Pain Improvement after Intradiskal Lidocaine Administration in Provocation Lumbar Diskography: Association with Diskographic Contrast Leakage

Lumbar Diskography Technique

Diskography was performed in a standard fashion after appropriate informed consent. The patient's pain history, including leg-versus-back pain contribution, was clarified, and the most-severe-versus-immediate preprocedural pain level was documented using the 0–10 visual analog scale (VAS). Limited intravenous conscious sedation (fentanyl, 0.05 mg; midazolam [Versed]), 1 mg) was given before the procedure with a dose of fentanyl typically given at the end of the examination. Administration of additional fentanyl during the examination was necessary only on rare occasions when disk provocation resulted in extremely severe patient pain that was not responsive to intradiskal anesthetic, and the level of consciousness was never affected. Supplemental administration of additional conscious sedation during the diskography was generally avoided. The lower lumbar region was cleansed and draped, trajectory to target disk space was identified with C-arm fluoroscopy (OEC Series 9800; GE Healthcare, Milwaukee, Wis), and local anesthetic was applied. Disk access was achieved employing a double-needle technique using a 20-gauge guiding spinal needle followed by a curved long 25-gauge spinal needle into the center of the disk space. All needles were placed concordantly opposite the side of leg pain before individual disk injection. Anticipated normal/control disk level was typically studied first.

Disk spaces were studied and provoked by a moderate-to-rapid hand injection of 1.5- to 4.5-mL iohexol, 240 mg Iodine/mL, delivered by a 3-mL syringe and a small 0.6-mL-volume connection tube under direct fluoroscopic guidance. Injection volume was dependent on 1) disk-volume end point, 2) clearly established severe pain response, or 3) exaggerated capacity in degenerative disks. Patients were kept unaware of whether a level was being provoked or which level was being studied. Response to injection was initially observed by the operator, and with a positive pain response, the features of the pain were clarified, VAS level of pain was established, and these items were recorded in a manner similar to the technique of Walsh et al.²⁷ Concordant pain was recorded if the provoked pain was the patient's typical or familiar pain, and nonconcordant pain was recorded if the provoked pain was not the typical pain. Fluoroscopic spot film imaging was obtained for each disk level in anteroposterior and lateral projections during and following the injections.

During disk testing, the patient's immediate response to injection, response to injection end point (if present), and perception of provoked pain (concordant or diskordant) were the primary focus of the diskographer. Syringe disk pressures were not recorded during injection.

If a significantly painful disk space (typically VAS ≥7, concordant or nonconcordant) was encountered, preservative-free lidocaine (2% strength, 1- to 1.5-mL lidocaine (Xylocaine-MPF) was injected into the disk in an attempt to reduce the patient's provoked pain and to allow response clarity in subsequently studied lumbar levels.

The patient was routinely questioned regarding any pain reduction; after administration of the intradiskal lidocaine, the response was recorded as either 1) complete or near-complete pain relief, 2) partial pain relief, or 3) no significant pain relief or as a specific VAS-grade reduction from the pain generated by disk provocation relative to the baseline pain, depending on patient's ability to express the change. Responses initially reported with the VAS were converted to the 3-point scale by calculating the percentage of pain reduction relative to baseline VAS pain level (>66% pain reduction, complete or near-complete pain relief; 33%–66% pain reduction, partial pain relief; and <33% pain reduction, minimal or no significant pain relief).

Postdiskogram CT (GE Healthcare, Milwaukee, Wis) was obtained in all patients immediately following the diskogram employing bone and soft-tissue algorithms with either direct axial 3-mm-section acquisition or with a spiral technique and 3-mm axial and sagittal reformatting with isotropic voxels.

Imaging and Lidocaine Response Analysis

Diskographic fluoroscopic images with the postdiskographic CT of all significantly painful lidocaine-treated disk spaces were independently and blindly reviewed by 2 neuroradiologists experienced in lumbar diskography. Identification of epidural diskographic contrast leakage was primarily established by assessment of the fluoroscopic images with secondary inspection and correlation with the postdiskographic CT. Significant leakage at the disk margin around the needle-entry site was classified as a true leak, but minimal contrast identified at needle entry only after needle withdrawal was not classified as leakage. Disks were judged as either “contained” (no contrast leaking from the disk space) or “leaking” (epidural contrast leakage from the disk space due to radial tear, combined annular and radial tear, or annular and nuclear herniation), itemized, and tabulated. Discordant judgments were resolved by consensus.

Imaging identification of diskographic leakage in the significantly painful treated disks was compared with the 3-point response to lidocaine administration: 1, complete or near-complete relief; 2, partial pain relief; and 3, minimal or no significant relief.

Statistical Analysis

Statistical analysis was performed by χ² comparison between complete response and no response to lidocaine in contained and leaking disks and a logistic regression model adjusting for the disk state to compare the 2 diskographers regarding the proportion of responses among patients (SAS release 8.2, SAS Institute, Cary, NC).

Contained Painful Disks: No Relief

In 69% of painful contained disks, intradiskal lidocaine administration resulted in no significant improvement in their provoked pain (Figs 1 and 2). Despite the degenerative fluoroscopic and postdiskographic CT imaging appearances, pain relief was not achieved when these internal deranged surfaces (disk degeneration, annular degeneration, and annular tears) were bathed with local anesthetic. LBP in these disks is likely generated peripheral to the internal disk surfaces, perhaps in the peripheral annulus or posterior longitudinal ligament. One could question, in these instances, whether peripheral annular injury was present and not recognizable at imaging.

