Differential Discography

 

Differential Discography

Background

A common reason for the failure of conventional spinal surgery is operation at the wrong spinal level. Differential discography is a technique to identify the predominant spinal level responsible for disc-related symptoms. This study examines the validity of differential discography as a diagnostic tool.

Methods

Between 1 January 1996 and 31 June 1998, prospective data were collected on 131 consecutive patients undergoing lumbar differential discography. The diagnostic validity of differential discography was assessed by evaluating the clinical outcome of endoscopic laser foraminoplasty at 3 months in patients treated according to conventional indications and those treated according to the results of differential discography.

Results

Overall, 89 of 131 patients yielded a definitive response to differential discography and 38 proceeded to endoscopic laser foraminoplasty. Of these, 31 had the procedure at the level indicated by differential discography and seven had the procedure at the level indicated by conventional methods. In the differential discography group, 74% obtained ‘Excellent’, ‘Good’ or ‘Satisfactory’ outcomes. In the group where differential discography was overridden in favor of clinical and MRI findings, only 43% obtained ‘Excellent’, ‘Good’ or ‘Satisfactory’ outcomes. Based on these results, differential discography yielded a sensitivity of 88.5%, specificity of 37.6%, a positive predictive value of 82.1% and a negative predictive value of 50%.

Conclusions

Differential discography is more sensitive than conventional techniques at diagnosing the segmental level responsible for the predominant presenting symptoms of disc disease. A large, randomized, controlled clinical trial is warranted to investigate the potential of this diagnostic tool further.

Introduction

Low back pain is the most common and expensive cause of chronic disability in adults less than 45 years of age and one of the most common reasons for early retirement in industrialized societies. The socio-economic costs account for nearly 1% of the entire gross domestic product of the UK each year around £6,000M.1

Conventional diagnosis of back pain relies upon multidisciplinary specialist evaluation, clinical examination, electromyography and spinal imaging. However, the relationship between spinal pathology and pain is uncertain. Studies have shown a poor correlation between back pain and disc degeneration on imaging, suggesting that only certain elements of degenerated discs are likely to be painful and that extra-discal sites may also be sources of pain.2, 3 Other degenerative changes may simply be signs of aging.4

Surgery directed towards specific discal pathology has demonstrated variable results and achieves acceptable outcomes only in selected patients.5-12 A common reason for failure of spinal surgery is operation at the wrong spinal level,13 due to multilevel disc disease and overlap of symptoms from adjacent levels. It is therefore imperative to identify the symptomatic spinal level accurately before surgical intervention.

Pain provocation studies have confirmed that chronic back pain can be reproduced by mechanical stimulation of degenerated intervertebral discs,14-16 leading to the development of aware-state evaluation that identifies pain sources by direct feedback from the patient during intervention.17 However, where stimulation of two or more discs produces concordant symptoms, selecting the appropriate level for surgery still relies heavily on imaging findings.

Differential discography is an investigation designed to distinguish the causal level of back, buttock or leg pain in patients with multilevel chronic lumbar spondylosis, failed fusion surgery and multilevel degenerative lumbar disc disease. It involves injection of up to three adjacent intervertebral discs with different pharmacologic agents (hygroscopic radio-opaque contrast medium, corticosteroid or anesthetic), each of which modifies symptoms via a different mechanism and provides analgesia over a different time scale, from a few hours to several weeks (Figure 1). By monitoring the course of the patient’s symptoms over the subsequent weeks, it is possible to assess which agent has provided the most clinical benefit and therefore which spinal level is responsible for the predominant symptoms.

Figure 1 Temporal symptom modification by differential discography

This prospective study assesses the validity of differential discography as a diagnostic tool in patients suffering from multilevel chronic lumbar spondylosis, spondylolytic spondylolisthesis, degenerative disc disease with or without failed back surgery syndrome, presenting as back pain, compressive radiculopathy or non-compressive radiculopathy.

Methods

Between 1 January 1996 and 31 June 1998, prospective data were collected on all patients undergoing lumbar differential discography at a single center in the UK.

