10 Management of Benign Skull Base Tumors in Neuro-oncology: Systemic Cytotoxic and Targeted Therapies



Sophie Tailibert and Marc C. Chamberlain


Summary


The treatment of benign tumors of the skull base relies mainly on a surgical approach whose purpose is to obtain a maximal safe resection as well as, if possible, a complete resection. However, the proximity of functional neurovascular structures such as cranial nerves and the brainstem to skull-based tumors results in frequent inability to obtain a complete resection, leading to frequent local recurrences, significant neurologic disabilities, and decreased survival. Most benign skull-based tumors display a slow growth rate, which often results in intrinsic radioresistance and chemoresistance. Though biologically defined as “benign,” these tumors are ultimately complicated by clinical evolution that is far from “benign,” being mainly caused by the therapeutic limitations and high risk of recurrence already noted. Nevertheless, considerable progress has been achieved recently in understanding the underlying genetic, epigenetic, molecular, and cellular signaling mechanisms of these tumors, opening a pathway to potentially new therapeutic strategies. Nonetheless, the clinical development of targeted drugs for the different types of benign tumors of the skull base is at an early stage. Limitations such as the rarity of several of these tumors, the lack of preclinical animal model, and the lack of understanding of their clinical relevant biology remain to be addressed. The current review focuses on those benign skull-based tumors that exhibit the genetic and molecular understandings most promising for targeted therapies. The review provides an update on the current management, pathogenesis, and recent therapeutic advances of these tumors.




10 Management of Benign Skull Base Tumors in Neuro-oncology: Systemic Cytotoxic and Targeted Therapies



10.1 Introduction


Tumors of the skull base are categorized and defined as “benign” according to their pathologic and biologic behavior. Nevertheless, their anatomical location in proximity to critical neurovascular structures such as the cerebral vasculature, cranial nerves, and brainstem generally does not safely permit optimal surgical treatment. The absence of complete resection leads to high local recurrence rates, significant neurologic disabilities, compromised quality of life, and decreased overall survival. A comprehensive approach is thus necessary for management of benign skull base tumors, which often involves multiple therapeutic modalities. The slow growth rate that characterizes “benign tumors” is in large part responsible for their commonly seen intrinsic radioresistance and chemoresistance. This chapter provides an update on the pathogenesis and current neuro-oncology management of these tumors, including the related therapeutic advances from 2018 and earlier. Some benign tumors of the skull base, such as chondroma, paraganglioma, and pituitary adenoma, are not discussed, whether because they are discussed elsewhere or because there are insufficient data regarding active systemic cytotoxic or targeted treatment. The current review focuses on the benign skull-based tumors that are candidates for new systemic therapeutic approaches based on recently identified genetic and molecular targets. These tumors mainly comprise World Health Organization (WHO) grade I meningioma, chordoma, craniopharyngioma, and giant cell tumor of the bone.



10.2 WHO Grade I Meningioma



10.2.1 Introduction and Treatment


Benign or WHO grade I meningioma, when located in the skull base area, remains a challenge for complete and safe surgical resection; consequently, residual tumor remains postoperatively and there is often a significant risk of neurologic morbidity due to proximity to adjacent eloquent brain. The 2007 WHO classification, updated in 2016, did not change the grading of meningioma except to include existence of brain invasion with a mitotic count ≥ 4 that now categorizes WHO grade II or atypical meningioma.1 It was previously established that the prognosis and clinical evolution of grade I meningioma with brain invasion was like those characterized as grade II lesions.2


Skull base WHO grade I meningioma has a typically indolent onset, with minimal symptoms until tumors become large enough to compromise adjacent functional neurovascular structures. The slow onset results in a delay in diagnosis and presentation, often at an advanced stage.3 Depending on the location of the WHO grade I meningioma in the skull base, presenting deficits are usually referable to proximate cranial nerves.


Meningiomas of the skull base originate from the olfactory groove, sphenoid wing, suprasellar and parasellar regions, petroclival, cerebellopontine angle, and foramen magnum.4 Imaging is often characteristic, although radiographic findings may vary among patients. On CT, meningiomas are hyperdense and sometimes calcified. On MRI, meningiomas often are isointense with gray matter on all sequences, and contrast enhancement is intense and homogenous, although necrotic areas may be observed in large lesions.4


Meningiomas are highly vascularized tumors that have high relative cerebral blood volume (rCBV) values on MR perfusion. MR perfusion may assist in the differential diagnosis of skull-based lesions.5 ,​ 6 ,​ 7 Diffusion tensor imaging provides additional information about the tumor’s cellular content and may help differentiate among atypical, fibroblastic, and benign meningiomas.4 ,​ 8 ,​ 9 High alanine and low N-acetylaspartate peaks are observed on MR spectroscopy (MRS), but bone artifacts in the skull base often limit the use of MRS except for large lesions.10


The initial management of patients who have WHO grade I meningioma of the skull base usually consists of surgery, surgery plus radiation therapy (RT), RT alone, or observation.3 ,​ 11 ,​ 12 Observation is an option in asymptomatic and small lesions, those defined by National Comprehensive Cancer Network (NCCN) guidelines as having a diameter of less than 3 cm, but an observation-only approach is predicated in part on the tumor’s anatomical proximity to neurovascular structures. Surgery is recommended for symptomatic lesions whenever possible if it can be performed without sustaining functional morbidity. In instances of incomplete resection, postsurgical RT is often administered.3 ,​ 11 ,​ 12 Differing methods of RT may be used, including conformal external fractionated RT, stereotactic radiosurgery (SRS), stereotactic radiotherapy, intensity-modulated RT (IMRT), and volumetric-modulated arc RT.13 ,​ 14 ,​ 15 ,​ 16 ,​ 17 As of 2019, systemic treatment does not have an indication in the frontline treatment of meningioma. Particularly with skull-based WHO grade I meningioma, recurrence is frequent, and surgery or RT are most often deployed again as first salvage therapy. A variety of systemic therapies have been evaluated in surgical and RT refractory meningiomas, including chemotherapy, hormonal therapy, immunotherapy, and targeted therapies. These therapies are discussed hereafter. Regardless, most systemic treatments show modest activity. Recent identification of several molecular and druggable targets may allow development of more active systemic therapies in the future.


In all grades of meningioma, the prognosis is related to extent of resection (Simpson grade), patient characteristics such as gender (unfavorable for male), and tumor pathologic profile, such as WHO grade, Ki67/MIB-1 index, percentage of tumor cells in the S phase, p53 overexpression, telomerase activity, and telomerase reverse transcriptase (TERT) promoter mutation status (unfavorable when mutated).18 ,​ 19 ,​ 20



10.2.2 Pathogenesis, Genetics, Molecular, and Cellular Biology


Nearly 60% of WHO grade I meningiomas exhibit mutations of the NF2 gene on chromosome 22 (location: q12.2).21 ,​ 22 ,​ 23 ,​ 24 The NF2 gene codes for the protein merlin, which belongs to the 4.1 family of structural proteins, and behaves as a tumor suppressor. Merlin is named for its similarity to moesin, ezrin and radixin-like protein. Variable levels of merlin loss are observed in different subtypes of grade I meningiomas (lower loss in the meningothelial type than in fibrous and other).25 Merlin localizes to the cell membrane and is involved in controlling cell–cell contact and cell motility.