Contained Painful Disks: Partial or Complete Relief

Complete pain relief was seen in 20% of painful contained disks, and partial pain relief occurred in 11% (Fig 3). This observation strongly suggests that the visualized internal degenerative features are related to pain provocation for that disk level. The existence of pain nociceptive fibers in the disk has been long debated. Free nerve endings suggesting pain fibers have been documented in the peripheral one third of the annulus and PLL, most commonly associated with small penetrating blood vessels.^1,31,34,38 Free nerve endings have been observed extending as far as the peripheral half of the annulus in patients with disk herniation.⁵¹ Neovascularization and substance P immunoreactive innervation (nociceptive pain-mediator fibers) of the inner one third of the annulus (46%) and nucleus (22%) have been described in operative biopsy of painful disk levels⁵² and recently reconfirmed with calcitonin gene-related peptide-expressing nociceptor innervation demonstration in the outer and inner annulus of postoperative specimens.⁵³ Local anesthetic action on these annular or nuclear pain fibers could be responsible for the pain resolution or improvement noted in this subset of our patients with painful contained disks.

Painful Disks with Diskographic Leakage: Complete Response

In 74% of painful leaking disks, complete or near-complete resolution of LBP occurred after intradiskal lidocaine administration (Figs 4 and 5). In these lumbar disks, local anesthetic was either reaching the location of the LBP generator or blocking its central neural transmission.

LBP in these patients could be generated at the annular/PLL surface at the radial tear or herniation. Alternatively, irritation of epidural structures (vessels) or dura could be the source of pain and relieved by the local anesthetic. It is well known to diskographers that leaking disks can reproduce severe and concordant LBP after only minimal contrast injection or leakage. Patients are also encountered in whom nonspecific/semiconcordant LBP is provoked with initial disk injection, but exact replication is recognized only when greater contrast volume and pressure are used and focal contrast leakage finally occurs.

The irritant effects and by-products of disk degeneration such as phospholipase A2, substance P, inflammatory cells, and hydrogen ions have been suggested as a source of LBP and radiculopathy.^40–48 Unmyelinated fibers (somatic and autonomic) are known to be present in the peripheral annulus/PLL but are also concentrated in epidural blood vessels and the ventral dura. Radiating pain patterns (radiculopathy) at times resemble a nerve root distribution and at times are more indistinct, variable, or referred, perhaps reflecting the distribution of the somatome (somatic and autonomic innervation including dermatome, myotome, and sclerotome), and this distribution has been termed the “radiating zones of head.”² A vascular pain fiber (epidural, paravertebral vessel, or dural) source of pain fiber irritation and LBP might explain a portion of the response to lidocaine in painful leaking lumbar disks.

Painful Disks with Diskographic Leakage: Partial Response or No Response

The fourth important observation is related to the absent improvement (11%) or partial improvement (15%) to local anesthetic noted in the painful leaking disks. Considering the response of contained disks, it is likely that some leaking disks also have internal derangement or peripheral annular injury as the source of their pain, which does not come in contact with the intradiskal local anesthetic. Absent pain improvement (11%) despite leakage suggests that the leak pathway and leaking substances may not be the pain generator and that pain is being produced elsewhere in the disk. Partial pain improvement (15%) suggests that the leak pathway and leaking products may partially contribute.

Pain origin in leaking disks, therefore, could be related to the visualized internal derangement, peripheral annular margin injury, leakage path, or irritant effects of epidural leakage of inflammatory material. Although a simple epidural anesthetic mechanism of pain extinction could be suggested for relief in the leaking disks, clearly complete pain ablation is not encountered in all instances. In addition, in many leaking disks, the leak is small and focal, and widespread leakage of contrast (and presumably lidocaine) is not present.

We have not embraced injection-pressure measurement during provocation lumbar diskography, despite familiarity with the technique. In our experience, pressure measurements have rendered limited insight into confirmation of or characterization of (concordant-versus-diskordant) objectively provoked diskogenic pain beyond a perceived end point to injection. This result is in part supported by the lack of clear postoperative improvement when pressure-controlled diskography is used in patient selection.^54–56 Of interest, a statistically significant difference in surgical outcome between diskectomy/intradiskal fusion in low-pressure (leaking?) disks compared with fusion only in high-pressure (contained nonleaking?) disks has been demonstrated in a subset comparison within 1 study.⁵⁶

Several limitations are present in this retrospective study, despite the large number of patients and disks evaluated. In contained disks, internal disk derangement is complex, flow of injected material is unpredictable, and uniform spread of the injected anesthetic may not occur, leading to a reduced rate of improvement in these disks. In leaking disks, early epidural contrast reabsorption may lead to under-recognition in subtle cases with small leaks, not originally appreciated at fluoroscopy. Second, many leaking disks cannot be completely distended, and areas of annular injury may not be fully tested with anesthetic or adequately provoked without more aggressive biomechanical strain. Third, anesthetic leakage, if excessive, could potentially mask the response at another level, but in our experience, adjacent levels can be correctly tested after anesthetic administration.