Article 3

From The New England Journal of Medicine
December 5, 1996 -- Volume 335, Number 23

Treatment of Spinal Pain Syndromes
by Richard B. North, M.D.

Controlled studies of surgical and other invasive procedures to treat pain are few and far between. By their very nature, such procedures are not amenable to the study designs that have become standard in pharmacologic trials.

Many of the procedures developed for pain management, therefore, have not been rigorously evaluated. Their study is confounded further both by the nature of pain, which is intrinsically subjective and difficult to quantify, and by the multifactorial nature of pain syndromes, which often involve psychological factors and coexisting social problems.

Neurosurgical and interventional procedures to treat pain may be grouped in three categories: anatomical, ablative, and augmentative. Anatomical procedures are intended to address the structural cause of pain and are exemplified by spine surgery -- for example, spinal fusion. Ablative procedures destroy some portion of the nervous system in an attempt to block or interfere with the neural transmission of pain. Augmentative procedures use electrical stimulation or the local application of drugs to achieve the same effect in a reversible manner.

The study by Lord et al. in this issue of the Journal makes an important contribution by demonstrating the benefit of a peripheral ablative procedure, percutaneous radio-frequency neurotomy, in treating cervical zygapophyseal (facet) joints, as compared with a placebo procedure. Earlier studies of ablative procedures have not been placebo-controlled. In most reported case series of patients undergoing such procedures to treat pain not caused by cancer, who were followed for extended periods by disinterested observers, the pain eventually recurred.

This new study corroborates the results of these reports, improving on most of them by providing Kaplan-Meier survival estimates; the median time to a recurrence was 263 days. Fortunately, this minimally invasive procedure can be repeated. It will be interesting to determine the long-term outcomes in patients with recurrent pain who undergo repeated neurotomy of cervical zygapophyseal joints. Such studies will address not only the hypothesis that the regeneration of nerves is responsible for the recurrent pain, but also the role of the procedure in clinical practice.

Lord et al. used temporary relief of pain by reversible blockade with local anesthesia as a diagnostic test to identify candidates for neurotomy. The specificity of the test, however, has not been validated in long-term studies of ablative or anatomical procedures. This partly reflects the limited long-term efficacy of "permanent" ablative procedures, which Lord et al. show to be important even for a procedure with proven efficacy. Still open to question, however, is the specificity of local anesthetic blocks and even of supposedly permanent procedures in defining the anatomical and physiologic causes of persistent pain.

This study used double-anesthetic blockade -- with anesthetic drugs that have different durations of action -- to identify physiologic (as opposed to functional or placebo) responses and putative sources of pain. The attainment of long-lasting relief with a longer-acting drug can be interpreted as a specific response, indicating that the source of the pain has been identified. Lord et al. built on earlier reports indicating that patients with cervical pain attributable to facet-joint disease can be identified by controlled double-anesthetic blockade; they used this method to select patients to undergo nerve ablation. The much longer lasting effect of nerve ablation, as compared with the placebo procedure, is taken as additional evidence that the cervical facet joints are clinically important sources of pain.

Temporary relief of pain achieved with local-anesthetic nerve blockade, however, has limited specificity in localizing sources of pain. My colleagues and I recently reported the results of a controlled study of the specificity of diagnostic nerve blocks for sciatica of spinal origin. Not only did blockade of the paraspinal lumbosacral nerve root and the medial branch of the posterior primary ramus (at the site of disease or proximal to it) bring temporary relief to a majority of patients, but so also did blockade of the sciatic nerve (distal or collateral to the disease site). Subcutaneous injections of the same dose of the local-anesthetic drug were ineffective.

These findings confirmed that false positive results are common with nerve blocks used to localize or diagnose sources of pain and that the sensitivity and specificity of such blocks are low. For sciatic-nerve blocks, the specificity ranged from 24 to 36 percent. Patterns of responses specific to the patients' established diagnosis of radiculopathy (i.e., the root block was most effective) had sensitivities ranging from 9 to 42 percent.

Temporary blockade with local anesthesia can be effective when it is directed to areas of referred pain, as opposed to sources of pain. Even central pain resulting from central nervous system injury or disease can be relieved by distal local-anesthetic blocks in the area or segment where the pain is perceived or projected. Can the same be true of double blocks? Does a long-acting local-anesthetic blockade of a collateral or distal nerve, supplying an area of referred pain, provide relief that lasts longer but is nonspecific? Can the same be true of ablative procedures?

Recent anatomical and physiologic studies may explain the nonspecific relief of pain on the basis of the interruption of peripheral input. Primary afferent neurons from a number of spinal and paraspinal structures converge on second-order neurons in the dorsal horn of the spinal cord. Recordings from these centrally projecting afferent neurons have shown the convergence of nociceptive as well as nonpainful sensory inputs, with excitatory effects on the same central pain-signaling neurons.

The use of placebo-controlled trials in the treatment of pain raises interesting practical questions. Is it necessary to perform multiple separate anesthetic-block procedures to select patients for a neurotomy procedure that has low morbidity? We have had similar rates of success with radio-frequency neurotomy of lumbar facet joints, without major complications, in patients selected on the basis of a clinical diagnosis corroborated by a single block. Are placebo studies cost effective? No one knows. The difference between single and multiple, placebo-controlled anesthetic blocks in selecting patients to undergo neurotomy requires further study.

The use of local-anesthetic blockade and radio-frequency neurotomy to treat pain has no normal physiologic counterpart, and there is no a priori reason to expect the central nervous system to interpret their effects specifically and meaningfully. Diagnostic and mechanistic interpretations of nerve blocks, neurotomy, and other interventions should accordingly be undertaken with caution.

Studies such as the one by Lord et al. provide convincing evidence that pain apparently originating in the facet joints can be treated empirically. Nevertheless, the underlying mechanisms of spinal pain remain to be defined fully, as does the clinical role of ablative procedures.

Richard B. North, M.D.
Johns Hopkins University School of Medicine
Baltimore, MD 21287-7713

Copyright © 1996 by the Massachusetts Medical Society