The intervertebral disc is estimated to be the source of chronic low back pain (CLBP) in approximately 40% of adults.1,2 History, physical examination, and advanced imaging provide inadequate sensitivity and specificity to reliably diagnose discogenic pain.2–5 As such, provocation discography (PD) was introduced to provide more accurate diagnostic information to confirm or refute the disc as the true source of CLBP. However, the original and early methods of performing and interpreting PD were associated with an unacceptably high false positive rate.6–9 Further, a study of PD by Carragee et al10,11 demonstrated accelerated disc degeneration compared with the normal aging process and an increase rate of future symptomatic CLBP. These reports have been challenged on multiple grounds, including the fact that study participants did not have clinically significant CLBP and thus do not reflect a realistic clinical population with an indication for PD.12 However, the results raised enough concern to change patterns of clinical practice.
Subsequently, technical and operational criteria were developed and endorsed jointly by the Spine Intervention Society (SIS) and the International Society for the Study of Pain (IASP)13 in an effort to improve the diagnostic accuracy of the test and to mitigate the possibility of increasing the rate of disc degeneration and long-term sequelae. The SIS/IASP method includes the use of fine-gauge needles to minimize the diameter of annular puncture, limits to the maximum volume of contrast and intradiscal pressure introduced during PD, as well as rigorous operational criteria for the interpretation of a “positive” versus “negative” result. The implementation of these operational criteria has been shown to substantially decrease the false positive rate.14,15 However, no study has reported on long-term adverse event rates following lumbar PD using the SIS/IASP method. The present study aimed to assess the long-term effect of lumbar PD on the rate of subsequent disc degeneration, disc disruption, and disc herniation in a true clinical population of patients who seek treatment for refractory CLBP. This study was undertaken not only to assess long-term safety of lumbar PD, but also long-term safety of disc access for therapeutic purposes, as future biologic16–19 and/or gene therapy20,21 treatments for refractory discogenic pain are likely to require intradiscal access.
MATERIALS AND METHODS
This longitudinal matched cohort study was conducted at a tertiary academic medical center and approved by the Mayo Clinic Institutional Review Board (16–007228). The electronic medical record (EMR) at the study institution was queried to identify consecutive patients who presented for clinical care of refractory CLBP, had a baseline MRI, underwent PD within 6 months of the baseline MRI, and then underwent a second MRI a minimum of 7 years after PD (PD Group). EMR query was restricted to dates between January 1, 1998 and December 31, 2007. The EMR was also queried to identify consecutive patients who had clinical care for refractory CLBP, had a baseline MRI, did not undergo PD, and then underwent a second MRI a minimum of 7 years later (Control group). The following exclusion criteria were applied to both groups: (1)age less than 18 at the time of the first MRI, (2)age more than 60 years at the time of the second MRI, (3) prior or subsequent PD relative to the index procedure, (4) prior thoracic or lumbosacral spine surgery, (5) thoracic or lumbosacral spine surgery between the time of the first and second MRIs, (6) baseline diagnosis of fracture, dislocation, tumor, or rheumatologic condition affecting the lumbar spine, (7) development of fracture, dislocation, tumor, or diagnosis of a rheumatologic condition between the first and second MRI.
Patients in the PD Group who met the study selection criteria were matched to patients in the Control group for age, body mass index (BMI), composite Pfirrmann score (at L3–L4, L4–5, L5–S1 levels) ±2, number of disc herniations (at L3–L4, L4–5, L5–S1 levels) ±1.
Demographic, historical, and procedural data were collected for each patient by chart review (Appendix A, http://links.lww.com/BRS/B433). The following occurrences were tallied in order to inform of the possible consequences of discography and to maintain transparency about potential selection bias, but also served as exclusion criteria (as above) given their potential for causing accelerated disc degeneration: (1) microdiscectomy, level(s), and side(s) following PD; (2) other lumbar spine surgery, level(s), and side(s) following PD; (3) intradiscal therapy, any type/method, level(s), and side(s) following PD; (4) spinal infection following PD.
