Clinical outcome of proton therapy for patients with chordomas

Abstract

Purpose: To investigate the clinical outcome of proton therapy (PT) in patients with chordoma.

Materials and methods: Fifty-eight patients with chordoma treated with PT between June 2007 and December 2015 at the National Cancer Center, Korea, were retrospectively analyzed. The median total dose was 69.6 cobalt gray equivalent (CGE; range, 64.8 to 79.2 CGE). Local progression-free survival (LPFS), distant metastasis-free survival (DMFS), overall survival (OS), and diseasespecific survival (DSS) rates were calculated by the Kaplan-Meier method.

Results: With the median follow-up of 42.8 months (range, 4 to 174 months), the 5-year LPFS, DMFS, OS, and DSS rates were 87.9%, 86.7%, 88.3%, and 92.9%, respectively. The tumor location was associated with the patterns of failure: the LPFS rates were lower for cervical tumors (57.1%) than for non-cervical tumors (93.1%) (p = 0.02), and the DMFS rates were lower for sacral tumors (53.5%) than for non-sacral tumors (100%) (p = 0.001). The total dose was associated with both the LPFS rate and DMFS rate. The initial tumor size was associated with the DMFS rate, but was not associated with the LPFS rate. Three patients had grade 3 late toxicity with none ≥grade 4.

Conclusion: PT is an effective and safe treatment in patients with chordomas. The tumor location was associated with the patterns of failure: local failure was common in cervical tumors, and distant failure was common in sacral tumors. Further refinement of PT, such as the utilization of intensity modulated PT for cervical tumors, is warranted to improve the outcome.

Keywords: Chordoma; Complications; Proton therapy; Treatment outcome.

Fig. 1.

Survival curves using Kaplan-Meier method. (A) Local progression-free survival, (B) distant metastasis-free survival, (C) overall survival, and (D) disease-specific survival curves and the 5-year survival rates.

Fig. 2.

Local progression-free survival (LPFS) estimation using Kaplan-Meier analysis (log-rank test). (A) The 5-year LPFS rate: skull base vs. cervical spine vs. sacrum (97.0% vs. 57.1% vs. 86.5%, respectively, p = 0.059). (B) The 5-year LPFS rate: total dose <69.6 CGEa) vs. ≥69.6 CGE^a) (63.5% vs. 92.8%, respectively, p = 0.050). CGE, cobalt gray equivalent in Gy. ^a)75.2 Gy of EQD₂ (equivalent dose delivered in 2 Gy fractions, assuming an α/β ratio of 3 Gy.

Fig. 3.

Distant metastasis-free survival (DMFS) estimation using Kaplan-Meier analysis (log-rank test). (A) The 5-year DMFS rate: skull base vs. cervical spine vs. sacrum (100.0% vs. 100.0% vs. 53.5%, respectively, p = 0.001). (B) The 5-year DMFS rate: initial tumor size <5 cm vs. ≥5 cm (100.0% vs. 59.5%, respectively, p = 0.001). (C) The 5-year DMFS rate: total dose <69.6 CGE a) vs. ≥69.6 CGE ^a) (58.3% vs. 95.0%, respectively, p = 0.016). CGE, cobalt gray equivalent in Gy. ^a)75.2 Gy of EQD₂ (equivalent dose delivered in 2 Gy fractions, assuming an α/β ratio of 3 Gy.

Fig. 4.

A 68-year-old man with a >25-cm of sacral chordoma involving pelvic cavity and gluteal soft tissue. He was successfully treated with partial removal of the chordoma followed by proton therapy (PT). Red-line indicates gross tumor volume; yellow-line, omentum as a spacer between the tumor bed and other pelvic organs (e.g., bladder and rectum); cyan-line, bladder; greenline, rectum; MRI T2WI, axial T2-weighted image on magnetic resonance imaging; CT, computed tomography.