Provocative discography in volunteer subjects with mild persistent low back pain

Abstract

Background context: Whether discographic injections would be positive in subjects with benign persistent “backache” who are not seeking treatment is unknown. This information is important, because benign backache undoubtedly co-exists in patients with chronic low back pain (CLBP) illness that is not discogenicin origin. If these subjects had a high rate of positive discography, the high background incidence of common backache would allow many positive tests in patients in whom discogenic processes were unrelated to their severe CLBP illness. Conversely, if subjects with benign low back pain rarely if ever had significant concordant pain reproduction on disc injections, the basic tenet of discographic diagnosis would be strengthened.

Purpose: To compare, using a strict experimental design, the relative pain and concordancy response to provocative discography in subjects with clinically insignificant “backache” and clinical subjects with CLBP illness considering surgical treatment. Study design: Comparison of experimental disc injections in subjects with persistent mild backache and those with chronic low back pain (CLBP) illness. Patient sample: Twenty-five subjects with mild persistent low back pain (LBP) were recruited for an experimental discography study. Subjects were recruited from a clinical study of patients having had cervical spine surgery. Inclusion criteria required that subjects not be receiving or seeking medical treatment for LBP, be taking no medications for backache, have no activity restrictions because of LBP, and have normal psychometric scores. To more closely approximate the pain behavior in CLBP illness, 50% (12) of the “backache” group were recruited with a chronic painful condition (neck/shoulder) unrelated to the low back. CLBP subjects, patients coming to discography for consideration of surgical treatment, were used as control subjects. Outcome measures: Results of discography were determined using the criteria of Walsh et al.: pain response of 3 or greater, two or more pain behaviors, a negative “control” discographic injection, and a similar or exact concordancy rating. Methods: Discography was performed on experimental subjects and control patients. Experienced raters, who were blinded to control versus experimental status of the subjects, scored the magnetic resonance image, discogram, psychometric tests and discography videotapes of the subjects' pain behavior. Results: Thirteen of 25 volunteer subjects had pain rated as “bad” or worse with disc injection. There were 12 painful and fully concordant disc injections in 9 of these 25 “backache” subjects (36%). These injections met all the Walsh et al. criteria for a positive diagnosis of discogenic pain. All positive discs had annular disruption to or through the outer annulus. Of the 9 subjects with positive discograms, 3 had no chronic pain states and 6 did. All subjects with positive injections had negative control discs. In comparison, in 52 subjects with CLBP illness 38 (73%) had at least one positive disc injection.

Conclusions: In a group of volunteer subjects with persistent “backache,” 36% were found to have significant pain on disc injection, which is reported to be concordant with their usual pain. The presence of positive concordant pain responses and negative control discs in 33% of subjects without CLBP illness seriously challenges the specificity of provocative discography in identifying a clinically relevant spinal pathology.

Introduction

Chronic backache of a mild degree, unaccompanied by significant illness behavior, is a very common human condition 1, 2, 3, 4, 5, 6. Although these minimal symptoms are usually not brought to the physician's attention, this sort of benign backache is reported on close questioning by 25% to 50% of adult populations studied [7]. Most people with common “backache” have little or no restriction of activity and participate in life without significant disability. However, it is common in medical practice to evaluate patients with complaints of persistent severe low back pain and disability. These patients represent a much smaller portion of the population, estimated at 2% to 3% 5, 7. Some of these patients are found on close analysis to have gross instability, deformity, tumor, or infection, but usually such conditions are not found. This latter group of patients may be labeled as patients with chronic low back pain (CLBP) illness. Researchers have long sought a lesion unique in either character or degree to this group of patients. If a specific identifiable lesion were found, one would be far more certain that in treating this lesion, specific effective remedies might be developed.

Radiologic investigations, including plain X-ray films, computed tomography scans, and magnetic resonance imaging (MRI) studies, have been unable to reliably distinguish patients with CLBP illness from asymptomatic or minimally symptomatic subjects 1, 8, 9. Both commonly have age-related changes of the lumbar spine. The provocative test of intervertebral disc injection, discography, has been purported to be the standard whereby clinically significant primary discogenic pain could be identified in patients with CLBP illness as distinct from common but relatively benign degenerative changes 10, 11, 12, 13, 14. A primary lesion in this instance means a pathologic process of the disc, which alone would be expected to cause chronic disabling back pain and the accompanying illness independent of other structural findings or psychosocial co-morbidities. This specific disc lesion is most commonly postulated to be an inflamed annular fissure that chemically or mechanically irritates local neurologic structures 15, 16. It is assumed that this lesion, indirectly detected by discography, is severely painful, functionally limiting and when present will, in and of itself, cause serious disabling illness 10, 11, 12, 15, 16, 17.

For a discographic examination to be considered positive, three criteria must be met: 1) the injection must on radiographs demonstrate an annular fissure in the disc in question; 2) the injection must evoke “concordant” pain similar to the patient's usual pain; 3) the reported pain must be of significant intensity 10, 13, 18, 19. Once established, the diagnosis is used to direct treatment interventions aimed at alleviating the patient's pain, including fusion or such “minimally invasive” options as chemonucleolysis, microdiscectomy and thermal nucleoplasty. Because the clinical results of such procedures in this patient population have been notoriously unpredictable 20, 21, 22, researchers have questioned both the methods of diagnosis (discography) and treatment 23, 24, 25.