All discography procedures were performed by a single radiologist board certified in radiology with significant experience in the practice of interventional spine care (Timothy Maus). PD was performed in accordance with the SIS/IASP standards. A 22-gauge spinal needle was used to access the disc in all cases. Intradiscal contrast volume was limited to 3 mL and the maximum intradiscal pressurize did not exceed 50 psi above opening pressure. Other operational criteria are detailed in clinical practice guidelines.13
A study-specific imaging review and analysis tool developed with the Analyze imaging software development platform (Mayo Biomedical Imaging Resource, Rochester, MN) was used to evaluate all MRIs, scoring, and measurements (Figure 1A, B).22,23 The MRI grading protocol has demonstrated consistent and acceptable inter/intra-rater reliability and has been used in other studies which assess lumbar disc degeneration.10,24,25 Two double-board certified neuroradiologists with a spine-focused practice and more than 5 years of clinical experience interpreting MRIs of the lumbar spine, were blinded to group allocation and used a standardized data collection sheet specific to each lumbar disc level (Appendix B, http://links.lww.com/BRS/B433) to record outcomes (below). Disagreements between the two MRI examiners were settled by consensus discussion.
- 1. Pfirrmann score. The Pfirrmann score is an integer on a five-level ordinal scale developed to grade lumbar disc degeneration.26
- 2. Disc to cerebrospinal fluid (CSF) T2-signal intensity ratio. T2-signal intensity loss is a measure of loss of disc fluid (i.e., disc desiccation), a process indicative of disc degeneration. The quantification of T2-signal intensity used methods described by Videman et al27 (Figure 1).
- 3. Disc height (millimeters [mm]). Disc height was quantified by measuring the maximal distance between vertebral end-plates in the mid-sagittal plane (Figure 1).
- 4. The presence/absence of disc bulge or herniation was characterized using the nomenclature defined by Fardon et al.28
- 5. The presence/absence and side of an annular high intensity zone (HIZ), which has been shown to correlate with the presence a grade 4 annular fissure29 and discogenic pain.30
- 6. The presence/absence and type (1 or 2) of lumbar vertebral endplate (Modic) changes, which have been associated with symptomatic of disc degeneration.31
Demographic, historical, and radiographic differences were compared between the PD group and the Control group in order to identify any potential confounding factors.
MRI changes from baseline to repeat scan were compared between the discs injected during PD (PD group) and the disc(s) at analogous level(s) in patients from the Control group. The primary outcome was the change in Pfirrmann score from baseline to time of repeat MRI, as it represents an overall measure of disc degeneration and has been used in a prior similar study.13 The frequency and 95% CI of Pfirrmann scores grouped as I to II, III to IV, and V were calculated. Secondary outcomes included the change in the following measures from baseline to repeat MRI: (1) mean disc-to-CSF T2-signal intensity ratio, (2) mean disc height, (3) number of disc bulge(s) or herniated nucleus pulposis(s) and congruence with the level and side of PD in the PD group, (4) number of HIZ(s) and congruence with the level and side of discography in the PD group, and (5) number of disc levels with Modic changes (type I, II, III),31 as well as congruence with the level of PD in the PD group.
The above data were also analyzed with regard to MRI changes from baseline to repeat scan between non-injected discs in the PD group and corresponding non-injected discs in the Control group in order to assess for any possible group differences in propensity toward more or less rapid disc degeneration in the lumbar spine.
McNemar test of equality of paired proportions with an α = 0.05 two-sided significance level demonstrated 93% power to detect a difference in group proportions of Pfirrmann score category changes of 0.21 (0.140 and 0.350 based on Bogduk et al's study),13 when the sample size in each group is 50 with a 0.25 proportion of discordant pairs. To further increase power, we planned a 4:1 matching ratio of in the PD group to the Control group.
Data were checked for distributional form and outliers using summary statistics and graphical displays. Data were analyzed using SAS version 9.4 (Cary, NC). The level of significance was set at 0.05. Two-sided testing was used for all hypothesis testing. To illustrate the demographic, radiologic, and procedural characteristics of the study sample, means and standard deviations (SD) were calculated for continuous variables. Proportions and percentages are shown for categorical variables. Multivariable multi-level mixed models were used to adjust for multiple discs per patient, and multiple matched controls per discography cases, in addition to age, sex, and BMI confounders. In the cases of small sample sizes, chi-squared or Fisher exact tests were used for comparison of categorical data. Sensitivity analyses were conducted by comparing the mixed models including the total cohort of cases (n = 24) that were matched to between two and four controls, to the subset of cases (n = 21) matched to three or four controls.