The following proteins show some degree of interaction with merlin: CD44 and β1-integrin, βII-spectrin, paxillin, actin, and syntenin. These proteins are involved in cytoskeleton dynamics, sodium–hydrogen exchange regulatory factor, hepatocyte growth factor–receptor regulation, and endocytosis.26 Merlin downregulates the Yes-associated protein (YAP), a protein that controls cellular proliferation. NF2 mutations result in increased proliferation of arachnoid cap cells and meningioma development.27


Besides merlin loss, other protein 4.1 family members are also downregulated in grade I meningiomas. Up to 50% of WHO grade I meningiomas display a dual loss of protein 4.1B (DAL-1) and merlin expression.28 In up to 85% of sporadic meningioma, loss of TSLC-1 gene (tumor suppressor gene for lung cancer-1) is observed, triggering dysfunctional interactions between its protein and DAL-1.29 In normal conditions, the protein of the TSLC-1 gene (tumor suppressor gene for lung cancer-1) interacts with DAL-1, which is present in the plasma membrane, close to cell–cell contact points. Both proteins are spectrin–actin-binding proteins, and when they interact by direct binding, they influence cell motility through actin reorganization.30


The therapeutic targeting of TSLC-1 and DAL-1 might be part of a future strategy of anti-invasion and antimigration in meningiomas that exhibit intact expression of these two proteins.26 ,​ 31 AKT1 mutation is found almost exclusively in skull-based grade I meningiomas and has become another potential therapeutic target since the development of AKT inhibitors.32 ,​ 33


Multiple growth factors and their cognate receptors are observed in meningiomas and likely determine, at least in part, progression through overexpression and dysregulation. These include platelet-derived growth factor BB (PDGF-BB), epidermal growth factor receptor (EGFR), vascular endothelial growth factor (VEGF), transforming growth factor alpha and beta (TGF-alpha and -beta), SDF1 and receptor CXCR, bone morphogenic protein, insulin growth factor I and II (IGF), HER2, somatostatin, fibroblast growth factor, placental growth factor, and cyclo-oxygenase-2 Cox-2 (34–36). EGFR and platelet-derived growth factor receptor beta are both overexpressed in WHO grade I meningiomas.34


Meningiomas show alterations of chromosomes 1, 9, 10, 14, and 22, and their proliferation has been associated with mutations of genes in chromosomes 2, 4, 7, 8, 11, 12, and 16, leading to the activation of multiple signaling pathways. The most prominent ones are summarized hereafter.26 ,​ 35 The Wnt/β catenin pathway is involved in grade I meningiomas, in which deletions of the tumor suppressor APC gene are observed.36 ,​ 37 A third meningiomas of all grades display losses of E-cadherin function, leading to increased invasiveness and higher recurrence rates.37 The Notch pathway is an intracellular signaling pathway mediated by the transmembrane proteins Notch 1–4.38 ,​ 39 Hes-1 is a protein that is a target of the Notch signaling pathway, and its expression is found in all grades of meningioma and related to an overexpression of the Jagged ligand, Notch 1, and Notch 2. TLE2/TLE3 are enhancers that increase the activity of the Split family (co-repressors) and that modulate Hes-1, a NOTCH signaling protein but appear to be upregulated in high-grade meningiomas only.39


The hedgehog (Hh)/patch (PTCH) pathway is involved in cell growth and is regulated by the (Smoothened) SMO gene. When Hh binds to PTCH, it activates SMO, leading to the activation of GLI transcription factors such as GLI1 (growth activator), GLI2 (growth activator), and GLI3 (growth repressor). The Hh pathway is also involved in angiogenesis, matrix remodeling, and stem cell homeostasis.26 A recent large series demonstrated that 28% of SMO mutations (L412F or W535L) and 15% of AKT1E17K mutations are seen in meningiomas of the olfactory groove.40 Among WHO grade I meningiomas, the group having SMO mutations had a significantly poorer prognosis. Meningiomas in the SMO-mutant group had an overall 36% recurrence rate, significantly higher than in the AKT1-mutant group (16%) and the “SMO and AKT1 wildtype” group (11%). All late recurrences, defined as 5 years after diagnosis, occurred in the SMO-mutant group. The authors suggest that the high frequency of SMO mutations in meningiomas arising from the anterior and medial skull base might be explained by the central role of the sonic hedgehog pathway in craniofacial development during embryogenesis. These results support the systematic determination of SMO mutations, both for prognosis and for potential inclusion of patients in future trials targeting this pathway using Hh inhibitors. The p53/pRB pathway controls the G1 to S phase cell cycle transition through the tumor suppressor gene pRB and is involved in meningioma progression and development of the anaplastic grade.26


The PI3K/AKT and mitogen-activated protein kinase (MAPK) are two pathways involved in grade II/III meningiomas only, in which high levels of phosphorylated Akt (PI3K/Akt) are found.21 ,​ 41 ,​ 42 ,​ 43 ,​ 44 Decreased levels of MAPK lead to higher recurrence rate. Other pathways, such as the phospholipase A2-arachadonic acid cyclo-oxygenase pathway, the PLC-gamma1-PKC pathway, and the transforming growth factor-beta (TGF-b)-SMAD signaling pathway, have all been related to meningioma pathogenesis.45 ,​ 46 ,​ 47 The levels of expression of cox-2 and of tumor necrosis factor–related apoptosis-inducing ligand receptor 4 (TRAIL-R4) seem correlated to the grade of meningioma.48 KIT expression is upregulated in 20.6% of meningiomas and is another potential druggable target.18 TERT promoter mutations, which are correlated to a decreased time to progression and a higher risk of recurrence, are present in only 1.7% of WHO grade I meningiomas.47



10.2.3 Systemic Therapies


An accurate assessment of the therapeutic efficacy in meningiomas has not been possible until recently, mainly because of the heterogeneity of response criteria used in different series. The Response Assessment in Neuro-Oncology (RANO) working group recently proposed use of the 6-month progression-free survival (PFS-6) as the standard in clinical trials in recurrent meningiomas.49 The weighted PFS-6 of WHO grade I meningioma determined by the RANO authors was 29% (95% confidence interval: 20.3–37.7%). The authors concluded that a new therapeutic compound should demonstrate a PFS-6 > 50% in WHO grade I meningioma if it is to be considered a therapy of interest.49 Nevertheless, there is still a lack of standardization regarding the optimal response criteria. Due to the complex shape of meningioma, the potential superiority of volumetric measures as compared with 2D McDonald criteria has been highlighted.50



Chemotherapy

Hydroxyurea (HU), temozolomide, and irinotecan have been evaluated in phase II trials, retrospective cohorts, and case studies in patients having recurrent WHO grade I meningiomas.20 ,​ 51 ,​ 52 ,​ 53 ,​ 54 ,​ 55 ,​ 56 ,​ 57 ,​ 58 ,​ 59 ,​ 60 ,​ 61 ,​ 62 The reported median PFS ranged from 4 to 90 months.54 ,​ 55 ,​ 58 ,​ 59 PFS-6 was retrospectively assessed in one series of 60 patients and was only 10%.59 Using a combination of HU and fractionated conformal RT in 13 patients who had recurrent or progressive lesions, median PFS-12 was 84%, but the specific contribution of HU cannot be disentangled from the effects of radiotherapy.56 A phase II trial did not show any efficacy of temozolomide in 16 patients with a PFS-6 of 0%.60 Similarly, irinotecan did not display activity in a series of 16 patients with a reported PFS-6 of 6%.61 Trabectidin, a novel marine-derived antineoplastic agent, has shown a significant cytotoxic activity in grade II/III meningioma in preclinical studies.63 Synergy with HU, cisplatin, and doxorubicin was observed as well. These results led to the development of a phase II clinical trial in recurrent high-grade meningioma, but WHO grade I meningioma are currently excluded.



