Surgical management of skull base chordomas and chondrosarcomas: insights from a national cohort study

Abstract

Objective: Skull base chordoma and chondrosarcoma are distinct sarcomas of the skull base but share significant therapeutic challenges due to their proximity to critical neurovascular structures, making surgical resection difficult. We sought to establish factors associated with outcome predictors in a national cohort of patients.

Methods and analysis: Data for all patients referred with a diagnosis of skull base chordoma or chondrosarcoma from April 2017 to December 2022 were obtained. We performed analyses of data pertaining to the first cohort of patients treated in the UK with proton beam therapy (PBT) to determine factors associated with obtaining gross total resection (GTR) and adequate clearance of the brainstem and optic apparatus.

Results: Of 230 patients with skull base chordoma or chondrosarcoma referred for PBT, 71% were accepted for PBT, with a wide regional variation between referring neurosurgical units (29%-93%). Of the first 75 consecutive patients treated with PBT, the only factor predictive of obtaining GTR was surgical resection at a unit with higher volumes of patients accepted for PBT (OR 1.32, 95% CI 1.11 to 1.63, p=0.004). Use of intraoperative MRI (OR 4.84, 95% CI 1.21 to 27.83, p=0.04) and resection at a higher volume unit (OR 1.29, 95% CI 1.07 to 1.64, p=0.013) were associated with increased rates of tumour clearance from the brainstem/optic apparatus.

Conclusions: Treatment at a higher volume centre was a key determinant of the optimal surgical outcome in this cohort. These data support the management of skull base chordomas and chondrosarcomas in higher volume centres where multidisciplinary experience can be accumulated.

Keywords: Radiotherapy; Sarcoma.

Figure 1

(A) Illustration demonstrating the anatomical classification of tumour location used for this study, as previously described by Wang et al. (1) Sphenoclival type—tumour in the upper two-thirds of the clivus, main body in the middle line, anterior to the sphenoid sinus and posterior to the prepontine cistern. (2) Occipitocervical type—tumour in the lower one-third of the clivus and the foramen magnum, main body in the middle line, anterior to the pharynx and posterior to the premedullary cistern. (3) Sphenopetrosal type—tumour in the parasellar or petrous apex, main body in the middle fossa, medial to the cavernous sinus and intracavernous segments of the carotid artery displaced medially or laterally. (4) Petrooccipital type—tumour in the posterior fossa laterally, from the lateral side of the Meckel cavity to the cerebellopontine angle and to the jugular foramen. (5) Ethmoid-Sphenoid type—tumour in the anterior fossa and posterior to the pituitary fossa and (6) extensive type (not shown)—tumour with a huge volume involving at least two of the five parts mentioned above. (B) Stacked column graph illustrating the relationship between pathology and anatomical location, as classified by classifications of tumour location. As expected, the majority (35/40, 88%) of chordomas were located in the sphenoclival or occipitocervical regions. A significant proportion of chondrosarcomas (29/35, 82%) were located in the spheno-petrous region or spheno-occipital region.

Figure 2

(A) Stacked bar graph demonstrating the wide regional variation in outcomes following surgery for skull base sarcoma; the acceptance rate ranged from 29% to 93% (B) Scatterplot with regression line demonstrating the colinearity between the number of referrals from each centre and the proportion of cases accepted for PBT. An outlying high-volume, low-acceptance rate centre is highlighted in red. 95% CI of the regression coefficient is indicated by grey shading.

Figure 3

Graphical representation of a linear regression model demonstrating the influence of the number of patients accepted for PBT on the predicted probability of obtaining a GTR during resection of a chordoma/chondrosarcoma. The probability of achieving GTR increased linearly as the number of patients per unit accepted for PBT increased. 95% CI of the regression coefficient is indicated by grey shading. GTR, gross total resection; PBT, proton beam therapy.