Radiation Treatment for Chordomas

Chordomas are rare tumors that arise from remnants of the notochord. They are encountered in the skull base and upper cervical spine or in the sacrum. Skull base chordomas are slow-growing tumors that grow in a locally infiltrative pattern within which an oncologic resection is precluded in the vast majority of cases. ¹^,²^,³ Most of our knowledge on the optimal management of clival chordomas is based on retrospective studies and meta-analyses. ³^,⁴^,⁵^,⁶^,⁷ The optimal treatment strategy for skull base chordomas remains poorly guided due to the lack of quality evidence. Surgical resection was established as the backbone of management. Radiation therapy was not used after complete resections but was utilized when there was residual tumor tissue, recurrent tumor, or if the tumor was unresectable at presentation. ⁸^,⁹^,¹⁰ Due to the close proximity of the clival chordomas to critical neurovascular structures such as the brainstem, cranial nerves, and major intracranial vessels, the extensive surgical resection is challenging and carries risk of significant morbidity. These factors limit a complete surgical excision, and macroscopic or microscopic residual disease is encountered in a vast majority of cases. In a recent analysis of the Surveillance, Epidemiology, and End Results (SEER) database of the USA, published in 2014, 394 histologically confirmed clival chordoma patients were identified and almost half of them (45%) were treated with postoperative definitive radiation therapy (RT). ⁸ A meta-analysis from Israel, also published in 2014, of 467 skull base chordoma cases has shown that 84% of the patients have been treated with RT as part of their management strategy. ¹¹ Technical advances in surgery now permit a safer, more complete tumor resection than has been possible in the past. Nevertheless, the role of adjuvant RT for these patients remains undefined. In the setting of a subtotally resected tumor, local recurrence of progression is inevitable and causes severe disability and ultimately death. RT plays an important role in the multimodality treatment for such tumors. Many institutions, regardless of the extent of resection, recommend adjuvant radiation as the standard therapeutic approach after surgery for all skull base chordoma patients due to significant local failure rates with surgery alone. ¹²

Initial high failure rates after adjuvant radiation therapy (in the range of 45 to 60 Gy) created a widely accepted concept that chordomas are radioresistant tumors. However, various studies have proven that tumor growth control can be achieved with radiation therapy if sufficiently high doses are used. As in the case of surgery, delivery of high doses of radiation to these tumors can be challenging due to the close proximity of dose-limiting organs such as the optic tract, brainstem, and cavernous sinus and the cranial nerves within it. In historical series of patients treated with radiotherapy, where moderate doses of around 50 Gy in 25 fractions have been used, low tumor control rates of around 20 to 40% ³^,⁴ and low complication rates have been reported. However, this is mainly due to application of ineffectively low doses of RT due to fear of probable severe complications resulting from delivery of higher doses to nearby critical structures. However, more recent series have used higher doses in the range of 70 to 75 Gy and reported better local control (LC) rates and relatively modest rates of complications. ¹³ The probability of complications is high with radiation doses above 50 Gy, as the critical structures have lower tolerance doses than the doses needed to adequately ensure control of the tumor. Fortunately, dramatic technical improvements have allowed the developed of intensity-modulated radiotherapy (IMRT), stereotactic radiosurgery (SRS), and charged particle radiotherapy techniques. These innovations allow for a more precise dose delivery to the areas at risk for residual tumor while reducing the dose delivered to nearby critical normal structures. This chapter will focus on the technical characteristics and outcomes of contemporary external beam radiotherapy and SRS for clival chordomas using gamma rays and X-rays. Charged particle therapy will be the topic of Chapter 26.

25.2 Conventional External Beam Radiotherapy

Historically, radiotherapy was delivered in multiple daily fractions of 1.8 to 2.0 Gy per fraction, five times a week, to total doses of 45 to 50 Gy for microscopic disease using megavoltage photon beams. For macroscopic residual disease, this would be followed by a 10-16 Gy boost. Because these tumors are considered relatively radioresistant, high doses of radiation are required to reduce the risk of local recurrence and to increase disease-free (DFS) and overall (OS) survival. ¹⁴^,¹⁵ However the optimum radiation dose is yet to be determined, as no randomized study have addressed this issue yet. The medical literature to date consists of a small number of retrospective studies with a low number of patients treated with different methodologies; consequently, the rarity of these tumors prevents conducting large randomized controlled trials. Due to the indolent nature and late recurrence pattern of clival chordomas, studies with long follow-up are needed in order to draw definitive conclusions regarding the efficacy and safety treatments. Interpretation of the current literature is very difficult due to short median follow-up times. To our knowledge, there is no level I evidence to suggest an optimal dose of conventional radiotherapy for chordomas, and it seems that it will not be available in the near future. Herein, we will focus on available studies and reviews in the literature ²^,⁴^,¹²^,¹⁶^,¹⁷^,¹⁸^,¹⁹^,²⁰^,²¹^,²²^,²³^,²⁴^,²⁵^,²⁶ ( ▶ Table 25.1).