Hormonal Therapy

Thirty percent and 70% of meningiomas express receptors to estrogen and progesterone, respectively.45 ,​ 64 ,​ 65 ,​ 66 ,​ 67 Although PFS-6 is lacking in most published studies, the reported data fail to show any compelling efficacy of any antiestrogen or progesterone agents in recurrent and surgery-naive meningioma WHO grade I. Tamoxifen has been assessed in recurrent meningioma, including a single phase II trial, without significant effect.45 ,​ 64 ,​ 65 ,​ 66 ,​ 67 ,​ 68 No radiographic responses were observed, although prolonged stabilization was seen in some studies.69 ,​ 70 The progesterone-targeting agents megestrol acetate, medroxy-progeterone acetate, and mifepristone (RU-486) have been evaluated in grade I meningioma at recurrence or in meningiomas for which there was no prior surgery.71 ,​ 72 ,​ 73 ,​ 74 ,​ 75 Like antiestrogen-targeting agents, these agents achieved, at best, long-lasting disease stabilizations.71 ,​ 74 Mifepristone was reported to have resulted in limited prolonged volumetric reductions in a few cases but had no significant impact when assessed in a randomized double-blind phase III study in grade I nonsurgical meningioma.75



Somatostatin Analogs

Ninety percent of meningiomas express somatostatin receptors, mostly of the sst2A subtype76 Several retrospective cohorts and phase II studies have assessed the effect of somatostatin analogs such as somatostatin, pasireotide, octreotide (Novartis), Sandostatin LAR (Novartis), pasireotide (Novartis), 90Y-DOTATOC (Abbott), 177Lu-DOTATOC (Novartis) on progressive or recurrent meningiomas of all grades.50 ,​ 76 ,​ 77 ,​ 78 ,​ 79 ,​ 80 Some studies report global PFS-6 rates for tumors of all grades between 32 and 44%.50 ,​ 76 A PFS-6 of 50% was observed in recurrent or progressive grade I meningiomas with pasireotide LAR, albeit without any objective radiographic response.50 The activity of these agents seems limited, notwithstanding a subpopulation of responders’ appearing to benefit from somatostatin-based therapy. Factors such as differing expression of somatostatin receptors (sst2 and 3) and level of octreotide uptake might be predictors of response and longer PFS and survival.50 ,​ 76 ,​ 77 ,​ 78 ,​ 80



Interferon Alpha

An encouraging PFS-6 rate of 54% was obtained in a phase II trial of patients who had refractory WHO grade I meningioma treated using interferon-alpha, an active agent in vitro.45 ,​ 67 ,​ 81 ,​ 82 ,​ 83 Although no objective radiological response was reported, 74% of patients were stabilized with a median PFS of 7 months and a median overall survival of 8 months (range 3–28 mo).81



Molecular Targeted Therapies

The lack of current effective treatment in recurrent or progressive WHO grade I meningioma is related in part to an insufficient understanding of the molecular pathogenesis of these tumors.26 Despite the identified overexpression of multiple growth factors, including platelet-derived growth factor (PDGF), Epidermal growth factor (EGF), vascular endothelial growth factor (VEGF), insulin-like growth factor (IGF), transforming growth factor-beta (TGF-β), and their receptor tyrosine kinases, and the dysregulation of their downstream signaling pathways, including the RAS/MAPK, PI3K/Akt, PLC-γ1-PKC, and TGF-β-SMAD pathways, their respective importance is poorly understood, and druggable targets need to be identified.26 Moreover, other approaches involving insulin-like growth factor receptor 2 (IGFR-2), histone deacetylase, NFκB, HSP90, JAK/STAT, immune checkpoint inhibitor, telomerase inhibitors, BRAF inhibitors, Src kinase, focal adhesion kinase (FAK), and hypoxia-inducible factor 1α might lead to potential new treatments.26 ,​ 84 ,​ 85


Sixty percent of meningiomas express EGFR.86 ,​ 87 A phase II trial evaluating either gefitinib or erlotinib showed poor efficacy in recurrent meningioma WHO grade I patients who had a PFS-6 rate of 25%, 13 months of median overall survival, and 9 weeks median PFS.88 Most meningiomas of all grades express platelet-derived growth factor (PDGF)-b receptor and manifest an increased level of PDGF.89 ,​ 90 In vitro, meningioma cell proliferation is inhibited by anti-PDGF antibodies.91 ,​ 92 Imatinib, an inhibitor of PDGF receptor (PDGFR), was given as a single agent in a phase II trial.93 It showed limited efficacy, with a PFS-6 of 45% in recurrent and progressive WHO grade I meningioma. When combined with HU in the same setting, PFS-6 reached 87.5%.94 In a small randomized study of recurrent meningioma (all grades) that compared HU with HU + imatinib, PFS-9 (9-month PFS) was 0% and 75%, respectively.95


Meningiomas of all grades are highly vascularized and express VEGF and VEGF-R at variable levels related to grade (increased levels with higher grade).46 ,​ 96 ,​ 97 ,​ 98 ,​ 99 ,​ 100 This pathway is involved in tumor growth vis-à-vis angiogenesis as well as in induction of peritumoral edema.45 ,​ 66 ,​ 83 ,​ 101 Sunitinib (SU011248), a small molecule inhibitor that targets VEGF, PDGF, c-KIT, FLT, RET, macrophage colony–stimulating factor (CSF-1R), and valatinib (PTK787/ZK22584), a VEGFR 1–3 inhibitor, was ineffective in recurrent grade I meningiomas.102 Bevacizumab, a monoclonal antibody targeting VEGF, has been evaluated mostly in the retrospective setting and is often coadministered with either temozolomide or etoposide.99 ,​ 100 ,​ 103 ,​ 104 In a phase II trial, bevacizumab was given to patients who had recurrent meningioma who were not otherwise considered candidates for local treatment. Results were encouraging in meningioma WHO grade I: PFS-6 of 87%, median PFS of 22.5 months, median overall survival of 35.6 months, and 100% tumor stabilization.105 Additionally, bevacizumab appears to have a mitigating effect on peritumoral edema and meningioma growth rate.106


Preclinical data show inhibiting activity of temsirolimus, an mTOR inhibitor, in meningioma in which the pathway is activated. Furthermore, a synergistic effect was seen when octreotide was added to everolimus.107 ,​ 108 Consequently, this combination is being pursued in all grades of recurrent meningioma patients in a clinical trial with PFS-6 as the primary objective (CEVOREM/NCT02333565).


