than NSAIDs Return to at least 75% previous work capacity	39%
Penta and Fraser²⁶	Complete relief or Good deal of relief	39% 39%
Greenough et al.²⁷	Complete or almost complete relief A good deal of relief	17% 23%
NSAIDs, nonsteroidal anti-inflammatory drugs.