Comparative analysis of pelvic/sacral versus vertebral chordomas in the US from 2000 to 2020.

PURPOSE: Chordomas are rare tumors of the axial skeleton, with 55% of cases affecting the pelvis/sacrum and 10% affecting the vertebrae of the spine. We seek to provide a comparative analysis of demographic, disease, and treatment variables between these two disease sites.

METHODS: Patients with chordomas were identified via the Surveillance, Epidemiology, and End Results (SEER) Program. Inclusion criteria were malignant chordoma (ICD-O-3 codes 9370/3-9372/3) limited to the pelvis, sacrum, or vertebrae of the spine (PSV). Fisher's Exact, Cox regression, and Kaplan-Meier survival analysis was conducted to assess variation between PSV disease sites.

RESULTS: Eight-hundred-ninety-six patients diagnosed with chordomas of the PSV were identified. Patients with pelvic/sacral chordomas were more likely to be older (p < 0.001), proportionally more non-White (16.7% vs. 9.5%; p = 0.010), come from urban counties (p = 0.042), and have more advanced disease (p < 0.001) compared to those with vertebral chordomas. Patients with pelvic/sacral chordomas were also less likely to undergo radiotherapy (p < 0.001) but were more likely to undergo gross total resection (p < 0.001). Five-year overall survival (5y OS) was 75.4% [71.71-79.16] for pelvic/sacral versus 77.0% [72.81-81.24] for vertebral chordomas, with no difference on Kaplan-Meier analysis (p = 0.729); however, Cox regression demonstrated vertebral chordomas worsened survival (hazard ratio 1.63, p = 0.004). Disease histology was associated with variation in 5y OS (p < 0.001); treatment with surgery was associated with increased 5y OS (p < 0.001 for both disease sites).

CONCLUSION: Primary tumor site was independently associated with survival on multivariable analysis, despite no difference in unadjusted 5y OS between pelvic/sacral and vertebral chordomas. Significant differences in demographic, disease, and treatment characteristics were observed between groups, indicating that underlying factors may contribute to prognosis in unadjusted survival comparisons. Further investigation into the impact of disease histology and treatment modalities is warranted.