Jennifer Massie, MS, Alexander Giurea, MD, Sonya Waters, MD, David Amiel, PhD, Steven R. Garfin, MD, Wayne H. Akeson, MD, San Diego, CA, USAIntroduction: Epidural fibrosis is one of the contributing factors that have led to chronic pain after spinal surgery or spinal cord injuries. The dense scar formation on the dura surrounding the neural elements has been described as the “laminectomy membrane.” Separate from potential neurologic compromise, it makes subsequent spine surgery technically difficult. A number of substrates have been used in an attempt to inhibit posterior laminectomy scar formation. For instance, a single therapeutic topical application of a high-molecular-weight hyaluronan (HA) gel has been applied to the epidural space experimentally after a laminectomy just before wound closure. This approach employs the concepts of fetal wound healing. Fetal healing occurs in a unique extracellular matrix environment in which one of the key conditions is a high concentration of high-molecular-weight hyaluronan. Presently, Adcon-L (anti-adhesion barrier gel) is used clinically with mixed postoperative results. Laminar barriers, as opposed to gels, have been recently advocated as they reduce cellular trafficking and decrease vascular invasion into the epidural space from the overlying muscle and bone, without interfering with the more posterior muscle and other tissues critical for wound healing. Control studies between the multiple options have not been performed. The purpose of this study is to compare high-molecular-weight HA gel (Healon; Pharmacia) and Adcon-L gel (poly-sulfo-alpha-[1-6]-D-glucan, etc., Gliatech Inc.) to the effect of a thin barrier sheet (a 70:30 absorbably poly L-lactide-co-D and Poly L-lactide, Macropore Corporation).Methods: Forty-seven male adult Sprague-Dawley rats, weighing 400+grams, were used (approved by Veterans Affairs Institutional Animal Care Use Committee). Under anesthesia a posterior midline incision was made exposing the bony posterior elements from L4 to L7. Bilateral laminectomies were performed at L5 and L6. The dura and neural elements were then gently restracted medially, exposing the posterior lateral aspect of the L5–6 disc. A 26-gauge needle was then inserted into the exposed disc creating a controlled disc injury. After obtaining hemostasis and irrigation, an anti-inflammatory agent was applied over both laminectomy sites. The rats were divided into five groups: 1) normal controls without surgery, 2) laminectomy without treatment, 3) 0.1 cc of HA gel, 4) 0.1 cc of Adcon-L gel and 5) an application of a Macropore barrier sheet. The Macropore barrier sheet was inserted between the dural sleeve and the paravertebral musculature conforming readily to the ronguer laminae. The wound was closed in a routine manner, and the animal recovered. After 3 weeks, the animals were euthanized.The L5 segmental nerve roots were dissected free bilaterally through an anterior approach. The segmental nerve roots were excised, including the portion of the nerve root within the foramen (1 cm in length). Additionally, the dura was removed from the caudal aspect of the body of L4 to the cephalad aspect of the body of L7 (1.5 cm in length), including the attached scar. The samples were analyzed biochemically by extracting the fat, then obtaining dry weights, followed by hydrolyzing the tissues using hydrochloric acid to determine the amount of total collagen from the hydroxyproline content. This collagen was expressed in milligrams of fat-free dry tissue.Each treatment group was compared with both the normal controls and the operated but untreated controls using a Fisher's multiple comparisons paired t test. Additionally, each treatment group was compared with each other using a one-way analysis of variance.Results: In the untreated laminectomy specimens, the total collagen increased more than two-fold in the dura (p value of .0009). Additionally, the percent collagen increased significantly in both the dura and nerve roots (p values .0008 and .005, respectively). Treatment with HA (p value .0103), Adcon-L (p value .0039) or the Macropore barrier sheet (p value .0020) significantly reduced the amount of total collagen in the dura. However, the Macropore barrier sheet led to a decrease approximately 50% more in total collagen compared with the HA and Adcon-L groups. In the nerve roots, the amount of total collagen was not significantly affected by the treatment.Discussion: Measurements of total collagen provides quantitative data on postlaminectomy scar formation. Gross findings and biochemical analysis demonstrate that the untreated laminectomy scar becomes adherent to the dorsum of the dura mater, an undesirable outcome. Both HA and Adcon-L demonstrated a beneficial effect on scar inhibition at the level of the dura. Because of their short half-life, these compounds probably act by impeding the inflammatory component of wound healing in the area of topical application. All gels, as they migrate posterior, could have the potential to interfere with normal posterior wound healing. The barrier sheet, by walling off the overlying muscle, may be more effective in protecting against cellular trafficking and vascular ingrowth, and because it is a confined substrate, it should not interfere with normal posterior wound healing.
Read full abstract
Aug 26, 2006
Jean E Hsu + 4
Open Access