In a series of 13 skull base chordomas out of 48 cranial and spinal chordomas, Catton et al demonstrated a LC rate of 8% (1/13) with a median survival of 62 months using doses of 40 to 60 Gy, with no severe complications. ⁴ However, others have proposed that doses of > 70 Gy are needed to control the tumor and that doses of < 40 Gy are inadequate. ¹⁷ Fuller and Bloom, from the Royal Marsden Hospital, reported on 13 patients with skull base chordomas treated with conventional RT after subtotal resection or biopsy between 1952 and 1981 with a follow-up more than 5 years. The radiotherapy total dose ranged from 45 to 65 Gy (median: 55 Gy) delivered in 1.5 to 1.7 Gy per fraction. Nine of the 13 patients died from locally recurrent tumor in 2 to 85 months; the authors reported that subtotal resection before RT did not favorably influence survival. ² Debus and colleagues from Heidelberg, Germany, have reported the outcomes of 37 chordoma patients after postoperative fractionated stereotactic RT (SRT), all with macroscopic residual tumors. ²³ A median radiotherapy dose of 66.6 Gy with a median daily fraction size of 1.8 Gy was delivered. Fractionated SRT was preferred because of the sharp dose gradient required to spare surrounding normal tissues. However, radiosurgery was not used due to the large size of tumors (median: 55 cc). They have shown that 5-year progression-free survival (PFS) and OS rates are 50% and 82%, respectively. However, the follow-up was relatively short, 27 months, in this study. ²³ Another study reported 18 chordoma patients treated with conventional RT after subtotal resection. ²⁴ The median dose was 60 Gy, and the mean follow-up was 43 months. Even though the dose was reported to be over 60 Gy, the outcomes were worse than most of the other studies using similar doses, with 23% PFS and 35% OS rates at 5 years. ²⁴

Due to the disappointing results with conventional RT, charged particles, with their unique physical and biological characteristics, have been utilized in the treatment of clival chordomas. ²⁷ In the planning of the delivery of radiotherapy, attempts are made to maximize the therapeutic ratio, which is the theoretical ratio of the risk of injury to normal tissues in relation to the likelihood of tumor control by maximizing the dose of received by the tumor. However, despite the theoretical advantages of charged particles, the reported outcomes with their use have not clearly demonstrated their superiority over contemporary external beam radiotherapy. ⁵^,⁶^,²⁸^,²⁹

Since the mid 2000s, linear accelerator technology has improved immensely to permit the delivery of highly conformal, computer-modulated beams that produce superior dose distributions with a very sharp dose fall off at the edge of the tumor in comparison with photon therapy plans of the past. Additionally, image-guided radiotherapy (IGRT) technology allows for a reduction of the margin of normal tissues included in the treatment of malignant tissues. Sahgal et al published the outcomes of image-guided IMRT (IG-IMRT) for skull base chordomas and chondrosarcomas. ¹² They have defined the gross tumor volume (GTV) as all gross disease visible on imaging and the postoperative surgical bed. To this, they have added a 0.5-cm margin beyond the GTV and areas at risk of microscopic extension to create a CTV (clinical target volume). The planning target volume (PTV) typically consisted of a 0.2- to 0.3-cm margin beyond the CTV, depending on the immobilization technique. With respect to the PTV, their aim was to achieve 90 to 95% coverage with a median total dose of 76 Gy (range: 60–78 Gy). The authors have reported very favorable preliminary rates of LC for both chondrosarcoma and chordomas following high-dose photon linear accelerator IG-IMRT, a crude LC rate of 67% for chordoma patients and 88% for chondrosarcoma patients. The 5-year actuarial LC and OS rates were reported to be 65.3% and 88.1%, respectively, for chordoma patients. ¹² The authors have concluded that their results are consistent with expected outcomes following proton therapy. ³⁰ Radiation induces hearing loss, hypopituitarism, vestibular nerve injury, and cranial nerve IV injury; resulting diplopia and secondary malignancy were reported in 6 of 46 patients. ¹²

Given that technical advancements have improved the treatment outcomes, a 2012 review by Gil et al evaluated the trends in survival in patients with anterior skull base cancers and showed that the patients treated after the year 1996 have a statistically improved survival (66%) when compared with patients treated before the year 1996 (55%) (p = 0.02). ³¹ Interestingly, the presence of adjuvant RT was a prognostic factor for improving OS (p = 0.02), but it was not a significant predictor of disease-specific survival. ³¹ Another comprehensive review by Jian et al analyzed 560 cranial chordoma patients. ⁷ A total of 210 patients were treated with surgery alone, 190 patients had surgery in addition to adjuvant radiation treatment, and 56 patients underwent definitive radiotherapy as their sole therapy. The difference in 5-year survival rates between the surgery-only group versus the surgery plus radiation group was not significant (54% vs. 56%; p = 0.8). No difference was reported between conventional surgical, SRS, and fractionated SRS techniques. The only factor affecting the OS was young age (< 5 years old), and the authors concluded that adjuvant RT may not be associated with an improved survival for this group of patients, ⁷ which has relevance to the decision on whether or not to recommend adjuvant radiotherapy in pediatric patients given the secondary cancer risk following radiotherapy.