Patients presenting with ≥2-level disc disease demonstrated by MRI scanning and at least a 1-year history of back, buttock or leg pain resistant to ≥3 months of muscle balance physiotherapy were eligible to participate in the study. Patients were excluded if they had evidence of:
• Facet joint cysts
• Cauda equina syndrome
• Painless motor deficits
• Tumors.

Participants were consented for a staged procedure consisting initially of spinal probing and discography on two spinal segments that clinically reproduced the site of back pain or peripheral radiation or were shown to evidence clinically relevant pathology on radiological or imaging investigations. The details of these techniques have been described elsewhere17 but in brief, involve radiologically guided probing of the intervertebral discs of a sedated but conscious patient in an attempt to reproduce the patient’s symptoms and identify their source. The distribution and intensity of evoked responses are recorded on a data sheet describing the patient’s response to both discal and extra-discal probing. Contrast medium is also injected to evaluate the integrity of the discs.

If spinal probing and discography were able to localize symptoms to a single spinal level, patients underwent endoscopic laser foraminoplasty17 at the appropriate level and were excluded from the differential discography cohort.

In patients where spinal probing and discography demonstrated concordant or overlapping symptoms arising at adjacent spinal levels, two-level differential discography was performed under antibiotic cover:

  • The disc with the greater clinical likelihood of causing symptoms (based on examination, spinal probing and discography findings) was instilled with methylprednisolone 80mg (Omnipaque™; Pharmacia Ltd, Milton Keynes, Bedfordshire, UK). This was termed the ‘therapeutic discogram’.
  • The less likely disc was instilled with additional contrast medium (iohexol [Omnipaque™]; Nycamed Imaging, Oslo, Sweden) until the capacity of the disc space was filled with iohexol alone. This was termed the ‘hydraulic discogram’.

Short-term modification of symptoms was assessed using the Visual Analog (Pain) Scale (VAP) and Oswestry Disability Index (ODI). Scores were recorded before and after surgery. The outcome was measured by observing the percentage change in scores. An index of ³90% was deemed an ‘Excellent’, ³50% deemed ‘Good’, ³20% deemed ‘Improved or satisfactory’ and the remainder deemed ‘Poor’.

Longer-term modification of symptoms was assessed using a pain diary relating back, buttock and leg pain levels to activity levels. Patients completed the diary 3 times a day for 6 weeks.

Patients were discharged the day of, or morning following, differential discography. A muscle balance physiotherapy regime was re-commenced on the first day following surgery, amplified with neural mobilization drills and continued on a monitored self-help basis for 3 months. Patients were reviewed at 6 weeks with their pain diaries, follow-up questionnaires and pain manikins.

At the 6-week review, symptom modification post-differential discography was assessed. Short-term improvement in symptoms (lasting 12–18 hours) was attributed to the hydraulic discogram and longer-term benefits (lasting a week or more) were attributed to the therapeutic discogram. Patients requiring further intervention were scheduled for endoscopic laser foraminoplasty. Administration staff unfamiliar with the patients or their condition allocated patients to surgical lists. Depending on the list to which they were allocated, patients underwent endoscopic laser foraminoplasty either:

  • At the spinal level indicated by differential discography OR
  • At the spinal level indicated by clinical and MRI findings regardless of the results of differential discography.

Clinical outcomes of endoscopic laser foraminoplasty was evaluated 3 months post-operatively using the VAP and ODI scores described previously. This evaluation point was considered more likely to reflect the appropriateness of the targeting rather than later evaluation that would reflect the ability of the technique to sustain outcome. The outcomes for patients undergoing endoscopic laser foraminoplasty according to the results of differential discography were compared to those for patients undergoing endoscopic laser foraminoplasty according to conventional indications.

A postal review was performed 2 years following surgery to ascertain whether the patient had required additional surgical intervention.

Results

Baseline characteristics

A total of 131 patients participated in the study. A summary of their baseline demographics is shown in Table 1. Diagnoses included multilevel chronic lumbar spondylosis, spondylolytic spondylolisthesis and degenerative disc disease with or without failed back surgery syndrome, presenting as back pain, compressive radiculopathy or non-compressive radiculopathy.

