Background – Chordomas are rare ectodermal bone malignancies derived from transformed notochordal remnants. Histologic variants include conventional (80%–90%), chondroid (5%–15%), and dedifferentiated (2%–8%). Because chordomas are relatively resistant to chemotherapy and radiotherapy, novel targeted agents are needed to expand treatment approaches and improve outcomes. This study analyzes the genomic landscape of chordoma and identifies potential pathogenic and druggable targets.

Methods – Eighty‐six tumor samples derived from chordoma patients treated at Massachusetts General Hospital, University of California, Los Angeles, and the University of Miami were included. Tumor specimens were sent for comprehensive molecular profiling using next‐generation sequencing. The most frequently mutated genes were identified and categorized by subtype, and microsatellite instability and programmed death ligand‐1 (PD‐L1) staining were assessed.

Results – Histologic subtypes included 70 conventional (81.4%), nine chondroid (10.5%), and seven dedifferentiated chordomas (8.1%). The most common mutations were cyclin‐dependent kinase inhibitor 2A/B (CDKN2A/B) (28 of 86, 33%), low‐density lipoprotein receptor‐related protein 1B (10 of 86, 12%), polybromo‐1 (9 of 86, 11%), and epidermal growth factor receptor (EGFR) (8 of 86, 9%). By subtype, CDKN2A/B mutation was most common in conventional chordoma (24 of 70, 34%), and chondroid chordoma (3 of 9, 33%). CDKN2A/B and EGFR mutations were most common in dedifferentiated chordoma (2/7, 29%). Microsatellite instability was not detected in seven of 69 (10.1%) samples. PD‐L1 staining of tumor and immune cells was scarce, with scores <1 in 38 of 41 (92.7%) and 22 of 25 (88%) patients, respectively.

Conclusions – This study provides a robust, high‐dimensional sequencing assessment from 86 chordoma tissue samples and a descriptive overview of the genomic landscape of this rare, difficult to treat malignancy. Future studies should include in vitro assessment of gain and loss of function of frequently altered pathways to validate these findings.

VIEW THE PUBLICATION