The therapeutic potential of TRAIL agonists has been evaluated in meningioma cell lines.108 TRAIL-induced apoptosis was seen in 29.7% of cell lines, but native TRAIL receptor expression was not predictive of TRAIL agonist sensitivity. The coadministration of bortezomib, which induces NOXA expression and downregulation of c-FLIP, may increase sensitivity to TRAIL-induced apoptosis.109 FAK inhibitors that target NF2 and merlin loss seen in up to 60% of sporadic meningiomas represent another trial in progress in patients who have recurrent meningiomas.110 ,​ 111


The FAK inhibitor GSK2256098 and vismodegib, a hedgehog pathway inhibitor, are being studied in progressive meningiomas with SMO/AKT/NF2 mutations. A histone deacetylase inhibitor and its impact on suppression of p-AKT are also being explored.18 ,​ 112


Some of the numerous potential targets may preferentially be expressed in high-grade rather than low-grade meningiomas. For example, programmed death-1 ligand (PD-L1) is frequently expressed in higher-grade tumors, indicating that it may play a biological role in the aggressive phenotype of meningiomas and may be targetable by PD-1 and PDL-1.113


Other approaches, such as gene silencing, supplementation of lost genetic factors, regulation of YAP levels in NF2 mutated tumors, and addressing of deficient intrinsic and extrinsic pathways, may improve management of recurrent and progressive meningiomas.26


Insufficient understandings of the evolution of grade I meningiomas and the lack of a preclinical reliable animal model are significant barriers to the development of novel active therapies. The RANO criteria will help standardize results of future trials in recurrent meningioma.



10.3 Chordomas



10.3.1 Introduction to Chordomas and Treatment Overview


Chordomas of the skull base are rare, slow-growing, locally aggressive, and destructive bony tumors derived from embryonic remnants of the notochord. Chordomas located in the skull base represent a third of all chordomas and are most frequently located in the spheno-occipital region.114 Headaches and intermittent diplopia are the predominant clinical features at onset, followed by additional cranial nerves palsies that differ according to the topography of the chordoma (e.g., supra- or infraclival). Delays in diagnosis are frequently caused by nonspecific symptoms such as nasal congestion or dysphagia that are related to cranial nerve involvement but not always initially identified as being so.


CT imaging is the best technique for assessing bone destruction; it may show characteristic ring-forming calcifications.115 MRI better shows the extent of tumor, including soft tissue components and dural extension.116 Chordomas on MRI are hypo- or isointense on T1-weighted and hyperintense on T2-weighted MRI sequences. Contrast enhancement is heterogeneous, with a lobulated “honeycomb” appearance, reflecting the liquid and gelatinous mucoid components and necrotic-hemorrhagic areas.117 A fusion of CT and MRI ideally provides the most accurate assessment of the anatomical relation between soft tissue tumor components surrounding normal soft tissue and bone.


Though chordomas are classified as benign tumors, their prognosis is more akin to that of a malignant lesion, with high local recurrence rates, including systemic metastases (40% incidence overall, 12.5% from skull base chordoma), largely responsible for tumor-related lethality.118 ,​ 119 ,​ 120 ,​ 121 ,​ 122 The median survival rate is approximately 6.29 years with 5-year OS and PFS rates of 78.4 and 50.8%, respectively.114 ,​ 121 ,​ 123


Histologically, chordomas, like the notochord, are composed of large vacuolated physaliphorous cells surrounded by an extracellular myxoid matrix.118 ,​ 124 Chordomas comprise three subgroups: conventional chordomas, chondroid chordomas, and dedifferentiated chordomas (i.e., those with sarcomatous transformation). Conventional chordomas, which comprise chordomatous features only, are most frequent, followed by chondroid chordomas, which comprise additional chondromatous features. Both entities share the same prognosis, unlike dedifferentiated or sarcomatous chordomas, which behave in a more malignant manner.114 ,​ 125 Most chordomas stain positive for cytokeratin, and 80% are epithelial membrane antigen–positive. A limited number in addition are S100 protein–positive.126 ,​ 127 ,​ 128 Routine immunohistochemical staining for brachyury, a key transcription factor in notochord development that is overexpressed by chordomas, is a sensitive and specific diagnostic tool that also rules out cartilaginous tumors.129 ,​ 130 ,​ 131 ,​ 132 ,​ 133


The treatment of chordoma is continually evolving owing to continued progress in understanding underlying biology. Chordomas are radioresistant to standard radiation protocols and chemoresistant to classic cytotoxic therapies, in part as a result of their slow-growing kinetics. Standard radiation to the skull base is also limited by the radiosensitivity of the brainstem and other proximate intracranial neurovascular structures that increase the risk of radiation-related injury.134 ,​ 135


For localized disease, consensus guidelines published in 2015 recommend a complete surgical resection with tumor-free margins.136 Classic open skull base approaches and less invasive endoscopic approaches are used to maximize the resection.121 ,​ 122 ,​ 137 ,​ 138 ,​ 139 ,​ 140 ,​ 141 ,​ 142 When residual disease is remaining, often having a relationship to adjacent neurovascular structures, radiotherapy is recommended.143 ,​ 144 ,​ 145 ,​ 146 Definitive (proton beam) RT is the preferred alternative in instances in which resection is not feasible. Recent advances in heavy particle RT, such as proton and carbon ion beam, are promising in the adjuvant setting of skull base chordomas.147 ,​ 148 ,​ 149 Adjuvant proton therapy, when added to maximal safe resection in skull base chordomas, led to 5-year local control rates of 75.8% and improved survival.147 ,​ 148 Adjuvant carbon ion radiation provided similar local 5-year control rates of 70.0%.149 Image-guided IMRT (IG-IMRT) is also under evaluation for skull base chordoma.150 Cytotoxic chemotherapy or other systemic therapies have no proven efficacy and are thus not recommended in any line of treatment according to current guidelines.136 ,​ 151 Nevertheless, several molecular therapies targeting essential pathways in chordomas are currently under evaluation and are discussed hereafter.



10.3.2 Pathogenesis, Genetics, and Molecular and Cellular Pathways


Advances in the molecular understanding of chordomas have led to the identification of promising prognostic markers and targetable pathways. These include genetic and epigenetic alterations that involve brachyury, downstream pathways, and receptor tyrosine kinases (RTKs). The brachyury or T gene located on chromosome 6q27 encodes a transcription factor essential for the development of the notochord.152 Normally silenced, brachyury is reexpressed at high levels in chordoma cells. Several lines of evidence also suggest a causative role of brachyury overexpression in chordoma formation. Further brachyury might be involved in the metastatic process by facilitating the epithelial–mesenchymal transition.153 ,​ 154 ,​ 155 ,​ 156 As already mentioned, the expression of brachyury is very helpful for pathologic diagnosis, but its level does not seem correlated with prognosis.157


Other genetic alterations in chordomas include loss of p16, PTEN, CDKN2a/CDKN2b, and PDCD4.158 ,​ 159 ,​ 160 ,​ 161 ,​ 162 ,​ 163 ,​ 164 ,​ 165 ,​ 166 ,​ 167 ,​ 168 Losses on chromosome 1p and the FHIT gene and gains on 1q and 2p have been detected in skull base chordomas.152 ,​ 160 Loss of chromosome 1p36, 9p loss of heterozygosity, and an elevated Ki67 proliferative index are associated with shorter OS in skull base chordomas.152 ,​ 161 Additionally, loss of chromosome 1q, gain of 2p, and aberrant brachyury copy number are associated with recurrence.


