BackgroundEnchondroma, a subtype of chondroma, originates from the medullary cavity of the bone and produces an expansile growth pattern. Enchondroma located in the spine is rare and a few cases of large thoracic enchondroma have been reported. The authors document a rare case of large enchondroma in the thoracic spine of a 49-year-old woman, and discuss its clinical, radiological and histopathological characteristics.Case presentationThe patient presented with rapidly progressive and severe pain on her upper back. Magnetic resonance imaging revealed an expansile lesion at the posterior elements of T3 that was hypointense on T1-weighted images and mixed iso- to hyperintense on T2-weighted images. Administration of gadolinium-diethylenetriamine pentaacetic acid (Gd-DTPA) resulted in heterogeneous enhancement. During surgery, a large tumor of 4.2cm × 4.7cm × 2.1cm was resected along with the lamina and spinous process. Histological examination revealed that the tumor consisted of mature hyaline cartilage with typical chondrocytes, indicating that it was an enchondroma.ConclusionsDespite its benign-growing nature, enchondroma should be examined closely for signs of enchondromatosis and enchondrosarcoma. Complete surgical resection is the treatment of choice for immediate relief of symptoms and avoidance of recurrence.

BackgroundSignificant prominence of iliac crests with a deep seated L5 vertebra can potentially interfere with the screw trajectory when placing percutaneous pedicle screws (PPS) at the lumbosacral segment. The objective of this study was to investigate the influence of L5 position in relation to the iliac crests on the accuracy of percutaneous placement of lumbosacral pedicle screws.MethodsFrom Oct 2012 to Sep 2014, 54 patients who underwent PPS placement at L5-S1 segment were recruited. Patients were divided into 2 groups: the L5-Seated Group (L5-S Group, n = 34) including patients with intercrest lines passing through the L4 vertebra or L4/5 intervertebral disc; whereas the L5-Non-Seated Group (L5-NS Group, n = 20) including patients with intercrest lines passing through the L5 vertebra. Postoperative computerized tomography was obtained in all patients, and PPS accuracy was evaluated by grading pedicle breach (Grade 0, no breach; Grade 1, ≤2mm; Grade 2, >2mm without neurological compromise; Grade 3, with complications). Screw convergence angle (SCA), defined as the angle subtended by the screw axis and vertebral midline, was also recorded.ResultsIn the L5-S Group, 82.4% (56/68) screws were measured as Grade 0 at L5, and 66.2% (45/68) were Grade 0 at S1; meanwhile, in the L5-NS Group, 77.5% (31/40) and 75.0% (30/40) screws were Grade 0 at L5 and S1, respectively. Misplacement rate was numerically higher at S1 in the L5-S Group (P > 0.05). There were significantly more medial pedicle violations at S1 in the L5-S Group as compared to the L5-NS Group (25.0% vs 7.5%, P = 0.024). No statistical difference was found in L5 SCA between the 2 groups (L5-S Group 23.7° ± 7.4° vs L5-NS Group 23.4° ± 10.6°, P = 0.945); however, S1 SCA was significantly smaller in the L5-S Group (14.7° ± 5.8°) when compared with the L5-NS Group (20.8° ± 5.2°) (P = 0.036).ConclusionsA deep seated L5 vertebra with respect to the iliac crests might compromise the accuracy of PPS placement at S1 vertebra. Severe iliac prominence may interfere with the screw trajectory and limit the medial angulation of pedicle screw for percutaneous S1 fixation.