This study investigated the developmental basis for the human phenotypic morphology of the interaction between the vertebrae and the nerve plexus by evaluating changes in the human lumbar plexus according to various thoracolumbar formulas. The dissection found that the changes in lumbar nerve roots reported by experimental embryology studies to be concomitant with thoracolumbar trade-off, i.e., a change in vertebrae from thoracic to lumbar with no change in the overall thoracolumbar count, were not apparent in humans with the usual 17 or mutant 16 thoracolumbar vertebrae. When vertebral changes in two segments were examined by comparing spines with a reduced thoracolumbar count of 16 to those with an increased count of 18, this tended to show only a single-segment caudal shift of the lumbar plexus. We cannot provide evidence for the phylogenetic difference in the concomitant changes of lumbar nerves and vertebrae, but comparisons between experimental rodents and humans highlighted fewer and shorter lumbar vertebra and more complicated lumbar plexus in humans. Therefore, these multiple differences may contribute to a human phenotypic morphology that is not evident in the concomitant transformation of vertebrae and lumbar nerves reported in experimental rodents.

Background: Experimental evidences identified that thoracolumbar mutants caused by Hox genes 7-10 mutants also involve a craniocaudal shift and/or the addition or reduction of segments of the limb plexus roots. This study investigated whether the theoretical concomitant shift of the brachial plexus roots in human different thoracolumbar counts is shared as confirmed in those of the human lumbosacral plexus. Material and methods:The phenotypic morphology of the brachial plexus and its arterial interaction on 20 sides of 10 atypical human thoracolumbar counts out of the 354 sides of the 177 cadavers, were compared with those of 52 sides of 26 cases in a typical human vertebral formula (7C_12T_5L_5S).Results: Regardless of the course and branching patterns of the axillary artery, our results showed that the main brachial plexus roots were composed of only five segments of the 5 th -9 th spinal nerves, with small contributions from the 4 th and/or 10 th nerves. This root composition is identical to a typical human thoracolumbar formula, and therefore, neither a craniocaudal shift nor additional/reduced main roots occurred in our thoracolumbar variants. Conclusions:Unlike the concomitant shift of the lumbosacral plexus roots, our present cases suggest that the phenotypic morphology of the human brachial plexus may be less likely to 1 show theoretical craniocaudal shifts, further data on the root changes in different vertebral formulae are needed for its accurate validation.