Many alterations in the signaling pathways and growth factors are involved in chordoma biology as well. Both the PI3K/AKT/TSC1/TSC2/mTOR and STAT3 pathways are activated.162 STAT3 level of expression has reported prognostic value. Changes in EGFR signaling, activation of IGF1R, and loss of MTAP have also been identified.118 ,​ 163 ,​ 164


Protein tyrosine kinases (PTKs) mediate phosphorylation of selected tyrosine residues, resulting in functional activation of many proteins and thus playing a crucial role in cancer development. RTKs are specialized transmembrane PTKs that mediate signaling via sampling of the external environment. RTKs are composed of extracellular domains that bind cognate environmental ligands and of an intracellular domain that mediates the signaling event via dimerization and binding to other signaling molecules. EGFR overexpression is observed in 69 to 79.6% of chordomas and is related to tumor aggressiveness.144 ,​ 163 ,​ 165 ,​ 166 ,​ 167 PDGFR expression is also observed in chordoma. Although PDGFR promotes chordoma cell proliferation through activation of the PI3K/AKT, RAS/ERK, and STAT pathways, PDGFRβ is observed in the stromal component of the tumor and is likely involved in microenvironmental regulation.144 ,​ 165 ,​ 168 ,​ 169 ,​ 170 ,​ 171


Increased expression of angiogenesis-related factors such as VEGF, hypoxia-inducible factor-1α (HIF-1α) and matrix metalloproteinase (MMP)-2 and -9 has been identified in chordoma tissue and cell lines.172 ,​ 173 MMP-9 expression is correlated with higher rates of local recurrence and poor prognosis.173


Epigenetic alterations, especially modifications of the expression of microRNA (miRNA) have been identified in chordomas, and may be therapeutically targetable as illustrated in vitro.174 ,​ 175 ,​ 176 ,​ 177 ,​ 178 ,​ 179 For example, the administration of nonselective histone deacetylase inhibitors significantly triggers apoptosis in chordomas cells.174 Similarly, the blockade of the constitutive downregulation of microRNA-1 in chordomas inhibits cellular proliferation, migratory and invasive properties, and the expression of the oncoprotein SLUG as well.176 ,​ 177 The restoration of two constitutively downregulated miRNAs targeting EGFR, MET, and Bcl-xL block cell proliferation.175



10.3.3 Systemic Therapies


As already mentioned, cytotoxic chemotherapy has not proven efficacious and thus is not recommended in any line of treatment of chordomas. Brachyury based on its determinant role in notochord development and selective reexpression in chordoma makes this an attractive and potential target.145 This is illustrated in vitro by the differentiation and senescence of chordoma cells, triggered by silencing the brachyury gene.180 Currently no treatment directly targets brachyury. An alternative approach would be instead to target downstream or interacting signaling pathways such as FGFR/MEK/ERK signal-transduction and EGFR signaling pathways.181 ,​ 182 Immunotherapy may be promising, as illustrated by the partial responses observed in advanced chordoma patients treated with an immune-stimulating therapeutic cancer vaccine designed to elicit brachyury-specific T-cell responses.183


Multiple PTKs are posited to be involved in chordoma development and progression, some of which are functionally redundant. Therapeutic strategies comprise simultaneous inhibition of different PTKs or blocking of downstream signaling pathways. In chordomas, the PI3K/AKT/mTOR pathway is activated and PTEN is inhibited.162 ,​ 168 ,​ 169 ,​ 184 It was hypothesized that therapeutic activity might be seen with inhibition of signaling mediated by EGFR and PDGFR, both of which converge on the PI3K/AKT/mTOR pathway. Some data suggest a potential therapeutic role for combined AKT and mTOR inhibitors, such as rapamycin.162


Preclinical use of the combination of a mTORC1 (rapamycin), mTOR (MLN0128), and PI3K/AKT/mTOR (PI-103) inhibitors presaged clinical activity seen with both mTOR and PDGFR as single agents in recurrent chordoma. The combined clinical efficacy of the combination of an mTOR inhibitor and imatinib in imatinib-resistant patients who had chordomas illustrates another possible strategy.184 ,​ 185 ,​ 186 ,​ 187 Improved activity-based molecular profiling of a patient’s tumor and identification of specific aberrant signaling pathways is likely the most effective approach, as illustrated by the spectacular efficacy of rapamycin on tumors having aberrant mTOR-pathway signaling.188 ,​ 189


PDGFR inhibition with imatinib mesylate (alone or combined with sirolimus) has shown some antitumoral activity in case series and one phase II trial in chordoma patients, including several who had skull base chordomas.170 ,​ 185 ,​ 190 ,​ 191 An overall clinical benefit of 64% and stabilization rate of 70% were observed in advanced chordoma patients during the phase II trial.190 Sorafenib, a dual inhibitor targeting PDGF and VEGF pathways, has also been prospectively assessed in a phase II trial of patients who had advanced and metastatic chordomas.192 The response rate was low (3.7%), although the 9-month PFS was 73% and the 12-month OS 86.5%. The modest efficacy of PDGFR inhibition probably reflects the limitation of drug delivery into the chordoma and thus drug exposure.144 ,​ 145 ,​ 169 ,​ 192 EGFR inhibition has also shown responses and clinical improvement in a few case series and in a single phase II trial in patients who had recurrent chordoma, including skull base tumors.145 ,​ 193 ,​ 194 ,​ 195 ,​ 196 ,​ 197 Anti-EGFR monoclonal antibodies such as cetuximab, and EGFR tyrosine kinase inhibitors such as gefitinib, erlotinib, and lapatinib, have been studied.145 ,​ 193 ,​ 194 ,​ 195 ,​ 196 ,​ 197 Median PFS was 8 months in a phase II trial of lapatinib, suggesting the need for a larger trial to establish the role of this strategy.193


Inhibition of proangiogenic growth factors such as VEGF is another option for recurrent chordomas. In a small clinical series including patients who had skull base tumors, the combined use of erlotinib and bevacizumab, a humanized anti-VEGF antibody, led to disease stabilization.198 Small numbers of patients have been treated with the VEGF inhibitors pazopanib or sunitinib.120 One partial response was achieved with sunitinib, and 50% of patients benefited clinically from pazopanib, exhibiting 14- to 15-month disease stabilization. These data need further confirmation to confirm the place of anti-VEGF therapy in chordomas. Other potential therapeutic targets in chordomas, such as the inhibition of STAT3 with SD-1029, or the selective de novo purine synthesis inhibitor in MTAP-deficient chordomas, have shown promise in preclinical studies.164 ,​ 199 ,​ 200


Although epidemiologic data show an improvement over time in the management of chordomas, with improved survival and local control, this effect is related mostly to earlier diagnosis made possible by improved imaging techniques, to improved surgical resection using endoscopic endonasal approaches, and to recent developments in radiation techniques.118 ,​ 121 ,​ 136 ,​ 138 ,​ 139 The place of systemic therapies remains suboptimal and is confined to the recurrent setting when no other options are available. Multiple targeted therapies have been tried with modest results. These approaches need further development based on an improved understanding of the underlying molecular pathways relevant to chordoma and on the increasing use of molecular characterization of chordomas that match the molecular target with a targeted therapy.



