Differences in damage to CGRP immunoreactive sensory nerves after two lumbar surgical approaches: investigation using humans and rats.

STUDY
DESIGN: Immunohistochemical study.
OBJECTIVE: To evaluate invasive surgical approaches by analyzing the number of sensory nerve fibers at 2 back muscle sites in rats and humans and the number of injured nerve fibers innervating these 2 sites after muscle injury in rats. SUMMARY OF BACKGROUND DATA.: Several minimally invasive approaches have recently become popular in the treatment of lumbar spine disorders. Minimally invasive surgery (MIS) is not invasive to back muscle and is thought to reduce low back pain. Muscle damage has been generally evaluated by magnetic resonance imaging (MRI); however, damage to sensory nerve fibers in and around back muscle that is directly related to pain has apparently not been explored.
METHODS: Human muscle at L4-L5 was obtained from the paraspinous process (during a midline approach) and from paraspinal back muscle (during a Wiltse paraspinal approach) (n = 10 each). The muscle was sectioned and immunostained for calcitonin gene-related peptide (CGRP). To detect dorsal root ganglion (DRG) neurons innervating back muscle in rats, Fluoro-Gold (FG) was applied to the same 2 sites on the lower back muscle at L4-L5 (only application, n = 12; application of FG + muscle injury, n = 12). DRGs were harvested and immunostained for CGRP and activating transcription factor-3 (ATF-3: marker for nerve injury). The numbers of FG-labeled CGRP-immunoreactive or FG-labeled ATF-3-immunoreactive DRG neurons innervating the 2 sites were counted and compared.
RESULTS: CGRP-immunoreactive sensory nerve fibers were found at the 2 sites. The average number of CGRP-immunoreactive sensory nerve fibers in muscle obtained in a midline approach was significantly higher than that in muscle obtained in a Wiltse paraspinal approach (P < 0.01). The numbers of FG-labeled CGRP- and ATF-3-immunoreactive DRG neurons innervating paraspinous process muscle were significantly greater than those innervating paraspinal back muscle in rats (P < 0.01).
CONCLUSION: There are more CGRP-immunoreactive sensory nerve fibers and DRG neurons innervating muscle in the midline approach area than in the Wiltse paraspinal approach area in humans and rats. There are more ATF-3-immunoreactive DRG neurons innervating muscle in the midline approach area than in the Wiltse paraspinal approach area after muscle injury in rats. This result may show the differences in sensory nerve injury during the 2 surgical approaches.