23 Chordoma

10.1055/b-0040-177079

23 Chordoma

Moujahed Labidi, Shunya Hanakita, Kentaro Watanabe, Vincent Couloigner, Bernard George, and Sébastien Froelich

Abstract

Chordomas are rare benign lesions arising from notochordal remnants in the skull base region and the spine. The majority of pediatric chordomas are located in the spheno-occipital region. Within the pediatric population, the location of chordomas is dependent on age. Cranial chordomas are almost exclusively seen in very young patients, while the sacrococcygeal forms are seen in older adolescents. This chapter explores chordomas, their malignant clinical behavior and surgical removal as the foundation of treatment.

23.1 Epidemiology

Chordomas are benign lesions arising from notochordal remnants in the skull base region and the spine. Their prevalence is low, especially in pediatric patients, as they represent less than 1% of intracranial tumors in this population (with an estimated annual incidence of <1 in 1,000,000 children aged ≤10 years). ¹ ^, ² ^, ³ In a large surveillance, epidemiology, and end result (SEER) database study, only 6.3% of all chordomas affected patients younger than 20 years. ⁴ In chordomas affecting the skull base, the mean age of presentation is 8.8 years. ¹ The majority of pediatric chordomas are located in the spheno-occipital region (~50–60% of children with chordomas), while they are more evenly distributed in adult patients between the sacrococcygeal, mobile spine, and skull base regions. ⁴ ^, ⁵ Even in the subgroup of pediatric patients, the location of chordomas seems to be age dependent, with the very young patients presenting almost exclusively with cranial chordomas, while the sacrococcygeal forms are only seen in older adolescents. ²

23.2 Familial Forms

The majority of chordoma cases are sporadic, but, in some rare cases, their occurrence has been linked to tuberous sclerosis complex or familial forms of chordoma. ² These rare associations, which are overrepresented in the pediatric age group, might provide clues as to the pathophysiology and molecular biology of these tumors.

23.3 Clinical Presentation and Natural History

Cranial chordomas usually present as well-delineated multilobulated lesions of the midline cranial base. They are typically T2-hyperintense, and contrast enhancement is heterogeneous.

Notwithstanding their “benign” histology, these lesions tend to grow on follow-up, with resulting osteolysis of the cancellous bone in which they are embedded. Eventually, they break through the cortical bone and invade the surrounding structures and the intradural compartment. Clinically, children with cranial chordomas most frequently present with a cranial nerve palsy (60% of children), usually a sixth nerve palsy, while headaches (40%) and other symptoms of intracranial hypertension (28%) are also common complaints at presentation. ¹ ^, ⁶ Pediatric chordomas more often present with distant metastasis than adults, especially for those younger than 5 years, although distant metastases were also more associated with sacrococcygeal locations. ³ ^, ⁴ Prolonged survival can be achieved in the majority of the patients, with an overall survival (OS) of 63% at 15 years in a combined series of patients from Hôpital Lariboisière and Hôpital Necker – Enfants Malades. ¹ However, patients younger than 5 years seem to have a worse prognosis. ⁶

23.4 Management Options: Medical, Surgical, and Adjuvant

The mainstay of treatment of pediatric cranial chordomas is surgical resection. Gross total resection (GTR) with preservation of neurological function must be the goal of surgeons undertaking resection of these lesions. In a recent meta-analysis of adult chordoma cases, we found a clinically and statistically significant difference in tumor control between GTR and subtotal or partial resections. ⁷ In a cohort of 610 patients with skull base and clivus chordomas, recurrences were less frequent in complete resections than incomplete resections (24.3 vs. 55.0%, p < 0.0001) with an odds ratio of 0.289 (95% confidence interval [CI]: 0.184 –0.453) of having a recurrence for patients who had a GTR compared with incomplete resections. In the largest pediatric chordoma series, George et al and Ridenour et al also reported a survival advantage when GTR was achieved, although this association between increased survival and GTR did not reach statistical significance. ¹ ^, ⁵

Another important prognostic factor to take into consideration is the quality of the resection at first attempt. In fact, in most surgical series, GTR appears harder to achieve in residual and recurrent diseases than during the first surgical attempt. In chordomas, more than in any other skull base lesions, we very frequently combine different surgical approaches and stages to obtain the desired resection and resulting progression-free survival (PFS) and OS advantages. In our systematic review and meta-analysis of the literature, we found a statistically significant relationship between the rate of multiple surgeries undertaken for chordoma resection and the GTR rate achieved (p < 0.0001). This correlation confirms that in many cases, a GTR or maximally safe resection cannot be achieved through only one corridor.