A total of 585 consecutive patients underwent lumbar PD at the study institution during the period of EMR query. Of these patients, 49 had a follow-up lumbar MRI 7 to 10 years after their baseline lumbar MRI. Twenty-four of these patients were excluded, as 23 had undergone spinal fusion between the time of PD and follow-up lumbar MRI and one had severe scoliosis. During the same period, 950 consecutive patients were identified who had a baseline lumbar MRI and a follow-up lumbar MRI 7 to 10 years later, and 91 were matched to the PD cases. Seven of these patients were excluded; six were due to spinal fusion between the time of baseline and follow-up MRI and one was due to a spinal cord lipoma and tethered cord. Eighty-two of the remaining 84 control patients were a match by age (±8 yrs), BMI (±6 kg/m2), Pfirrmann score (±2), and disc herniation(s) (±1) to patients in the PD group. It was not possible to match one of the PD patients to an adequate number of control patients, and thus, this case was excluded, leaving a total of 24 patients in the PD group and 82 patients in the Control group for analysis. Given the 1:2 to 1:4 matching ratio used for the PD group discs and Control group discs, a greater power to detect intergroup differences was present with respect to the primary outcome than in prior study.13
Demographic information for all patients included in the study is shown in Table 1. Notably, all patients who underwent PD were male, and thus, subsequent data analyses of disc height and disc-to-CSF T2 signal intensity ratio are controlled for this sex difference. Lumbar PD was performed a median of 1.5 months (min: 2 d; max 11 mos) after baseline MRI. None of the patients in the PD cohort had a microdiscectomy, any subsequent intradiscal procedure, or a spinal infection following PD. The mean duration between initial and follow-up MRIs was significantly longer in the PD cohort at 9.8 years (SD 1.9) compared with 8.5 years (SD 1.2) in the Control cohort (P = 0.006).
Baseline and follow-up findings for injected discs in the PD group and corresponding discs in the Control group are shown in Table 2. Models were adjusted for age, sex, and BMI. At baseline, more discs in the Control group had a Pfirrmann grade of 1 or 2 compared with the corresponding discs that underwent puncture in the PD group, though 95% confidence intervals (CI) overlapped: 29%, 95% CI 23–35% versus 17%, 9% to 27%. At follow-up, there was no intergroup difference in the proportion of discs that progressed in Pfirrmann grade category (P = 0.3578). There were intergroup differences between Control group and corresponding PD group discs with regard to the decrease in disc-to-CSF T2 signal intensity ratio, number of associated Modic changes (both type 1 and type 2), new disc bulges, new disc herniations, and new HIZs (all P > 0.05). At follow-up, the mean decrease in disc height was 1.0 mm in punctured discs in the PD group compared with 0.6 mm in corresponding Control group discs (P = 0.0126).
Baseline and follow-up findings for non-injected discs in the PD group and corresponding discs in the Control group are shown in Table 3. Models were adjusted for age, sex, and BMI. There were no intergroup differences between non-injected discs in PD group and corresponding discs in the Control group with regard to any outcome measure (all P > 0.05).
The most important finding of this study is that PD, utilizing SIS/IASP standards, did not lead to accelerated disc degeneration, internal disc disruption, or disc herniation compared with corresponding discs in a matched control cohort at long-term follow-up. This finding is particularly notable given that patients who underwent PD in this cohort all warranted this procedure due to clinical symptoms of refractory low back pain, unlike prior study.13 Further, this finding was observed despite the fact that the mean duration between initial and follow-up MRIs was significantly longer in the PD cohort compare with the Control cohort (9.8 yrs vs. 8.5 yrs, P = 0.006); greater disc degeneration would be expected in the PD cohort due to normal aging. Additionally, 22 g needles were used for disc puncture in all patients in the PD group, yet acceleration of disc degeneration was not observed. In prior study, 86% of subjects underwent disc puncture with a 25 g needle rather than a 22 g needle. This finding suggests that use of a 22 g (vs. 25 g) needle may not affect the rate of disc degeneration or occurrence of new disc herniations. This finding is consistent with animal study which demonstrated that large bore needle punctures in relation to the size of the disc are needed to create degeneration. Elliot et al32 reviewed 23 in vivo disc puncture studies in rat, rabbit, dog, and sheep models, and found that significant disc changes were not produced when the needle gauge represented less than 40% of the disc height.
Beyond the safe practice of lumbar PD, the present findings have important implications for disc access in general. While confirmation is needed, our data suggest that percutaneous disc access itself, when applying SIS/IASP standards, may not result in long-term harm to discs. Safe percutaneous disc access is relevant to future treatments for refractory discogenic pain that may spare patients from spinal fusion surgery or chronic opioid use. Investigation of intradiscal biologic agents,16–19 and intradiscal gene therapy solutions20,21 for discogenic pain are underway and while it is not clear what treatments will emerge as truly safe and effective, percutaneous intradiscal access will likely be necessary.