Total number of patients
131
Age (years)
          Mean±SD
          Range
 
39±10.6
22–78
Male
57 (43.5%)
Duration of symptoms (years)
          Mean±SD
          Range
 
5.6±2.4
4–17
Nature of symptoms
          Ipsilateral
          Bilateral
 
31
100

 

Table 1 Summary of baseline patient demographics

Following differential discography, 89 patients had a marked improvement in symptoms, 28 had no modification of symptoms and in 14 cases, the outcome is unknown because patients were not referred back to the investigating unit (Table 2). Interestingly, 51 of the positive responders required no further treatment because the benefit afforded by differential discography allowed muscle balance physiotherapy to achieve sufficient sustained benefit such that no further surgical intervention was required

Improvement in symptoms
Number of patients
Marked improvement
No further intervention
Further intervention required
Predominant benefit at level of therapeutic discogram
Predominant benefit at level of hydraulic discogram
89 (68%)
51 (57%)
38 (43%)
29 (76%)
 
9 (24%)
No improvement
No further intervention
Further intervention required
28 (21%)
13 (46%)
15 (54%)
Unknown
14 (11%)

Table 2 Modification of symptoms following differential discography9 (24%)

In the 38 positive responders requiring further treatment, differential discography results were used to determine the spinal level responsible for the predominant symptoms. In 29 responders, benefits were sustained for ≥1 week and were attributed to the therapeutic discogram. In nine responders, benefits were sustained for <1 week and were attributed to the hydraulic discogram.

Fifteen of the 28 patients whose symptoms were not modified by differential discography required surgical intervention at a level determined by clinical and MRI findings.

Endoscopic laser foraminoplasty results

A total of 38 positive responders to differential discography and 15 non-responders proceeded to endoscopic laser foraminoplasty. Of the 38 positive responders, X were allocated by an unbiased administrator to surgical lists where the procedure was required to be carried out at the spinal level indicated by differential discography. The remaining Y were allocated to surgical lists where the procedure was required to be carried out at the spinal level indicated by clinical and MRI findings, regardless of the results of differential discography. In Z cases, clinical, MRI and differential discography results were in agreement. However, in seven cases, clinical and MRI findings conflicted with those of differential discography and the results of differential discography were overridden. This method allowed the opportunity to compare the outcome of endoscopic laser foraminoplasty in patients treated according to differential discography with those treated according to conventional guidelines.

The 15 non-responders who required endoscopic laser foraminoplasty were treated at the spinal level indicated by clinical and MRI findings. Again, this allowed outcomes with and without differential discography guidance to be examined.

Histogram to be shown

The clinical outcome of endoscopic laser foraminoplasty was assessed 3 months post-operatively. The results are shown in Figure 2.

Figure 2 Outcome of endoscopic laser foraminoplasty in patients treated according to clinical and MRI findings or differential discography results

When the results of differential discography were followed, 23 of 31 patients (74%) obtained ‘Excellent’, ‘Good’ or ‘Satisfactory’ outcomes and 17 of 31 (55%) attained ‘Excellent’ or ‘Good’ outcomes. When the results of differential discography were overridden in favor of clinical and MRI findings, three of seven patients (43%) obtained ‘Excellent’, ‘Good’ or ‘Satisfactory’ outcomes and two of seven (29%) attained ‘Excellent’ or ‘Good’ outcomes. When differential discography offered no clear guidelines, eight of 15 patients (53%) obtained ‘Excellent’, ‘Good’ or ‘Satisfactory’ outcomes and three of 15 (20%) attained ‘Excellent’ or ‘Good’ outcomes.

Sensitivity and specificity of differential discography

The sensitivity and specificity of differential discography as a tool for diagnosing the spinal level responsible for disc-related symptoms was calculated using the outcome of endoscopic laser foraminoplasty in patients treated according to the results of differential discography and those where differential discography results were overridden. For the purposes of this calculation, endoscopic laser foraminoplasty outcomes of ‘Excellent’, ‘Good’ and ‘Satisfactory’ were taken to indicate that the procedure had been performed at the appropriate level. Unknown outcomes were not included in the calculation.

In this study, differential discography yielded a sensitivity of 88.5%, specificity 37.6%, positive predictive value of 82.1% and negative predictive value of 50% (Table 3).