There is a lack of data specific to the pediatric population regarding the role of radiation therapy in skull base chordomas. However, most authors agree that high-dose irradiation is indicated in all cases where a partial resection was undertaken. After GTR, the role of adjuvant radiation therapy is controversial. ³ ^, ⁸ Chordomas are relatively radio-resistant tumors, and high doses of radiation are required to achieve adequate local control rates (in the 65–75 Gy range). In this context, charged particle radiation therapy (e.g., protons or carbon ions) appear ideally suited, with their steep dose falloff beyond the target that allows high doses to be delivered to the tumor and minimization of toxicity to the surrounding critical structures. ⁹ Even with these highly specialized techniques, late complications of radiation therapy are not uncommon, including partial or complete pituitary insufficiency, visual toxicity, and hearing loss.

Even if significant investigative efforts are engaged in developing new drug therapy for chordomas, none have so far been able to demonstrate a clear benefit in OS and PFS. Novel molecular markers associated with chordoma cells have been identified, and a number of targeted-therapy protocols are currently in trial. ¹⁰

23.5 The Role of Endoscopic Endonasal Surgery

23.5.1 Rationale and Surgical Corridor Selection

To achieve an optimal resection in chordoma surgery, it is fundamental to choose the most appropriate surgical corridor for each lesion treated. Although chordomas preferentially affect the midline skull base region, they also have the a tendency to extend locally and invade different compartments and anatomical structures around the skull base. An experienced multidisciplinary skull base team should carry out a careful examination of the preoperative CT scan and MRI to precisely delineate and record all extensions of the lesion to be resected. Compartments of the chordoma that are out of reach for radiation therapy, such as those in close proximity to the brainstem, optic apparatus, or metallic reconstruction material, should be considered important surgical targets. Larger compartments of the tumor or those causing severe compression of neurological structures or which are responsible for neurological deficits should also be resected. As mentioned previously, staged surgeries and combined approaches are often required to achieve maximal safe resection.

The role of the endoscopic endonasal approach (EEA) in chordoma surgery has increased significantly in the last decade, and the rationale for this approach is solid in many of these lesions. In addition to the fact that they frequently originate from midline bony structures and that cranial nerves need not to crossed to access the tumor through an EEA, these lesions are often soft and easily aspirated, with no brain retraction required. A few studies have compared surgical results between EEA and “classic” posterolateral approaches in adult patients, although selection bias greatly limits interpretation of such data. ⁷ ^, ¹¹ Unfortunately, there are no equivalent data in the pediatric population. However, in a systematic review of the adult literature, we observed significant differences in the rate of GTR obtained in series reporting on chordomas operated exclusively though the endoscopic midline corridor from that obtained in series also including resections through posterolateral approaches (60.7 vs. 42.0%, respectively, p = 0.02). There were also notable differences in postoperative complications between these two groups. As expected, there was a higher number of postoperative cerebrospinal fluid (CSF) leaks in series including endoscopic midlines cases only (22.1 vs. 9.5%, p = 0.06). The opposite was observed with CNS infection (1.9 vs. 6.1%, p = 0.09); however, these differences did not reach statistical significance. There was a trend toward a higher number of de novo postoperative cranial nerve deficits in the series including posterolateral techniques in the surgical armamentarium (16.1 vs. 7.7% in pure endoscopic midlines series, p = 0.10). ⁷ Nevertheless, it is our opinion that these different techniques should not be viewed as competing but as complementary.

The availability of vascularized mucosa and soft tissues for reconstruction needs to be part of the preoperative planning. In fact, these lesions tend to recur, and many patients require multiple surgeries, and a well-vascularized nasoseptal flap (NSF), generally used as first-line reconstruction material, may have already been used or compromised. In children younger than 10 years, septal development may be incomplete, and the surface the septal mucosa can eventually cover may be limited. ¹² Still, its length and surface may be sufficient to cover correspondingly smaller defects in children. ¹³ It is therefore important to pay particular attention to this element when considering the preoperative CT scan and to consider an alternate graft, if necessary. In some cases, when we deem the CSF leak risk too high or expect difficult dissection planes from the brainstem or neurovascular structures, we prefer leaving the intradural component of the chordoma for a second surgical stage through a posterolateral corridor. The use of endoscopic assistance has also proven to be an invaluable tool to increase the reach of these approaches (far lateral, lateral cervical, etc.).

When considering an EEA in a child, three particularly important anatomical limitations must be considered before surgery: the anterior nasal aperture (piriform aperture), the degree of sphenoid sinus pneumatization, and the localization of the internal carotid artery (ICA) within the sphenoid sinus. The piriform aperture is rarely an obstacle in children ≥3 years. As for the sphenoid sinus, pneumatization begins at 2 years, and a sellar-type pneumatization is usually reached between the ages of 6 and 13 years. With adequate preoperative imaging and intraoperative image guidance, the remainder of non-pneumatized sinus can usually be safely drilled away. Although there is only slight variation of the intercarotid distance, especially at the superior clivus, the degree of pneumatization of the carotid prominences should be carefully assessed, as it varies greatly depending on patient age. ¹⁴

Only gold members can continue reading. Log In or Register to continue