The findings of this study are consistent with other prior investigation. Ohtori et al33 compared Pfirrmann scores in subjects with symptomatic low back pain who had undergone lumbar PD and intradiscal bupivacaine injection (“discoblock”) with matched controls; there was no significant intergroup difference at 3 to 5 years follow up. Statistical power was limited, however (n = 28). Flanagan and Chung34 found no evidence of disc degeneration 10 to 20 years after lumbar PD in individuals who had originally presented with low back pain, although this study was limited by use of radiography rather than MRI to assess for disc degeneration.
Our findings must be interpreted within the limitations of this study, namely, a risk of selection bias. All patients in the PD cohort were male, and as such, the mean disc height at baseline was higher in the PD cohort compared with the Control cohort (55% male). This discrepancy was controlled for in the statistical analysis, but unrecognized sex differences in the rate of disc degeneration are possible.35 Only 8% of patients who originally underwent lumbar PD (n = 585) had an MRI available at a minimum of 7 years later (n = 49). This subpopulation who presented with symptoms warranting a repeat MRI may not be representative of the complete cohort of patients that originally underwent lumbar PD. However, a greater proportion of discs in the Control group, compared with corresponding punctured discs in the PD group, advanced in Pfirrmann grade category and developed new Modic changes, new disc herniations, or new HIZs. While these intergroup differences were not statistically significant, these findings do provide reassurance that even with some degree of selection bias, an intergroup difference in the opposite direction is unlikely. Further, lack of intergroup differences between non-punctured discs in the PD group and corresponding discs in the Control group provides further reassurance that neither group was more or less prone to accelerated disc degeneration or new disc herniations. Finally, patients underwent discography by a single experienced physician, so results may not be applicable to a less experienced proceduralist.
Supplemental digital content is available for this article. Direct URL citations appearing in the printed text are provided in the HTML and PDF version of this article on the journal's Web site (www.spinejournal.com).
1. DePalma MJ, Ketchum JM, Saullo T. What is the source of chronic low back pain
and does age play a role? Pain Med
2. Schwarzer AC, Aprill CN, Derby R, et al. The prevalence and clinical features of internal disc disruption in patients with chronic low back pain
. Spine (Phila Pa 1976)
3. Vanharanta H, Sachs BL, Spivey MA, et al. The relationship of pain provocation to lumbar disc deterioration as seen by CT/discography. Spine (Phila Pa 1976)
4. Boden SD, Davis DO, Dina TS, et al. Abnormal magnetic-resonance scans of the lumbar spine in asymptomatic subjects. A prospective investigation abnormal lumbar magnetic-resonance spine scans of the in asymptomatic. J Bone Jt Surg
5. Jensen MC, Brant-Zawadzki MN, Obuchowski N, et al. Magnetic resonance imaging
of the lumbar spine in people without back pain. N Engl J Med
6. Holt EP Jr. The question of lumbar discography. J Bone Joint Surg Am
7. Carragee EJ, Tanner GM, Yang B, et al. False-positive findings on lumbar discography: reliability of subjective concordance assessment during provocative disc injection. Spine (Phila Pa 1976)
8. Carragee EJ, Tanner CM, Khurana S, et al. The rates of false-positive lumbar discography in select patients without low back symptoms. Spine (Phila Pa 1976)
9. Carragee EJ, Alamin TF, Carragee JM. Low-pressure positive discography in subjects asymptomatic of significant low back pain illness. Spine (Phila Pa 1976)
10. Carragee EJ, Don AS, Hurwitz EL, et al. 2009 ISSLS prize winner: does discography cause accelerated progression of degeneration changes in the lumbar disc: a ten-year matched cohort study. Spine (Phila Pa 1976)
11. Cuellar JM, Stauff MP, Herzog RJ, et al. Does provocative discography cause clinically important injury to the lumbar intervertebral disc? A 10-year matched cohort study. Spine J
12. Mccormick ZL, Defrancesch F, Loomba V, et al. Diagnostic value, prognostic value, and safety of provocation discography
. Pain Med
13. Bogduk N. Practice Guidelines for Spinal Diagnostic and Treatment Procedures. 2nd ed.San Franciscso, CA: International Spine Intervention Society; 2013.