10.4 Craniopharyngioma



10.4.1 Introduction to Craniopharyngiomas and Treatment Overview


Craniopharyngiomas (CPs) are rare slow-growing solid or mixed solid–cystic tumors that arise from remnants of Rathke’s pouch found along the midline from the nasopharynx to the diencephalon. These epithelial tumors commonly occur close to the optic chiasm in the suprasellar area but may also be observed within the sella, the third ventricle, and the optic system.201 ,​ 202 ,​ 203 ,​ 204 ,​ 205 CP often decrease life span and can consequently be considered low-grade malignancies. These tumors show no gender predilection and have a bimodal occurrence between 5 and 14 years in children and 50 and 75 years in adults.206


Clinical presentation is variable, depending on the tumor location, and may consist of visual acuity and field disturbances, a wide range of endocrinopathies, headaches caused by compression or hydrocephalus or meningeal spread, and depression. Growth retardation in children and sexual dysfunction in adults are the most frequent endocrine manifestations. Diabetes insipidus is also frequent. Endocrine and visual assessments, including visual field testing, are part of the systematic presurgical and postsurgical usual and customary evaluations. Neuroimaging most often shows a calcified cystic mass in the parasellar region, but both characteristics may be lacking in up to 25% of cases. Based on histological patterns, adamantinomatous and papillary CP are separate entities in the WHO classification.207 Both subtypes seem to share the same evolution in terms of treatment efficacy and survival.208 ,​ 209


Surgery is the primary and major component of initial treatment. Surgery provides a diagnosis as well as relief of symptoms. Maximal safe resection is advocated, but cyst aspiration followed by partial resection is an alternative in some cases considering that the expected benefits of surgery should be balanced with potential surgical morbidities.


RT is indicated after partial resections and at recurrence.210 Stereotactic RT and RSR, IMRT, and proton beam RT are most often employed.210 ,​ 211 ,​ 212 ,​ 213 ,​ 214 ,​ 215 ,​ 216


Compressive cysts need to be addressed when causing hydrocephalus or visual or hypothalamic disturbances. Percutaneous aspiration or aspiration via an Ommaya reservoir, intracavitary irradiation with stereotactic administered radioisotopes, and intracavitary chemotherapy may be used for this purpose.217 ,​ 218 ,​ 219 ,​ 220 ,​ 221 ,​ 222 ,​ 223 ,​ 224 ,​ 225 ,​ 226 ,​ 227 ,​ 228 ,​ 229 Surgery and RT may result in a wide range of endocrine, visual, neurological, and cerebrovascular side effects that may appear early or late, depending on clinical context. These treatment-related deficits include panhypopituitarism and related complications such as obesity, diabetes insipidus, sleep and temperature regulation disorders, visual field impairment, cognitive and behavioral disorders, arterial stenosis and ischemic strokes, cerebral cavernomas and aneurysm, moyamoya syndrome, meningioma, and high-grade glioma.230 ,​ 231 ,​ 232 ,​ 233



10.4.2 Pathogenesis, Genetics, Molecular and Cellular Overview


The two histologic subtypes of CP show distinct molecular genetics. Activation of the Wnt signaling pathway via activation of mutations in the gene encoding β-catenin (CTNNB1) are reported in 96% of adamantinomatous CP (ACP), a subtype of CP seen mostly in the pediatric population.234 ,​ 235 ,​ 236 ,​ 237 ,​ 238 ,​ 239 This driver mutation leads to accumulation of the protein β-catenin, which plays a role in cell signaling and cell adhesion. The Wnt signaling pathway promotes cell proliferation and differentiation. When the Wnt pathway is activated, β-catenin localizes to the cytoplasm and nucleus and can be detected through standard immunochemistry and used as a marker of CTNNB1 mutation status.235 ,​ 236 ,​ 239 ,​ 240


In one study using whole-exome sequencing, mutations of BRAF V600E oncogene were reported in 95% of papillary CP (PCP), which are commonly observed in the adult population and are not found in ACP.239 In another study using targeted Sanger sequencing, the BRAF V600E mutation was shown in 81% of PCP and in 12.5% of ACP displaying the CTNNB1 mutation as well. These data suggest that these mutations are not mutually exclusive but rather tend to be predominant in one specific subtype of CP.240 Immunohistochemistry provides the same sensitivity for detection of the BRAF V600E mutation as sequencing and is much more widely available.235 ,​ 236 ,​ 239 ,​ 240 Circulating BRAF V600E DNA has been detected in blood during treatment with a BRAF inhibitor, indicating that “liquid biopsy” may become a tool in the diagnosis and response to treatment of BRAF V600E mutated CP.235 The activating mutation BRAF may serve as a potential therapeutic target.



10.4.3 Systemic Therapies



Intracavitary Chemotherapy

An alternative approach to managing cystic CP is administration of intracystic bleomycin by way of an implanted catheter and reservoir such as an Ommaya device.228 Experience is more limited than with intracavitary irradiation, but in one series of 17 children, intracystic bleomycin was well tolerated, with five complete remissions and a median progression-free interval of 1.8 years.229 This approach may have a role in delaying RT or aggressive surgery.



Biological Therapy

The intracystic administration of IFN-α via an intracystic catheter connected to an Ommaya reservoir was assessed in more than 75 patients.241 ,​ 242 ,​ 243 Three million units of IFN-α were administered three times per week. One cycle was defined as 4 weeks of therapy and a total of 36 million units. Patients received from one to nine cycles. In all studies, tolerance to therapy was good, no drug interruption occurred, and significant clinical and radiologic responses were observed in the majority of patients.