 
Excellent, Good or Satisfactory outcome
Poor outcome
 
Surgery as indicated by differential discography
23
(true positives)
5
(false positives)
Positive predictive value
82%
Differential discography overridden
3
(false negatives)
3
(true negatives)
Negative predictive value
50%
 
Sensitivity 88%
Specificity 38%
 

Table 3 Sensitivity and specificity according to ‘Excellent’, ‘Good’ and ‘Satisfactory’ outcomes

Discussion

With an aging population, the presentation of multilevel painful segments is becoming more prolific and distinction of the priority index level for targeted treatment, more important as minimally invasive spinal surgery becomes more widely utilized. The need to find a reliable means of distinguishing the target level is pressing because endoscopic laser foraminoplasty outcomes indicate that fewer than 2.5% of patients with multilevel degenerative disc disease need to have intervention at more than one intervertebral level.17

The results of this study indicate that differential discography is a valuable tool with a sensitivity of 88.5% for ‘Excellent’, ‘Good’ and ‘Satisfactory’ outcomes, a specificity of 37.6%, a positive predictive value of 82.1% and a negative predictive value of 50%. This is considerably more sensitive than MRI scans in the location of painful segments but has a lower specificity as a means of determining the causal lumbar segment (Table 4). However, in contrast to other studies in Table 4, this study was based upon the findings of two spinal levels that had been shown to produce concordant contributory symptoms. The specificity, or the proportion of patients without the target disease with a negative result, is therefore less relevant than sensitivity. Differential discography is employed only when two disc levels are symptomatically contributory.

Based on ‘Excellent’ or ‘Good’ outcomes only, differential discography still yields a sensitivity of 89.5%, specificity 26.7%, positive predictive value of 60.7% and negative predictive value of 66.7%.

Investigation
Sensitivity
Specificity
Positive predicted value
Negative predictive value
Differential discography
88%
38%
82%
50%
Modic changes18
23%
97%
91%
46%
High intensity zones

19

27%
95%
89%
47%
High intensity zones

Table 4 Sensitivity and specificity of diagnostic tools to determine the location of painful spinal segments
Investigation Sensitivity Specificity Positive predicted value Negative predictive value

Differential discography has the additional merit that it uses symptom modification as the benchmark for further intervention. It therefore reassures both patient and surgeon that there is a spinal locus of the pain and that it can indeed be modified. This can be very reassuring for patients and is a valuable means of confirming the spine as the causal centre for the surgeon treating a patient with atypical or elusive symptoms.

Differential discography hinges upon the sequestration and location of corticosteroid within the disc or hydraulic elevation of the disc by contrast medium, with monitoring of the subsequent response pattern. The severity of irritation may limit the response to differential discography. Annular leakage may render the hydraulic elevation non-existent, or certainly less sustained. Similarly, annular leakage may lead to dissipation of the corticosteroid, resulting in its dissemination over a wider area in the epidural canal with a shortening and a reduction of the beneficial response. Additionally, the presence of intradiscal fibrosis may be so marked that hydraulic elevation may not be possible or very limited in extent, thus negating the implementation of the hydraulic discogram.

Many patients with disc disease have considerable symptoms arising from multiple levels. Differential discography may fail because the severity of irritation from the adjacent level may mask the benefit arising from discrete modification of the symptoms from a particular level.

Animal studies have demonstrated that instillation of crystalline steroids in to rabbits discs can accelerate annular degeneration.21 However, settlement on repeat weight-bearing X-rays was noted to have progressed at follow-up in only three patients who underwent differential discography (at 17, 25 and 29 months post-procedure). This should be considered in the context of the 60 patients in this study who underwent differential discography and required no further intervention over the ensuing 3 years. Other procedures that involve instillation of corticosteroid into the disc space also report a low risk of aggravated settlement (no cases in 958 procedures), aseptic discitis (0.9%) and deep wound infection (0.1%).22.

Conclusion

In conclusion, differential discography is more sensitive than conventional techniques at diagnosing the segmental level responsible for the predominant presenting symptoms of disc disease. Subsequent endoscopic laser foraminoplasty can be focused upon a single level with a sensitivity of 88.5% and thus multilevel intervention can be reduced. Clinical outcomes following the application of this system are encouraging and worthy of further study in a randomized, controlled clinical trial.