Chordoma and giant cell tumor of bone are primary tumors of axial skeleton, with similar clinical symptoms particularly the pain. Although giant cell tumors of bone is considered as a benign disease, they tend to continue to enlarge, destroy bone, and may eventually erode the rest of the bone and extend into the soft tissues. These tumors are notorious for their tendency to recur, and need to be treated aggressively. Chordoma is rare, which is developed from the remnants of the primitive notochord. It tends to occur at the ends of the spinal column, usually in the middle of the sacrum or near the base of the skull. If chordoma is not correctly diagnosed preoperatively, implementing partial resection or scraping of tumor will seriously affect the prognosis of patients and quality of life due to damage of nerves. On imaging, both are characterized with osteolytic bone destruction, heterogeneous signal, and both can have cystic regions [1,2]. If they present typical imaging features, and combing with age and lesion location, they may be correctly diagnosed. In sacrum, chordoma and giant cell tumor of bone are the top one and two common primary tumors [3], and due to their similar imaging findings they can be easily mis-diagnosed. The purpose of this study is to characterize these two tumors with their morphological presentations and pharmacokinetic parameters analyzed from DCE-MRI.26 patients with giant cell tumor in the spine (mean age 33), and 12 patients with chordoma (mean age 47) who had received MRI with a DCE sequence were analyzed. The conventional imaging sequence included transverse T2WI, sagittal T2WI with and without fat suppression, and sagittal T1WI. After the abnormal region was identified, DCE-MRI was performed using the FLASH 3D VIBE sequence with TR of 4.1 ms, TE of 1.5 ms, flip angle of 10º, the matrix of 256 x192, FOV of 250 x250mm. A total of 30 slices, 3 mm in thickness with interval of 0.6mm, were used to cover the abnormal segment. The temporal resolution was 10-14 seconds. The contrast agents, 0.2 [mmol/kg] Gd-DTPA, was injected after one set of pre-contrast images was acquired. A total of 12 frames were acquired, so the total DCE-MRI time period was about 120-168 seconds. For each case, a region of interest (ROI) was manually placed on an area that showed the strongest enhancement, and the signal intensity time course was measured. The two-compartmental pharmacokinetic analysis was applied to obtain the in-flux transport constant Ktrans and the out-flux rate constant kep ( [1/min]), by using the fast population-based blood curve as the reference.The summary results are listed in Table 1. As expected, chordoma only occurred in the cervicales (7/12, 58%) and the sacrales (5/12, 42%), not in thoracalis or lumbalis. All chordoma did not show vertebral compression, and giant cell tumor showed different degrees of vertebral compression. Giant cell tumors were likely to show bone expansion changes (21/26=81%), but not chordoma (1/12=8%). Paraspinal soft tissue mass was a prominent feature for chordoma (100%), but only appeared in about half of giant cell tumor (15/26=58%). The presence of fiber separation in T2 soft tissue was a common finding in chordoma (9/12=75%), but not seen at all in giant cell tumor (0/26=0%). The signal intensity differences on T1WI and T2WI were small. In DCE-MRI, the giant cell tumor was more likely to present a rapid wash-in followed by wash-out pattern, while the chordoma was more likely to present the plateau/persistent DCE patterns. Giant cell tumor had a significantly higher Ktrans than chordoma (0.13±0.65 vs. 0.062±0.041 [1/min]), and a significantly higher kep (0.62±0.22 vs. 0.17±0.12 [1/min]). If using kep=0.43/min as the cut-off value, it achieved 100% sensitivity and 92% specificity to differentiate chordoma from giant cell tumor, with an overall accuracy of (36/38=95%).Chordoma occurs in the cervicales and the sacrales, presents as paraspinal soft tissue mass without vertebral compression, and often exhibits fiber separation in T2 soft tissues. These typical features may be used for differential diagnosis. However, when chordoma occurs in young patients and does not show these typical features, they may be mis-diagnosed as giant cell tumor. These two diseases may need different surgical strategy and subsequent adjuvant therapy, and a correct diagnosis will be very helpful to determine the best treatments for each patient to improve the outcome and prognosis. We have shown that the DCE-MRI enhancement kinetics and the parameters that can be analyzed by using pharmacokinetic models show significant differences between them, and may provide very helpful information to improve diagnostic accuracy.