Management of Recurrent and Progressive Skull Base Chondrosarcomas: A Retrospective Single-Institution Study

Background: Skull base chondrosarcomas are difficult tumors to cure, with a predilection for recurrence and progression over time. There are little data and no consensus on the best management strategy when these tumors recur.

Objective: To identify the presentation and treatment-related factors which impact the progression free survival (PFS) and disease specific survival (DSS) for recurrent chondrosarcomas, and to identify predictors of successful salvage treatment.

Methods: A single institution retrospective review was performed on 17 patients with recurrent/progressive chondrosarcoma over a 25-year period. Survival analysis for presentation and treatment-related factors impacting PFS and DSS was performed using a Cox proportional hazards model. Univariate and multivariate predictors of successful salvage treatment (defined as PFS after salvage treatment ≥ histology specific median PSF for nonrecurrent patients) were identified using logistic regression. Analysis was performed on first recurrence events only, as well as all recurrence events combined.

Results: A total of 47 recurrence or progression events were analyzed from 17 patients with a median of two events per patient (range: 1–8). The median overall PFS and DSS for the initial recurrence was 32 (range: 3–267) and 79 (range: 3–285) months, respectively. For first-time recurrences, failure of prior radiotherapy predicted worse PSF on multivariate analysis (p = 0.0004), while failure of prior chemotherapy or treatment with chemotherapy were univariate predictors of worse PFS (p = 0.04, and 0.03). Treatment with chemotherapy was a univariate predictor of worse DSS (p = 0.003). Gross-total resection (GTR) was the only multivariate predictor of successful salvage treatment (0.01). Across all recurrence events, histologic grade and failure of prior radiotherapy were multivariate predictors of worse PFS (p = 0.0001 and 0.0006), while number of recurrences, pattern of recurrence, and treatment with chemotherapy were univariate predictors (p = 0.031, 0.001, and 0.03). Multivariate predictors of DSS included histologic grade and treatment with chemotherapy (p = 0.0001 and 0.0002), while univariate predictors also included number of recurrences, pattern of recurrence, treatment with radiotherapy, and no surgical treatment (p = 0.04, 0.0003, 0.05, and 0.007). Histologic grade was the only multivariate predictor of successful salvage treatment (p = 0.008).

Conclusion: Recurrence after previous radiotherapy or with higher histologic grade portends a worse prognosis. For initial recurrences, the ability to achieve a GTR may result in PFS comparable to nonrecurrent cases and thus surgery should be seriously considered. While limited, these data provide insights to help guide management of these difficult tumors when they recur.