Targeted Therapies

The use of BRAF inhibitors in BRAF-mutated PCP may become a new therapeutic option, as suggested by dramatic isolated responses.244 ,​ 245 ,​ 246 Nevertheless, when vemurafenib was used, regrowth was observed as soon as the targeted therapy was interrupted.244 Combination of a BRAF inhibitor (dabrafenib) and a MEK inhibitor (trametinib) is the new standard treatment in BRAF-mutated melanoma, providing optimal inhibition of the MAPK kinase (MEK) and RAS pathways with a significantly reduced risk of death compared with that seen for use of vemurafenib alone.247 This approach combining dabrafenib and trametinib has recently been tried successfully in patients who have BRAF-mutated CP and will be further studied in a National Cancer Institute–sponsored multicenter clinical trial.235 ,​ 245 ,​ 246 ,​ 248


Future trials are needed to confirm the value of targeted therapies in CPs with respect both to time of implementation (neoadjuvant to facilitate resection, adjuvant after incomplete resection or at relapse) and to overall activity in the multidisciplinary approach of these tumors.249



10.5 Giant Cell Tumor of the Bone



10.5.1 Introduction to Giant Cell Tumor of the Bone and Treatment Overview


Giant cell tumor of the bone (GCTB) is a rare, benign, slow-growing but locally destructive intraosseous and osteolytic tumor of skeletally mature young adults in their second to fourth decade, predominant in females and Asians. There is an increased incidence of GCTB of the skull in Paget’s disease. Although GCTB is considered a benign lesion, its evolution is unpredictable. Neurologic deficits often result from local invasive and compressive effects of the tumor. Prognosis is impacted by recurrence, malignant transformation, and the rare (< 5%) occurrence of metastases, predominantly to lungs. The local recurrence rate is high in cases of incomplete surgery, mostly within 2 years of surgery. The recommended follow-up period is 5 years only, because recurrences after that period are quite rare. Clinically, patients present with multiple cranial nerves palsies and headaches. Minimal work-up should include serum calcium and phosphorous, serum parathyroid hormone, bone scan, chest X-ray, contrast brain CT, and MRI. Additionally, some investigators advise performance of a chest CT, at least at time of recurrence, because lung metastases are most common at recurrence but on occasion may be inaugural.


A lytic lesion resulting from an intratumoral hemorrhage is evocative but not pathognomonic of the diagnosis by either skull film or CT scan.250 ,​ 251 CT imaging is more accurate for assessment of bone architecture and MRI for the surrounding neurovascular structures. Either may show a hypervascularized and cystic lesion.252 ,​ 253 The Campanacci classification categorizes GCTB into three grades (I, II, III), based on clinical and radiological features, but does not correlate with pathology, nor is it prognostic.254


GCTB consist of three cell-types: neoplastic giant cell tumor stromal cells, which have high proliferative activity; recruited mononuclear histiocytic cells; and multinuclear giant cells. The histologic heterogeneity of GCTB increases the difficulty of diagnosis, which is compounded by the limited sampling associated with core or fine-needle biopsies. It can be challenging to identify the atypical neoplastic stromal cells among the reactive giant cells and thus to distinguish between benign and malignant GCTB.255 ,​ 256


Until now, surgery has been the mainstay of local treatment, achieving generally good local control rates when resection is complete. Recent advances in the understanding of the molecular pathogenesis of GCTB, and improvements in RT, have led to new and better-tolerated therapeutic options. Nevertheless, to preserve the adjacent neurovascular structures and minimize surgery-related injury, resection is often incomplete and adjuvant therapy thus required. To decrease the risk of post-RT malignant transformation, the NCCN has recommended RT as the last adjuvant option notwithstanding the established radiosensitivity of GCTB and its excellent local control rate. Stereotactic radiosurgery and IMRT have been recently used as adjuvant therapies but require further validation.251 ,​ 257 ,​ 258 Denosumab, a monoclonal antibody targeting the receptor activator of nuclear factor-κB ligand (RANKL), was approved by the FDA in 2013 for the treatment of GCTB.255 ,​ 256 Other systemic therapies are discussed hereafter.259 ,​ 260



10.5.2 Pathogenesis, Genetics, and Molecular and Cellular Pathways


The proliferative fraction of GCTB is thought to be derived from the stromal compartment, but the absence of cytologic aspects of malignancy in stromal cells, combined with the lack of clonal cytogenetic structural aberrations in GCTB in most studies, suggests that these cells may be reactive and not neoplastic.261 ,​ 262 ,​ 263 ,​ 264 However, neoplastic stromal cells and mesenchymal stem cells share similar differentiation characteristics. When present, the overexpression of p53 and centrosome amplification are correlated with a high risk of recurrence, and the overexpression of p53 is also associated with an increased risk of metastases. Centrosome amplification is more frequent in recurrent and malignant GCTB.265 ,​ 266 Fifty-four percent of GCTBs have 20q11 amplifications.267 Genetic instability is also suggested by the presence of a driver mutation in H3F3A, exclusively in the stromal cells in > 90% of GCTB.268 Furthermore, telomere dysregulation is present in up to 70% of GCTB, with a telomere protective-capping mechanism permitting telomere length maintenance.269 ,​ 270 ,​ 271 Telomere maintenance markers such as human telomerase reverse transcriptase and promyelocytic leukemia body-related antigens are expressed by GCTB.271 ,​ 272 ,​ 273 A telomeric fusion in chromosomes 11p, 13p, 15p, 18p, 19p, and 21p is present in more than 80% of GCTB, predominantly in grade III tumors.274


The multinucleated giant cells in GCTB result from the recruitment of CD68-positive monocytes attracted by SDF-1 and MCP-1, which are produced by stromal cells, and by the VEGF present in the stromal environment, because these monocytes express VEGFR1 (Flt1) among other macrophage markers.250 At a molecular level, RANKL (receptor activator of nuclear factor kappa B [NF-kB] ligand) plays a major functional role in the molecular pathogenesis of GCTB. The osteoblast-like mononuclear stromal cells display a high rate of expression of RANKL.255 ,​ 262 ,​ 275 ,​ 276 ,​ 277 ,​ 278


Receptor activator of nuclear factor-kappa beta (RANK) ligand (RANKL) interactions and macrophage colony–stimulating factor (M-CSF) stimulate recruitment of osteoclastic cells from normal mononuclear preosteoclast cells that become multinucleated osteoclast-like giant cells.279 ,​ 280 ,​ 281 ,​ 282 ,​ 283 Through a cathepsin K- and MMP-13–mediated process, these giant cells cause the characteristic bone destruction observed in these lesions.284 ,​ 285 ,​ 286 Runx2 may be a driver in cytokine-mediated MMP-13 expression in GCTB stromal cells.287 CCAAT/enhancer binding protein beta(C/EBPbeta), by activating RANKL promoter, seems to be a significant factor in GCTB pathophysiology.288 Given the predominant role of RANKL in the genesis of these multinucleated giant cells, the inhibition of RANKL signaling by the use of the targeted antibody denosumab has proven highly effective.


The CD33 + phenotype of the giant cells in GCTB may also in the future be targetable using an anti-CD33 antibody such as gemtuzumab.282 ,​ 283 Some cells also are positive for PDGFA, C-kit, and EGFR, all of which can be therapeutically targeted using commercial drugs.289 ,​ 290 The EGFR is expressed by neoplastic stromal cells and plays a role in their proliferation, in osteoclastogenesis, and likely in the progression of the disease.289


Less prominent RANKL-independent pathways are involved in GCTB osteoclastogenesis and may represent another druggable target.279 These RANKL-independent pathways include tumor necrosis factor-α, interleukin-6, tumor growth factor-β, a proliferation-inducing ligand, B-cell activating factor, nerve growth factor, insulin-like growth factor I (IGF-I), and IGF-II. Further understanding of GCTB molecular pathogenesis will further inform the development of novel molecular-based therapies.



