Uncovering Genes Behind Cartilage Tumors and Vascular Anomalies Using Genomic Sequencing

Study Design:

The study will recruit and enroll up to 100 patients with Ollier disease (OD) and Maffucci syndrome (MS) over five years. Each participant will be seen at the NIH Clinical Center (CC) for an inpatient study visit of 5 days or longer.

During that visit, the following procedures will be performed: dual-energy x-ray absorptiometry (DXA), whole-body X-ray, magnetic resonance angiography (MRA), whole-body positron emission tomography (PET) scan, a magnetic resonance imaging (MRI), and brain single-voxel magnetic resonance spectroscopy. After being seen at the NIH CC by the study team, the participants will receive a summary of all the results of the tests and consults completed at the NIH.

Family members enrolled for genome sequencing will not undergo any clinical evaluations at the NIH CC.

Dr. Gordon will oversee the implementation of the study at the NIH. Dr. Marini was involved in the original NIH grant application and has retired but she retains an appointment at the NIH as a Scientist Emeritus/ Volunteer. Dr. Marini will serve on the protocol as a Senior Clinical Consultant and work at the NIH CC. She will not consent participants but will be present at many of the study visits.

Members of the JHU study team will have permission to work and/or observe at the NIH; they will however not be considered part of the NIH study team and will not be identified as NIH Associate Investigators.

Objectives:

Ollier disease (OD) and Maffucci syndrome (MS), are characterized by multiple enchondromas that result in severe skeletal deformities during early childhood and an increased risk for chondrosarcomas and other malignancies. Patients with MS also have vascular anomalies. Chondrosarcoma is a malignant tumor that originates from cartilaginous cells. It is the third most common primary malignancy of bone after myeloma and osteosarcoma. It accounts for about 20% of bone tumors and is diagnosed in approximately 600 patients each year in the United States. Up to 40% of the chondrosarcomas arise from an enchondroma. The risk for chondrosarcoma in OD is up to 45.8% and in MS up to 57.1%. Currently, the only treatment for patients with these disorders is surgical; there is no effective pharmacologic therapy.

We identified heterozygous germline loss of function variants in PTPN11 (encoding a non-receptor protein tyrosine phosphatase SHP2) causing MC. In preliminary studies, we also identified the PTPN11 R138X variant in a retiform hemangioendothelioma of a patient with MS and the germline PTPN11 L560F variant in a patient with OD. PTPN11 encodes SHP2, a cytosolic protein tyrosine phosphatase involved in an early step in RAS/MAPK signaling downstream of several receptor tyrosine kinases including EGFR and FGFR. Using immunoblot analysis we measured the expression of the end products of this pathway, pERK1 and 2 in fibroblasts from patients with MC and OD and observed decreased expression of pERK1 and 2 in patients compared to the controls (Sobreira et al., unpublished). Similarly, experiments in the Ptpn11 conditional knockout mouse models showed reduced ERK1/2 signaling in chondrocytes lacking SHP2. Taken together, these results suggest reduction RAS/MAPK signaling can lead to the formation of enchondromas. Heterozygous somatic variants of IDH1 and 2 have been identified in the tumors of a fraction of the patients with MS and OD. Neither variant was identified in the germline DNA of the affected individuals. And our preliminary studies have identified germline variants KDM4C and HIF1A, 2 genes known to interact with IDH1 and 2, as the candidate genes in 2 unrelated probands with multiple enchondromas.

On basis of these results, we hypothesize that OD and MS are tumor predisposition syndromes caused by germline variants such as neurofibromatosis type 1 and schwannomatosis. Subsequent hits in the same or different genes such as IDH1 and IDH2 or other as yet identified genes are involved in the formation of enchondromas and chondrosarcomas. Moreover, these variants likely down-regulate the RAS/MAPK pathway or are in genes that interact with IDH1 and 2.

To test this model, we will pursue the following Specific Aims:

Specific Aim 1. Comprehensively define the phenotypic features of patients with OD and MS.

Hypothesis: OD and MS are distinct under-characterized cancer susceptibility syndromes with diverse features. Recently, we proposed diagnostic and surveillance guidelines for systematic evaluation of patients with OD and MS9. Here, we will: (a) identify the complete set of phenotypic features characteristic of patients with OD and MS by performing detailed assessment of their clinical and family histories and physical features at the NIH/CC; (b) determine the number, size, and location of enchondromas and vascular anomalies and identify other tumors in these patients by performing full body MRI or X-ray, MRA , and brain magnetic resonance spectroscopy; and (c) establish a biobank of specimens from patients and family members for genetic, metabolic, and functional testing.

Specific Aim 2. Search for missing germline (and/or somatic mosaic) causative variants (coding and non-coding) in DNA from blood, enchondromas, chondrosarcomas and associated hemangiomas. We will perform WGS on germline and tumor (enchondroma, chondrosarcomas and hemangiomas) DNA to identify the germline variants in the blood and the subsequent hits involved in the tumor s formation. We will also do RNA sequencing (RNAseq) on RNA isolated from tumor samples to perform a gene expression analysis and identify other possible pathways affected in these disorders; and to aid to the interpretation of variants detected by WGS, especially synonymous and non-coding variants affecting splicing and RNA abundance, respectively.

Specific Aim 3. To test the hypothesis that standard analysis may overlook a causative variant because of an alternative mode of inheritance, we will apply novel analytic approaches that investigate alternative modes of inheritance typically not considered. We will investigate paternal and maternal imprinting modes of inheritance and investigate variants identified in the genes that escape X-inactivation. With WGS, we will also be able to investigate the variants in X-linked genes that escape X-inactivation and the variants in Y-linked genes.

Specific Aim 4. Determine the effect of causative variants of OD or MS in the HIF-1 pathway.

Hypothesis: Causative variants of OD and MS cause downregulation of HIF1A and its target genes. Our preliminary analysis of RNAseq data from fibroblast and enchondroma samples from patients with OD and MS showed impaired induction of HIF-1-related genes in contrast to other cancer settings. HIF1A variants have not been identified as disease-causing, but seven of our patients have a causative variant in this gene. We found that HIF1A and VHL were mutated at a higher rate among patients with OD and MS compared to controls (p<0.05). Therefore, we will: (a) use knock-in cell-based models to investigate the effect of the causative variants we identified in HIF1A (mutated in seven patients), VHL (mutated in six patients), and KDM4C (mutated in four patients) on HIF-1 transcriptional activity using a dual-luciferase assay40; and (b) identify differentially expressed genes and disrupted pathways with enchondroma RNAseq.

Specific Aim 5. Data Sharing. As we are doing in other projects, we will submit our sequence data to dbGaP to facilitate data sharing. The JHM IRB will make the determination that the sharing is consistent with the NIH genomic policy. Also, to identify others with data relevant to our candidate variants and genes, we will use GeneMatcher and the Matchmaker Exchange. We will also submit our data to ClinVar and COSMIC database. We are eager to participate in a consortium to establish the Kids First Data Resource and to the overall genetic dataset for childhood cancers at the NCI Genomic Data Commons.