Brain Gliomas and Ollier Disease: Molecular Findings as Predictive Risk Factors?

Simple Summary As reported in the recent Classification of the Tumors of the Central Nervous System (2021), the role of molecular markers in CNS tumor diagnosis and classification, and accordingly in the decision making of the therapeutic process, has become even more decisive. In this scenario, we performed an accurate literature review of patients with Ollier disease harboring brain gliomas, disclosing only thirty cases; our own case was also included. Most of the reported cases belong to the pre-molecular era. We found a strong relationship between an IDH mutation in the enchondroma of patients with Ollier disease and the occurrence of brain gliomas; this finding may allow an early diagnosis of brain glioma in these patients, detecting the tumor when still small and silent, thus allowing the neurosurgeon, oncologist, and radiotherapist to plan the best management. Abstract Background: Ollier disease (OD) is a rare nonhereditary type of dyschondroplasia characterized by multiple enchondromas, with typical onset in the first decade of life. Surgery is the only curative treatment for primary disease and its complications. Patients with OD are at risk of malignant transformation of enchondromas and of occurrence of other neoplasms. Methods: A wide literature review disclosed thirty cases of glioma associated with OD, most of them belonging to the pre-molecular era. Our own case was also included. Demographic, clinical, pathologic, molecular, management, and outcome data were analyzed and compared to those of sporadic gliomas. Results: Gliomas associated with OD more frequently occur at younger age, present higher rates of multicentric lesions (49%), brainstem localizations (29%), and significantly lower rates of glioblastomas (7%) histotype. The IDH1 R132H mutation was detected in 80% of gliomas of OD patients and simultaneously in enchondromas and gliomas in 100% of cases. Conclusions: The molecular data suggest a higher risk of occurrence of glioma in patients with enchondromas harboring the IDH1 R132H mutation than those with the IDH1 R132C mutation. Thus, we suggest considering the IDH1 R132H mutation in enchondromas of patients with OD as a predictive risk factor of occurrence of glioma.


Background
Ollier disease (OD) is a rare nonhereditary type of dyschondroplasia, with a prevalence of 1:100,000 per year, characterized by enchondromatosis and areas of dysplastic cartilage, caused by a developmental anomaly in enchondral ossification mainly involving long tubular bones, such as the femur, tibia, and fibula, but also flat bones such as the pelvis In 2017, because of the occurrence of an epileptic seizure, the patient underwent a brain MRI, which disclosed small (<2 cm) intracerebral lesions in both frontal lobes. A stereotactic biopsy of the left frontal lesion, performed at another institution, was in favor of a WHO grade II IDH-1 (R132-H) mutant fibrillary diffuse astrocytoma, with proliferation index Ki67-MIB1 1%. The adjuvant treatment with temozolomide, for 6 months, resulted in partial regression of the two lesions.
Following treatment, the patient was symptom-free until July 2021, when he experienced left partial motor seizures. A contrast-enhanced brain MRI (Figure 2b,c) showed a multifocal intraparenchymal bi-hemispheric lesion, with the mass of maximum diameter in the frontal-temporal-insular region of the right hemisphere, as well as hypointense in T1-and hyperintense in T2-weighted sequences with deep small areas of contrast enhancement; spectroscopy and perfusion sequences showed increased choline peak, decreased N-acetyl-aspartate peak, and increased blood flow, respectively, at the site of lesion ( Figure 2d). The small left frontal lesion was stable. A bone lesion at the right medial sphenoid wing and parasellar region, with the MRI features of an enchondroma, was also evident (Figure 2a). Following treatment, the patient was symptom-free until July 2021, when he experienced left partial motor seizures. A contrast-enhanced brain MRI (Figure 2b,c) showed a multifocal intraparenchymal bi-hemispheric lesion, with the mass of maximum diameter in the frontal-temporal-insular region of the right hemisphere, as well as hypointense in T1and hyperintense in T2-weighted sequences with deep small areas of contrast enhancement; spectroscopy and perfusion sequences showed increased choline peak, decreased N-acetylaspartate peak, and increased blood flow, respectively, at the site of lesion (Figure 2d). The small left frontal lesion was stable. A bone lesion at the right medial sphenoid wing and parasellar region, with the MRI features of an enchondroma, was also evident (Figure 2a).
The detection of a contrast-enhanced area in the latest brain MRI was in favor of tumor progression.
At admission, the neurology examination was normal. The patient underwent a wide microsurgical resection of the frontal-parietal opercular tumor component of the right hemisphere, which appeared as a white-reddish and moderately vascularized mass. The part of the tumor encompassing the M3 branches of middle cerebral artery was left, to avoid postoperative neurological deficits and to preserve the patient quality of life. The component of the tumor that affected the frontal lobe of the left hemisphere was also not removed, because it did not change over time. The postoperative course was uneventful. A contrast-enhanced brain MRI at 48 h showed satisfactory tumor removal of the supra-sylvian component without neurosurgical complications. The detection of a contrast-enhanced area in the latest brain MRI was in favor of tumor progression.
At admission, the neurology examination was normal. The patient underwent a wide microsurgical resection of the frontal-parietal opercular tumor component of the right hemisphere, which appeared as a white-reddish and moderately vascularized mass. The part of the tumor encompassing the M3 branches of middle cerebral artery was left, to avoid postoperative neurological deficits and to preserve the patient quality of life. The component of the tumor that affected the frontal lobe of the left hemisphere was also not removed, because it did not change over time. The postoperative course was uneventful. A contrast-enhanced brain MRI at 48 h showed satisfactory tumor removal of the suprasylvian component without neurosurgical complications.

