The patient is placed in a prone position, ideally with a pillow or abdominal roll under the abdomen to help open up the lumbar interlaminar space by reversing the lumbar lordosis. The skin is then prepped and draped in a sterile manner. The targeted interlaminar space is identified using an anteroposterior (AP) fluoroscopic view, the vertebral body endplates at the targeted level are “squared off ” by adjusting the relative cephalad-caudad orientation of the fluoroscope, and the fluoroscope position is further adjusted so that the proposed needle entry site into the epidural space is centered with respect to the fluoroscopic view in order to reduce parallax error. After the local skin and underlying tissues are anesthetized with 1% lidocaine, a 17- or 20-gauge epidural needle (e.g., Tuohy or Crawford) of appropriate length, depending upon body habitus, is inserted at the injection site. The epidural needle then penetrates the skin, subcutaneous tissue, paraspinal muscles (paramedian approach) or the interspinous ligament (midline approach), and ligamentum flavum, where increased resistance is usually felt. At this point, the needle stylet is removed and the epidural needle is connected, ideally via extension tubing, to a Luer-Lok low friction glass or plastic syringe filled with about 2 mL of preservative-free saline. (Although the syringe can alternatively be filled with air, this can theoretically lead to an air embolus with inadvertent intrathecal injection and is believed to cause a higher incidence of postepidural headaches.) As the operator’s one hand advances the needle slowly into the ligamentum flavum, the other hand exerts steady gentle pressure on the plunger of the syringe. Depending upon the experience of the injectionist and the patient’s body habitus, the entire procedure can either be done using an AP view, or additional lateral views can also be obtained to help judge the depth of penetration. Once the needle penetrates the ligamentum flavum, loss of resistance should be detected by the hand holding the Luer- Lok syringe because saline will be suddenly injected owing to the negative pressure within the epidural space. Aspiration is then performed to ensure no CSF or blood return. (If blood is present, the needle position should be readjusted until no blood return is found. If CSF return is present, the needle is either withdrawn and the procedure attempted at an adjacent level or a caudal or transforaminal approach considered for the epidural.) A small amount of contrast (usually in the range of up to several milliliters) is then injected to visualize an epidurogram pattern that can be described as a Christmas tree, a bunch of grapes, or a vacuolated pattern (Fig. 68-1). Two other contrast patterns are possible if there has been false loss of resistance (in which the needle has not yet penetrated into the epidural space) or accidental needle penetration through the subarachnoid membrane. In these cases, contrast pattern recognition is essential. For example, in situations of false loss of resistance, the injected contrast typically appears as a local accumulation of contrast, whereas a typical myelogram revealing a relatively tubular (column-shaped) contrast pattern is generated when there has been subarachnoid membrane penetration. In the latter situation, the needle should be withdrawn, and the injection can be reattempted at an adjacent interlaminar space or by switching to a caudal or transforaminal approach. Once the needle is confirmed in the epidural space and no vascular pattern is observed upon contrast injection, a mixture of 4 to 10 mL of solution containing 80 to 125 mg of preservative-free methylprednisolone or 12 mg of preservative-free betamethasone sodium phosphate (Celestone Soluspan) and preservative-free 1% lidocaine with or without saline is injected into the epidural space through the epidural needle.
FIGURE 68-1. Lumbar interlaminar epidural injection. AP view showing a typical vacuolated epidurogram.
Several procedural modifications are recommended for cervical or thoracic interlaminar epidural injections due to the presence of the underlying spinal cord. For example, the cervical or thoracic interlaminar epidural injections should not be performed at the level of herniated nucleus pulposus or spinal stenosis, to avoid further potential spinal canal compromise and spinal cord compression. Furthermore, consideration should be given to directing the needle so that it contacts the inferior aspect of the lamina, to provide a clearly felt sense of depth prior to engaging the ligamentum flavum. The needle is then withdrawn slightly and directed into the ligamentum flavum. Further needle advancement should be performed using a lateral view and in addition to using the loss-of-resistance technique, the needle tip should not be advanced further than the laminar line to avoid the potential penetration of the dura mater or spinal cord injury. Epidural dye pattern recognition should be performed after a minimal amount of contrast has been injected since a total volume of less than 4 mL is recommended in these body regions (Figs. 68-2 and 68-3).
FIGURE 68-2. AP view of cervical interlaminar epidural injection demonstrating typical “honeycomb” pattern of epidurogram.