3.6 Surgical management of chronic low back pain

Patients with chronic low back pain are frequently referred to orthopaedic surgeons. The majority of these patients do not require surgical treatment, and should be managed by non-operative means. A small proportion may benefit from surgical stabilization of the spine or disc replacement. This chapter is concerned with the rationale for this approach, investigation, surgical methods, and the results.

The surgeon and patient have to take a decision after weighing up the risks and benefits of the procedure (often hard to define) against the disability. Spinal fusion may be regarded as either an attempt to speed up the natural progress of degenerative changes towards a functional ankylosis or a prevention of progressive deformity or instability. The aim is to reduce the severity of back pain and improve functional ability.

The evidence on which to base these decisions is developing. Several randomized trials of spinal fusion have now been completed:

In Sweden, trials showed an outcome advantage of fusion over conventional physiotherapy, but could show no difference between three different surgical techniques. In Norway and in the United Kingdom, studies could show no advantage of surgery over intensive rehabilitation. These trials have caused controversy, especially in the United States, where there are high rates of surgery for back pain.

‘Soft’ fusion using a variety of posterior devices has been practised for 20 years, but no large trial completed. The rationale remains experimental, not least because of our poor understanding of intersegmental biomechanics in relation to pain. Many ingenious devices have been developed to alter these, but results are anecdotal. In the Spine Stabilisation Trial in the United Kingdom, there was a small subset of patients who had the Graf procedure. This was underpowered, and there was no obvious advantage seen in these patients of surgery over rehabilitation and they required more revision surgery. Surgery was quicker with fewer complications than spinal fusion.

Disc replacement has evolved in the last 30 years, with cohort studies only reported until two Food and Drug Administration-approved trials. The Charité study was against a BAK (Bagby and Kuslich) anterior cage procedure (now abandoned) that showed non-inferiority. The ProDisc study showed non-inferiority using an unvalidated version of the Oswestry Disability Index (http://www.mapi-institute.com/; email: [email protected]) over 360-degree fusion. Both these trials show a non-statistically significant advantage to disc replacement and it is likely that they were underpowered. No study against rehabilitation has been reported yet. Surgeons should consider the difficulties of selecting patients and the high risks of revision surgery before recommending this procedure.

It is assumed that back pain may be generated from the low back by ‘mechanical’ means or through a source of ‘inflammatory’ agents (usually assumed to be the intervertebral disc) irritating neural tissues.

Unfortunately, reality has not always followed expectation, and the results of treatment have varied considerably. Carragee used discography to identify ‘best bet’ patients for spinal fusion. Only 27% met his strict criteria of success compared with 72% of a control-spondylolisthesis group (Box 3.6.1).

Spinal stability is a much abused term. There are at least four ways in which this can be conceived:

White and Punjabi defined stability as ‘a condition of the spine under normal physiological loading where there is neither abnormal strain nor excessive or abnormal motion in the functional spinal unit’. The functional spinal unit is a motion segment, consisting of bone, disc, and bone, as well as its supporting joints, ligaments, and muscles. Instability is the loss of the ability of the spine under physiologic loads to maintain relationships between vertebrae in such a way that there is neither damage or subsequent irritation to the spinal cord or nerve roots and, in addition, there is no development of incapacitating deformity or pain from structural changes.

Perceived or functional instability is sometimes called ‘instability syndrome’. The patient complains of giving way, getting stuck, a ratchety flexion in the spine, and, occasionally, a sensation of disconnection of the top of the body from the bottom. Panjabi has recently published a hypothesis based on the concept of a single trauma or cumulative microtrauma causing subfailure injuries of the ligaments and embedded mechanoreceptors leading to disturbed muscle control. Harris has suggested that mismatch between expected and actual proprioceptive information may be a potent source of ‘distress’ in the central nervous system, experienced as pain. There is a strong body of evidence suggesting proprioceptive dysfunction in chronic back pain patients.

Humans stand and walk best if they can keep their frame in balance, which means, when standing, the head is over the hips, which are in turn over the heels. In other words, the gravity line goes through the cervicothoracic junction and the lumbosacral joint. Deviations from this (usually forward) overload the posterior lumbar musculature and require compensatory hip extension and knee flexion. All may be well at skeletal maturity, but loss of balance may follow kyphosis due to disc degeneration and fractures. Imbalance has become increasingly recognized as an important cause of back pain. Surgery can be used to correct imbalance by osteotomy or facetectomy during fusion, but milder cases may well respond to rehabilitation and exercise.

There is anecdotal evidence that adults with spinal stenosis have a long history of back pain. Arguably the flexed posture these individuals adopt may be a contributory factor to sagittal imbalance. There is experimental evidence that these individuals are at more risk of trouble with their backs if they have a disc prolapse.

A good history is essential. The ‘ideal’ patient for spinal fusion, may have some or all of the following characteristics: a non-smoking ‘normal citizen’; no litigation outstanding or other potential secondary gains from his or her pain; a clear history; and a crescendo of symptoms, but with pain-free or low pain intervals. Some patients with recurrent attacks of pain report prodromal symptoms preceding an attack. Surgeons have tried many methods to identify good responders: for example, a trial period in external splintage, or a rational response to discography. External fixation has also been tried, but has not proved popular because of complications and poor predictive ability. Flexion–extension radiographs have been used for many years, but their value is very limited. Standing views may be of value if the gravity line or whole body can be included. This approach remains experimental.

Failures of the surgical approach can be attributed to two main areas: the patient and the surgical methodology.

A technically correct operation achieving a solid fusion may not relieve pain, and indeed it can make it worse. There may be a variety of reasons for this.

The objective of most procedures is to obtain a solid bony fusion, avoiding damage to the surrounding soft tissues. A wide variety of methods are available, and comparisons between them are difficult. Posterolateral fusions tend to be easier to perform, but are probably less reliable in terms of both fusion rates and in immobilizing a segment (or functional spinal unit). Interbody fusions, either from the back or from the front, are more likely to fuse with instrumentation, but are technically more difficult to perform, and carry a higher risk of complication. Spinal instrumentation has evolved rapidly, increasing in complexity and expense. It has been difficult to demonstrate that its use has any advantage. Some studies, but not all, suggest that the use of instrumentation increases the fusion rate. Unfortunately this is not necessarily accompanied by an improvement in clinical results. The evidence base would support the use of uninstrumented posterolateral fusion as the best option with lowest complication rate.

It is likely that a significant proportion of patients have discogenic or segmental pain. Proponents of this view use provocative discography to identify painful segments. Discography can identify segments where pain is reproduced by injection of saline into the disc and adjacent normal discs. It identifies patients with inappropriate responses (over-reaction to local anaesthetic or skin penetration by needle) and inappropriate or multilevel response to disc injection. The ‘normal’ disc is usually pain free when injected. There is a large literature on this, but many studies are flawed. Carragee’s studies are a good starting point. He was able to demonstrate only a predictive value for discography of 50–60% in spinal fusion patients.

The objective of treating these patients is to relieve symptoms and improve function. In theory this requires that a fusion be achieved. There is a close, but by no means total, correlation between fusion and pain relief.

There remains considerable uncertainty surrounding internal fixation. This may be unnecessary for a single-level posterior fusion, but may be useful if more than one level is involved. Certainly the experience with spinal fusion for scoliosis would support this view. Large trials are probably necessary to resolve these issues.