14. Wolfer LR, Derby R, Lee JE, et al. Systematic review of lumbar provocation discography
in asymptomatic subjects with a meta-analysis of false-positive rates. Pain Physician
15. Derby R, Kim BJ, Lee SH, et al. Comparison of discographic findings in asymptomatic subject discs and the negative discs of chronic LBP patients: can discography distinguish asymptomatic discs among morphologically abnormal discs? Spine J
16. Pettine KA, Suzuki RK, Sand TT, et al. Autologous bone marrow concentrate intradiscal injection for the treatment of degenerative disc disease with three-year follow-up. Int Orthop
17. Kumar H, Ha DH, Lee EJ, et al. Safety and tolerability of intradiscal implantation of combined autologous adipose-derived mesenchymal stem cells and hyaluronic acid in patients with chronic discogenic low back pain: 1-year follow-up of a phase I study. Stem Cell Res Ther
18. Levi D, Horn S, Tyszko S, et al. Intradiscal platelet-rich plasma injection for chronic discogenic low back pain: preliminary results from a prospective trial. Pain Med
19. Tuakli-Wosornu YA, Terry A, Boachie-Adjei K, et al. Lumbar intradiskal platelet-rich plasma (PRP) injections: a prospective, double-blind, randomized controlled study. PM R
20. Zhang YH, Zhao CQ, Jiang LS, et al. Lentiviral shRNA silencing of CHOP inhibits apoptosis induced by cyclic stretch in rat annular cells and attenuates disc degeneration in the rats. Apoptosis
21. Sowa G, Westrick E, Pacek C, et al. In vitro and in vivo testing of a novel regulatory system for gene therapy for intervertebral disc degeneration. Spine (Phila Pa 1976)
22. Hanson DP, Robb RA, Aharon S, et al. New software toolkits for comprehensive visualization and analysis of three-dimensional multimodal biomedical images. J Digit Imaging
1997; 10 (3 suppl 1):229–230.
23. Robb RA, Hanson DP, Karwoski RA, et al. Analyze: a comprehensive, operator-interactive software package for multidimensional medical image display and analysis. Comput Med Imaging Graph
24. Lurie JD, Tosteson AN, Tosteson TD, et al. Reliability of magnetic resonance imaging
readings for lumbar disc herniation
in the Spine Patient Outcomes Research Trial (SPORT). Spine (Phila Pa 1976)
25. Lurie JD, Tosteson AN, Tosteson TD, et al. Reliability of readings of magnetic resonance imaging
features of lumbar spinal stenosis. Spine (Phila Pa 1976)
26. Pfirrmann CWA, Metzdorf A, Zanetti M, et al. Magnetic resonance classification of lumbar intervertebral disc degeneration. Spine (Phila Pa 1976)
27. Videman T, Battié MC, Parent E, et al. Progression and determinants of quantitative magnetic resonance imaging
measures of lumbar disc degeneration: a five-year follow-up of adult male monozygotic twins. Spine (Phila Pa 1976)
28. Fardon DF, Williams AL, Dohring EJ, et al. Lumbar disc nomenclature: Version 2.0: Recommendations of the combined task forces of the North American spine society, the american society of spine radiology, and the american society of neuroradiology. Spine (Phila Pa 1976)
29. Aprill C, Bogduk N. High-intensity zone: a diagnostic sign of painful lumbar disc on magnetic resonance imaging
. Br J Radiol
30. Schellhas KP, Pollei SR, Gundry CR, et al. Lumbar disc high-intensity zone: correlation of magnetic resonance imaging
and discography. Spine (Phila Pa 1976)
31. Modic MT, Steinberg PM, Ross JS, et al. Degenerative disk disease: assessment of changes in vertebral body marrow with MR imaging. Radiology
1988; 166 (1 pt 1):193–199.
32. Elliott DM, Yerramalli CS, Beckstein JC, et al. The effect of relative needle diameter in puncture and sham injection animal models of degeneration. Spine (Phila Pa 1976)
33. Ohtori S, Inoue G, Orita S, et al. No acceleration of intervertebral disc degeneration after a single injection of bupivacaine in young age group with follow-up of 5 years. Asian Spine J
34. Flanagan MN, Chung BU. Roentgenographic changes in 188 patients 10-20 years after discography and chemonucleolysis. Spine (Phila Pa 1976)
35. Song XX, Yu YJ, Li XF, et al. Estrogen receptor expression in lumbar intervertebral disc of the elderly: gender- and degeneration degree-related variations. Joint Bone Spine