Diagnosis
The histology showed a tumor with fibrillary background and increased cellularity, with pleomorphic cells but without microvascular proliferation or necrosis; the immunohistochemical study was positive for GFAP and IDH1 R132H mutation. IDH1 mutation was confirmed by DNA sequencing. The mitotic activity was inconsistent, although the proliferation index Ki67-MIB 1 was 15% ( Figure 3). The positive methylation status of the MGMT gene promoter was assessed by Methylation Specific PCR (MSP) and confirmed by extrapolation of MGMT methylation status from raw data of EPIC array (Illumina 850k). To assess the codeletion status of 1p19q, we performed Multiplex Ligation Probe Amplification (MLPA): the absence of co-deletion was also confirmed by the analysis of

Diagnosis
The histology showed a tumor with fibrillary background and increased cellularity, with pleomorphic cells but without microvascular proliferation or necrosis; the immunohistochemical study was positive for GFAP and IDH1 R132H mutation. IDH1 mutation was confirmed by DNA sequencing. The mitotic activity was inconsistent, although the proliferation index Ki67-MIB 1 was 15% ( Figure 3). The positive methylation status of the MGMT gene promoter was assessed by Methylation Specific PCR (MSP) and confirmed by extrapolation of MGMT methylation status from raw data of EPIC array (Illumina 850k). To assess the codeletion status of 1p19q, we performed Multiplex Ligation Probe Amplification (MLPA): the absence of co-deletion was also confirmed by the analysis of Copy Number Variation (CNVs) extrapolated from METHYLOME raw data ( Figure 3e).
In addition, we analyzed the epigenome profile of the tumor using a previously used bioinformatic tool (DKFZ, Heidelberg, Germany, https://www.dkfz.de/de/index.html, accessed on 7 May 2022). Generated methylation data were compared with the Heidelberg brain tumor classifier to assign a subgroup score for the tumor, compared with 91 different brain tumor entities. The bioinformatic analysis showed a strong match for IDH mutant astrocytoma methylation class [31,32]. We considered significant a gain of genetic materials if the score, obtained from ratio of case and control sample, was >0,4; as well, we considered significant a loss of genetic materials if the ratio score was <0.4.
In addition, we analyzed the epigenome profile of the tumor using a previously used bioinformatic tool (DKFZ, Heidelberg, Germany, https://www.dkfz.de/de/index.html, accessed on 2022/5/7). Generated methylation data were compared with the Heidelberg brain tumor classifier to assign a subgroup score for the tumor, compared with 91 different brain tumor entities. The bioinformatic analysis showed a strong match for IDH mutant astrocytoma methylation class [31,32].
The detection of inconsistent mitotic activity allowed us to exclude a high-grade tumor; nevertheless, because of the high value of Ki67, unusual for low-grade gliomas, and the contrast enhancement of the lesion, a typical feature of high-grade gliomas, to further discriminate between high-and low-grade tumors we performed an analysis of two prognostic negative biomarkers implemented in the new classification of tumors of the central nervous system [33]: TERT promoter mutations and CDKN2A/B gene deletion. To detect We considered significant a gain of genetic materials if the score, obtained from ratio of case and control sample, was >0.4; as well, we considered significant a loss of genetic materials if the ratio score was <0.4.
The detection of inconsistent mitotic activity allowed us to exclude a high-grade tumor; nevertheless, because of the high value of Ki67, unusual for low-grade gliomas, and the contrast enhancement of the lesion, a typical feature of high-grade gliomas, to further discriminate between high-and low-grade tumors we performed an analysis of two prognostic negative biomarkers implemented in the new classification of tumors of the central nervous system [33]: TERT promoter mutations and CDKN2A/B gene deletion. To detect TERT promoter mutation we assessed direct DNA sequencing, but the analyzed sample did not present promoter mutation. To verify CDKN2A/B gene deletion we analyzed CNVs but, similarly, the deletion was not present, as shown in Figure 3e.
The global histopathological, genomic and epigenomic analysis oriented the diagnosis in favor of WHO grade 2, IDH mutant astrocytoma [34]. The contrast-enhanced brain MRI taken one month after surgery confirmed the wide resection of the right frontal-opercular component of the tumor (Figure 4).
TERT promoter mutation we assessed direct DNA sequencing, but the analyzed sample did not present promoter mutation. To verify CDKN2A/B gene deletion we analyzed CNVs but, similarly, the deletion was not present, as shown in Figure 3e.
The global histopathological, genomic and epigenomic analysis oriented the diagnosis in favor of WHO grade 2, IDH mutant astrocytoma [34].
The contrast-enhanced brain MRI taken one month after surgery confirmed the wide resection of the right frontal-opercular component of the tumor (Figure 4). The high value of Ki67-Li and the contrast enhancement on MRI witnessed aggressive clinical behavior, so the patient was discharged with indication to perform radiotherapy [35] and a close follow-up through contrast-enhanced brain MRI at 3 months.