Two large meta-analyses evaluating the outcomes of RT of skull base chordomas were published in 2011 and 2014. ⁶^,¹¹ Both analyses included patients treated with a variety of radiotherapy techniques. The objective of the first meta-analysis was to measure the relationship between complete resection and type of adjuvant RT and 5-year PFS and OS. The OS rate was 70% at 5 years and 63% at 10 years. Adjuvant radiation therapy was delivered using fractionated photon- and proton-based RT, Gamma Knife (GK), and carbon ion RT, and in several studies, more than one type of adjuvant radiation therapy was used. No significant difference in 5-year PFS or OS rate was found for different RT techniques. Among 517 patients, no significant difference in 5-year OS rate was observed for different types of adjuvant radiation therapy. Mean 5-year PFS rate was lower for GK surgery than that for carbon ion RT (p = 0.04); otherwise, there was no significant difference in 5-year PFS in comparisons between proton beam RT, carbon ion RT, and fractionated RT. Only 58 patients (22 patients with adjuvant RT and 36 patients without) were sampled for comparison of PFS and OS with and without adjuvant RT after complete resection, and no significant difference in 5-year PFS and OS rates was identified whether or not adjunctive radiation therapy was given. ⁶ The striking results of the meta-analysis showed that 5-year risk of recurrence was 3.83-fold higher in patients with incomplete resection than in those with complete resection (95% confidence interval [CI]: 1.62–9.00). Nevertheless, as in the majority of the studies, RT was used primarily in the subtotal resection group, which historically has had a poorer prognosis. The limitations of the meta-analysis or reviews are primarily due to inherent heterogeneity across individual studies and particularly for skull base chordomas, many of the included studies have relatively small numbers of patients with short follow-up. Furthermore, the details of RT such as dose, tumor volume, proximity to critical structures, treatment toxicity, surgical complications, functional outcomes, and quality of life were not accounted for in this meta-analysis. ⁶

A second meta-analysis was published in 2014 by Amit and colleagues from Israel. They assessed the efficacy of different surgical approaches and adjuvant RT modalities in the management of the skull base chordomas. ¹¹ The entire cohort consisted of 467 patients. As the other meta-analysis, the RT group patients were treated with conventional RT, SRT, or proton beam with combined RT techniques. The 5-year OS rate of patients treated with surgery followed by adjuvant radiotherapy was 87%, compared with 69% of those treated by surgery alone (p = 0.12). The 5-year OS and DFS rates of patients treated with adjuvant radiotherapy, including carbon ion RT, proton beam RT, and SRT, were not different statistically. In patients who had undergone total resection with or without radiotherapy, the 5-year OS and DFS rates did not differ significantly. However, for patients who had undergone partial resection with adjuvant radiotherapy, the 5-year OS and the DFS rates were significantly higher compared with the patients without adjuvant treatment (p < 0.001 and p = 0.01, respectively). ¹¹ Even though the heterogeneous and incomplete data limit the value of the meta-analysis of Amit et al, their findings, which have shown that adjuvant radiotherapy improves survival of patients undergoing partial resection, are important, and RT should be considered for appropriate patients.

Given the location of these tumors and proximity of radiosensitive structures, all radiotherapy techniques have the difficulty in maintaining conformality while sparing critical normal tissues. Consequently, high curative doses of RT may cause long-term toxicity, especially for the neurovascular structures. The recommended dose limits of critical organs for conventional and SRT are given in ▶ Table 25.2. In the literature, there are little data regarding the long-term toxicity of RT. Most meta-analyses and reviews focused primarily on LC and survival rates and not toxicity. ⁶^,¹¹ It is also hard to distinguish the source of toxicity as arising from surgery or RT. However, exceeding dose limits or the inherent radiosensitivity of the patient may result in serious complications such as visual deficits, cranial nerve dysfunction, brainstem injury, pituitary endocrinopathy, and temporal lobe necrosis. To our knowledge, the only trial that has focused on RT complications was published by Hauptman et al. ³² Their retrospective study consists of 13 cases of chordomas and 2 cases of chondroid chondrosarcomas of the skull base treated with linear accelerator–fractionated (28–42 fractions) SRT (frSRT; n = 10) or SRS (n = 5). After a median follow-up of 4.5 years, long-term complications were noted in five patients. Within this group, one frSRT patient developed endocrinopathy, two patients (one treated with SRS and the other with frSRT) developed cranial neuropathy, and one SRS patient developed visual deficits. Additionally, one patient who received both SRS and frSRT within 2 years for recurrence experienced transient medial temporal lobe radiation changes that resolved. ³² Although this paper provides some useful information regarding treatment toxicity, the study did not include patients treated with hypofractionated SRT. ³³