10.5.3 Systemic Therapies



Chemotherapy

There is limited evidence regarding the benefit of chemotherapy in benign GCTB, and because more efficient and better-tolerated systemic alternatives exist, it is usually not prescribed in these tumors except in cases of malignant GCTB. Doxorubicin, cisplatin, methotrexate, ifosfamide, and cyclophosphamide have been used in advanced aggressive, unresectable tumors, although in nonrandomized studies only.261 ,​ 291 ,​ 292 ,​ 293 ,​ 294 ,​ 295 ,​ 296 ,​ 297



Interferon Alpha

Limited retrospective data suggest activity of IFNα in the treatment of aggressive GCTB but at the cost of poor patient tolerance—a common side effect of systemic interferon-based therapy.296 ,​ 297 ,​ 298 ,​ 299 Consequently, the benefits and risks of IFNα should be balanced in clinical situations in which there are no alternatives.



Biphosphonates

Biphosphonates, with their antiosteoclastic properties, have shown activity in vitro in GCTB.300 ,​ 301 However, clinical data are scarce, with limited case reports, retrospective series, and one phase II trial having assessed zoledronic acid.302 ,​ 303 ,​ 304 ,​ 305 Modest clinical benefit and long-term local disease control have been reported in the adjuvant setting but without any significant preventive effect on local recurrence.250 ,​ 300 ,​ 302 ,​ 303 ,​ 304 ,​ 305 ,​ 306


Little information regarding the utility of calcitonin and sunitinib has been reported, but they represent other potential treatment options for recurrent GCTB.250 ,​ 307 ,​ 308 Local control is occasionally seen with oral steroids in patients who have Paget’s disease and GCTB.309 ,​ 310



Denosumab

Denosumab is a fully human monoclonal antibody against RANKL. Its clinical benefit has been shown in phase II trials involving recurrent or unresectable GCTB.260 ,​ 311 ,​ 312 Reported objective response rates of 86% are seen along with a reduced need for resective surgery. Long-lasting responses are seen in most patients (96% PFS at a median follow-up of 15 months), which combined with a favorable toxicity profile (1% incidence of osteonecrosis) make denosumab the therapy of first choice in patients who have GCTB (Fig. 10.1).251 ,​ 312

Fig. 10.1 Axial postcontrast T1-weighted MRIs of a 13-year-old girl who had a large skull base giant cell tumor. After the 6/15 scan, she was started on denosumab at 120 mg subcutaneous injection on days 1, 8, and 15 and then every 4 weeks. Scans show continuing response to therapy.

Based on results in 305 patients from two phase II trials, denosumab has been approved for patients who have unresectable GCTB or in whom a surgery would result in significant functional morbidity. NCCN guidelines emphasize indication for use of densoumab in instances of unresectable tumor. However, there are no data regarding the optimal duration or dose schedule of denosumab. Additionally, the optimal timing for the use of denosumab is also not well defined. Denosumab use has been proposed in the neoadjuvant setting, when surgery is expected to be incomplete, and when disease is clinically aggressive or is associated with a low performance status (PS) and impaired quality of life.273 The goals are to improve the resectability of the tumor by reducing its size (i.e., downstage) and to further achieve improvement in performance status (PS) and quality of life (QoL). Potential benefits in the adjuvant setting after surgery, to reduce local recurrence, is also under evaluation.312


There are limited data on the long-term safety of denosumab, especially relating to the most serious treatment-related adverse event, mandibular osteonecrosis. New challenges come with new therapies, and response assessment in GCTB is no exception. Postdenosumab treatment pathology can be challenging to interpret in some cases.313 Most often seen is a significant reduction of giant cells and neoplastic stromal cells, with some degree of bone reformation. However, cellular atypia and some patterns of ossification may suggest osteosarcoma or an undifferentiated pleomorphic sarcoma.255 ,​ 260 ,​ 313 Correlation of posttreatment pathology with clinical and imaging data is often necessary to achieve a correct diagnosis.


Although the role of (18F-FDG) PET scanning is not clear at time of diagnosis in GCTB, FDG-PET does appear to be a sensitive marker of early response to denosumab, as changes in standardized uptake values reflect tumor metabolism and angiogenesis.314 ,​ 315 ,​ 316 ,​ 317 Nevertheless, MRI should not be replaced by FDG-PET; rather, the latter provides correlative information that may augment MRI findings.


The introduction of denosumab has profoundly altered the management of GCTB of the skull base, providing a new treatment option that augments surgery. However, because many questions remain unanswered regarding the practical aspects and timing of denosumab treatment, further prospective randomized trials are needed.



10.6 Conclusion


“Benign” tumors of the skull base are characterized by a high local recurrence rate and attendant consequences, including diminished QOL and decreased survival. In some of these tumors, such as chordomas, improvements in global management have been observed over time with earlier diagnosis and better local control as a result of improved imaging and surgical techniques. Nevertheless, management of these tumors remains challenging, and despite the recent making of considerable progress in understandings of their genetics and molecular pathogenesis, many limitations remain with respect to current systemic therapies.


Because these tumors are rare, treatment of numbers of patients large enough to allow determination of the benefits of novel targeted therapies can be difficult. As a result, most studies of targeted agents have involved a single institution and comprised small numbers of patients. Furthermore, and as has become increasingly apparent in other tumor types, molecular characterization and matching of actionable targets with targeted therapies are essential to best determining the role of these new therapies. Studies such as these will require multi-institutional efforts and likely also industry sponsors to provide access to novel targeted therapies. The application of precision or personalized oncology based on the molecular profiling of aberrant pathways in individual patients holds great promise for advancing the treatment of patients who have skull-based benign tumors.



10.7 References














  • 12 NCCN Clinical Practice Guidelines in Oncology [Internet]. Available from: http://www.nccn.org/professionals/physician_gls/f_guidelines.asp#cns



































  • 46 Mawrin C, Chung C, Preusser M. Biology and clinical management challenges in meningioma. Am Soc Clin Oncol Educ Book ASCO Am Soc Clin Oncol Meet. 2015; 35:e106–e115























  • 68 Goldsmith B, McDermott MW. Meningioma.. Neurosurg Clin N Am 2006; 17 (2) 111-120, vi PubMed 16793503
























































  • 123 McMaster M. Update on the epidemiology of chordoma: SEER registry data 1973–2007. Poster presented at: Third International Chordoma Research Workshop, March 17–19, 2011; Bethesda, MD





















































































  • 207 Louis DN, Ohgaki H, Wiestler OD, Cavenee WK, eds. Craniopharyngioma. In: World Health Organization Classification of Tumours of the Nervous System, Editorial and Consensus Conference Working Group, Lyon, France: IARC Press; 2007



















































  • 257 Roeder F, Timke C, Zwicker F, et al. Intensity modulated radiotherapy (IMRT) in benign giant cell tumors—a single institution case series and a short review of the literature. Thieke C, Bischof M, Debus J, Huber PE




























































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

Stay updated, free articles. Join our Telegram channel

Feb 8, 2021 | Posted by in NEUROSURGERY | Comments Off on 10 Management of Benign Skull Base Tumors in Neuro-oncology: Systemic Cytotoxic and Targeted Therapies

Full access? Get Clinical Tree

Get Clinical Tree app for offline access