Literature Review
The incidence of brain glioma ranges from 2.4% [18,28] to 16% [18,28] in patients with OD. To the best of our knowledge, only 31 cases, including our own, are reported in the literature (Tables 1-3). The high value of Ki67-Li and the contrast enhancement on MRI witnessed aggressive clinical behavior, so the patient was discharged with indication to perform radiotherapy [35] and a close follow-up through contrast-enhanced brain MRI at 3 months.

Literature Review
The incidence of brain glioma ranges from 2.4% [18,28] to 16% [18,28] in patients with OD. To the best of our knowledge, only 31 cases, including our own, are reported in the literature (Tables 1-3).
Patients were 18 males (60%) and 12 females (40%) with age at diagnosis of the brain glioma ranging from 6 to 55 years (median 26 years) ( Table 2). The brain glioma was a single lesion in 16 patients (51%) (mainly with frontal location) and multicentric in 15 (49%) (of which, gliomas were only supratentorial in 2/3 and both supra-and infratentorial in 1/3). The histological study mainly showed astrocytic tumors (62%) and gliomas of WHO grades II and III, whereas only 2 cases out of 29 (7%) were glioblastomas. Among the 10 patients on which biomolecular studies were performed, eight showed positive IDH1 R132H mutations in the brain glioma; in three of them, this mutation was simultaneously detected in the enchondroma.
The management of brain gliomas in the reviewed cases was as follows (Table 3): at the initial diagnosis, 16 patients (12 with multicentric and four with single lesion) underwent a biopsy, six a craniotomy with tumor resection, and three were not operated on; a craniotomy was performed in response to tumor progression in four patients who had undergone a previous biopsy. Radiotherapy (RT), reported in 16 cases, was administered at the initial diagnosis of 12 and at progression in two. Finally, only five patients were treated by chemotherapy.   The outcome was as follows (Table 3): among the 13 patients whose data were available, nine were stable and progression-free at the median follow-up of 15 months, and four experienced progression after a median follow-up of 41 months. Fifteen of eighteen patients with available data were still alive at the median follow-up of 33 months, while three died from 11 to 96 months after the initial diagnosis.