Does a chronic positive discogram in CLBP illness necessarily establish that “painful” disc as a significant factor in the patient's illness? It is possible that other factors less easily isolated, such as occult pathology or psychosocial morbidities, play a critical role in the pain state of the patient (Fig. 1). If this were commonly the case, treatment specifically directed at disc removal and fusion would have decidedly mixed results, and very effective pain relief would be an unpredictable outcome. In fact, this mixed result has been the experience with spinal fusion for “discogenic pain.” In most series, complete pain relief without ongoing analgesic usage, resumption of normal occupational and social duties is clearly the exceptional outcome 21, 22, 26, 27, 28. Some authors believe that the illness itself in CLBP syndromes is more commonly related to primarily nonspinal processes: poor coping mechanisms, social factors promoting illness behavior, generalized or reactive pain intolerance, physical deconditioning and so forth 1, 3, 4, 6, 29, 30. Absent these co-morbidities, they believe few individuals with common age-related structural changes of the spine would have more than common backache and minimal disability. Other possibilities that may explain the relatively mixed outcomes of treatment for discogenic pain include other painful anatomic structures in the lower back, the significance of which is missed in an analysis focused narrowly on the results of discography.

In this milieu, discography has remained controversial. The discographic diagnosis assumes that disc injections would not be positive in subjects without significant LBP or with LBP primarily resulting from other factors. Previous work has explored the response to discographic injection in asymptomatic subjects with no history of back pain 13, 31, 32, 33, 34. In asymptomatic subjects with demonstrable annular disruption but no psychological distress or chronic pain behavior, the rate of painful injections is significant but low. With increasing psychological distress, compensation issues and chronic pain behavior, the rates of positive injections in experimental asymptomatic subjects become proportionately higher (40% to 80%). Critics of these types of studies have pointed out that, although painful, these injections are not in fact bona fide “false-positive” studies. If asymptomatic subjects can make no determination of pain concordancy during injection and without a concordant descriptor, the test is negative. This is a valid theoretical criticism. It is possible that patients in clinical practice report only significant concordant pain when the source of their clinically severe pain is stimulated. Case reports [35], a small experimental trial with iliac crest pain [32] and basic neurophysiologic studies 36, 37, 38, 39 have been published on this topic. These suggest that such clear discrimination between pain from different deep structures about the spine is unlikely.

In designing a more definitive study of discography, the authors chose an alternative model in which it would be possible for the concordancy criterion to be met: benign persistent backache. If on experimental testing, these subjects did not report significant and concordant pain on disc injections, the basic tenet of discographic diagnosis would be enormously strengthened. This would give persuasive, indirect evidence of a primary incapacitating disc pathology, unique to the condition, but not to date identified by imaging techniques. Conversely, if subjects with benign common backache had discographic results similar to those of patients with crippling CLBP illness, the meaning of a positive discogram would be fundamentally altered.

However, is it important that discography be able to reliably differentiate patients with discogenic CLBP illness from subjects with benign backache? Three lines of reasoning led the authors to believe that for the test to have validity, discography should be able to make such a discrimination. First, degenerative changes with only minor clinical relevance are common and thus must commonly co-exist with other nondiscogenic causes of low back pain (Fig. 1). If these “backache-only” discs were positive on injection when other axial pathology was the true pathologic feature, the diagnosis of “discogenic pain” would seriously misrepresent the patient's true illness (Fig. 1, CLBP 1). Second, it is very likely that a unique disc or annular lesion does exist that is extremely painful and generally disabling. This lesion may also co-exist with the “backache-only” discs at different levels. If these “backache-only” discs were also found to be positive on injection, a diagnosis of “multilevel disc disease” would be made when in fact the disc “disease” is really at only one level. Finally, if the patient's illness is related to primarily social or emotional issues, or a generalized pain intolerance, which magnify the distress associated with what would otherwise be a “backache-only” disc, a positive injection would again mislead the clinician. Instead of identifying the clinically relevant disease to be treated, discography would serve only to divert attention from more important co-morbid features. Clearly, if the test is positive in more than a very small number of subjects with benign backache, the ability of discography to be an effective diagnostic aid would be compromised.

The purpose of our current study was to test the hypothesis that subjects with benign backache rarely if ever would have significant concordant pain with disc injection. Would discography reliably discriminate between patients with CLBP illness and an experimental control group with common backache, that is, persons with no clinical impairment, no perceived disability and no intention to seek treatment? The subject entry criteria was designed to mitigate confounding variables, such as psychological and social stressors, as well as isolate, if possible, the variable of increased generalized pain sensitivity in chronic pain states. To this end, only control subjects with normal psychometric profiles and, as previously shown, a very low risk of overreporting pain 32, 33, 34 were recruited.