Discussion
Patients with OD are not only at risk of malignant transformation of enchondromas to chondrosarcomas [1,10,29], with a variable rate from 20 to 50% [1,10,29], but are also at risk of occurrence of other neoplasms, such as gliomas, ovarian tumors [4], pituitary and thyroid adenomas, and leukemia [3]. The prevalence of malignancies in individuals with OD is about 53% [28], among which the most common is represented by chondrosarcoma (30.6%) [28]. The diagnosis of cancer is made at a median age of 26 years (from 1 to 69 years), while the median time lapse between diagnoses of enchondromatosis and malignancy is eleven years [28].
The occurrence rate of glioma during the life of these patients ranges from 2.4% [18,28] to 16% [18,28], with only thirty cases reported in literature.
Prior to 2009, the only approved etiologic risk factors of occurrence of glioma were high-dose radiation and some rare familial cancer syndromes [40]. Since then, 10 inherited variants, around eight genes' defined risk loci, have been discovered to be associated to an increased risk of glioma's occurrence [40].
Most sporadic adult diffuse grade II and III gliomas harbor mutations of the IDH1 and/or IDH2 genes, which is considered the first event driving oncogenesis of these tumors [36,41].
Ollier disease is due to early post-zygotic IDH mutations, leading to somatic mosaic mutations of IDH1 or IDH2 [38,42]. IDH 1 mutations account for amino acid substitution in the active site of the enzyme in codon 132 (R132H), resulting in the abnormal production of 2-hydroxyglutarate, which causes histone and DNA methylation, as well as altered cellular differentiation, hence promoting tumorigenesis. In addition to gliomas, cartilaginous tumors experience IDH mutation [36][37][38]42], but unlike enchondromas, isolated IDH mutation is not enough to induce glioma genesis; further mutations, such as ATRX and/or TP53, are needed [41].
Compared to the sporadic forms of IDH mutant gliomas, gliomas occurring in OD patients are also mainly diffuse and low-grade, and more frequently involve frontal lobes; conversely, they are diagnosed at an earlier age (26 vs. 44 years) and show higher rates of multicentric lesions (49% in our review) and brainstem localization (29%), and significantly lower rate of glioblastomas histotype (7%).
Only 30 cases of gliomas associated to Ollier disease have been reported in the literature, and most of these belong to the pre-molecular era; therefore, only in seven studies [10,14,23,24,30], our own included, for a total of ten patients, the molecular analysis of gliomas has been performed ( Table 1): eight of these (80%) presented the IDH1 R132H mutation. Furthermore, only in three cases [23,30], our own included, molecular analysis was performed in both an enchondroma and glioma, and in all three cases (100%) the identical IDH1 R132H mutation was detected in both tumors.
Although from a small sample, these data suggest that patients with Ollier disease harboring the IDH1 R132H mutation in enchondroma are at major risk of developing gliomas, compared to enchondromatosis patients with the IDH R132C mutation in enchondroma. Furthermore, the early acquisition of IDH mutation accounts for the occurrence of glioma at a younger age in these patients. Therefore, once a diagnosis of OD with IDH R132H mutation is made, we consider it be useful to perform close surveillance through brain contrast-enhanced MRI, to detect possible early stage gliomas, when still asymp-tomatic, of small size, without compromission of the neurological functional areas nor vital neurovascular structures, so to plan the best tailored management. The incidence of postoperative complications and seizures is higher in patients with symptomatic compared to those with incidental low-grade glioma [43] and the overall survival (OS) of incidental low-grade glioma (LGG) patients treated with surgery is better than that of symptomatic LGG patients [39,44].
The median silent phase of glioma, defined as the time from the generation of the tumor to the occurrence of tumor-related symptoms, was estimated by Pallud et al. to be about 12 years (range 1.6-39.4 years) [45], and the average growth rate of LGG is 2.7 mm/year (range from 1 to 5 mm) [43].
The rate of malignant transformation of LGG ranges from 25% to 72% [46] and it is influenced by several risk factors, such as preoperative tumor size, velocity of diametric expansion, astrocytoma histology, extent of resection, and treatment with chemotherapy, radiotherapy, and adjuvant chemo-radiotherapy [46].
Being the extent of resection the main predictive risk factor of progression free-survival (PFS), overall survival and malignant progression free-survival (MPFS) in glioma patients [43], the possibility to detect a glioma in early stage, when of a small size, should allow to achieve the maximal safe tumor resection preserving the patient's quality of life.

Conclusions
We conclude that patients with Ollier disease which harbor the IDH1 R132H mutation in enchondroma could be at major risk to develop gliomas, compared to those with the IDH1 R132C mutation. Therefore, we suggest routine performance of molecular analysis of cartilaginous tumors in patients with OD, and, if IDH1 132H mutation is detected, performance of close surveillance through periodical ambulatorial and radiological control with contrast-enhanced MRI of the brain, to detect the eventual CNS tumor when asymptomatic and of small size, to plan the best management.
Author Contributions: S.C., conception, data collection, data review, drafting manuscript, reviewed drafted manuscript, and study supervision; G.M., study supervision; G.C., data collection; R.A.F., histology; M.D.B.D.C., study supervision; R.D.M., molecular analysis; L.C., study supervision; F.M., study supervision. All authors have read and agreed to the published version of the manuscript.
Funding: This research received no external funding.

Conflicts of Interest:
The authors report no conflict of interest associated with this study.

Limitations of the Study:
Although the literature review is wide and accurate, the sample of cases is rather small, due to the rarity of this pathological entity. Besides, many reviewed studies do not include complete data on several aspects, mainly concerning the molecular aspects.

Advantages:
The purpose of the study to establish a diagnostic protocol to make an early and prompt diagnosis of glioma associated to Ollier Disease, would allow to plan the best management for these patients improving their overall survival and quality of life. Further studies are needed to confirm our hypothesis.