RUNX2 (6p21.1) Amplification in Osteosarcoma

Abstract

Prior cytogenetic profiling of osteosarcomas has suggested that amplifications at the 6p12–21 locus are relatively common alterations in these tumors. However, these studies have been limited by variable testing methodologies used as well as by the relatively small numbers of cases that have been analyzed. To better define the frequency of this alteration, 111 osteosarcomas were profiled using hybridization capture-based next generation sequencing (NGS) platform (Memorial Sloan Kettering-Integrated Mutation Profiling of Actionable Cancer Targets, MSK-IMPACT) as part of an institutional clinical cancer genomics initiative. Using this platform, amplification at the 6p12–21 locus was determined by copy number assessment of the VEGFA and CCND3 genes. In addition, fluorescence in situ hybridization was used to assess copy number status for RUNX2, a known transcriptional regulator of osteoblastic differentiation which has previously been reported to be dysregulated in osteosarcomas. 6p12–21 amplification using NGS-based copy number assessment was confirmed in over a fifth of all cases tested (24 of 111, 21.6%). Most of these cases, when tested using FISH, were found to include RUNX2 within the amplified locus (17 of 18, 94.4%). While many laboratories lack access to large-panel NGS assays, the use of FISH to identify 6p12–21 amplification events by targeting RUNX2 represents a widely available diagnostic modality for the identification of such cases. This could help better define the role of RUNX2 in osteoblastic differentiation and serve as a surrogate for the identification of potentially targetable alterations such as VEGFA amplification at this locus.

1.0. Introduction

Initial functional studies of Runt-related transcription factor 2 (RUNX2) showed that transgenic mice with loss of function alterations of this gene showed a complete absence of bone formation [1, 2]. Currently, RUNX2 is regarded as a master transcriptional regulator of osteoblastic differentiation [3, 4]. Specifically, RUNX2 is thought to play a crucial role in the commitment of mesenchymal stem cells to an osteoblastic lineage and promotes osteoblastic differentiation through the expression of downstream genes that play a key role in bone matrix formation such as collagen type I alpha 1 chain (COL1A1), collagen type I alpha 2 chain (COL1A2), osteopontin/ secreted phosphoprotein 1 (OPN/SPP1), integrin binding sialoprotein (IBSP/BSP) and osteocalcin/ bone gamma-carboxyglutamate protein (OCN/BGLAP) [5–7]. This is supported by in vitro studies where for instance the induction of RUNX2 expression has been found to reproducibly promote osteogenesis [8].

It may therefore be hypothesized that dysregulation of RUNX2 may contribute to osteosarcoma pathogenesis. In fact, a few reports have documented chromosomal alterations of the 6p12–21 locus in osteosarcomas, which includes RUNX2 [6, 9–11]. In a study conducted by Lau et al, at least 7 cases of osteosarcoma (7/25, 28%) were found to harbor both amplifications and rearrangements at the 6p12–21 locus [9]. A follow up study by Lu et al documented amplifications at the same locus in 12 cases (12/48, 25%) with confirmatory gene expression profiling using quantitative RT-PCR. This study also revealed the presence of three candidate oncogenes at this locus: RUNX2, CDC5L and CCND3 [10]. Subsequently, Yang et al reported VEGFA amplification at this locus in 32 of 50 cases tested (64%) [12]. In addition, our initial studies had documented VEGFA and/or CCND3 amplification at this locus in 17 of 72 (23.6%) specimens tested [11].

Few studies have suggested that RUNX2 expression may predict response to chemotherapy and outcomes [6]. For instance, a study by Sadikovic et al assessed RUNX2 mRNA expression in a cohort of 22 cases of osteosarcoma with differential response to chemotherapy and concluded that osteosarcomas showed higher expression of this gene compared to normal osteoblasts [13]. In this study, tumors that lacked significant response to chemotherapy (defined as <90% necrosis) had on average a 3.3-fold higher expression of RUNX2 [13]. In addition, two separate studies have shown that either higher incidence of RUNX2 protein expression in metastases compared to primary tumors or higher RUNX2 protein expression correlated with poorer outcomes [14, 15]. In addition, our preliminary studies had shown that osteosarcomas with 6p12–21 gain showed a trend towards increased disease recurrence and/or metastasis within 5 years compared to cases that lacked this genomic amplification (32.1% vs 12.8%, p=0.05) [11]. At the molecular level, RUNX2 is thought to have a synergistic effect on cell cycle progression in the presence of TP53 and RB1 alterations, both of which are commonly seen in high grade osteosarcomas [6, 16, 17].

2.0. Materials and Methods

2.1. Patient Specimens

This study was approved by the Memorial Sloan Kettering Cancer Center institutional review board. Herein, we have profiled 111 osteosarcomas at the molecular level as part of an institutional clinical cancer genomics initiative [11].

2.2. Next Generation Sequencing Based Copy Number Assessment

This was done using a next generation sequencing assay that involves hybridization capture-based library preparation where the capture probes target approximately 1.5 megabases of the human genome and this is followed by deep sequencing of select non-coding and coding regions (Memorial Sloan Kettering Cancer Center Integrated Mutation Profiling of Actionable Cancer Targets, MSK-IMPACT). Details of this assay have been previously reported [18]. Accurate genome-wide copy number assessment in this assay is facilitated by homogenous distribution of single nucleotide polymorphism (SNP) tiling probes across the genome and based on previously reported criteria, amplifications at 6p12–21 were defined as a fold change ≥ 2.0 and borderline amplifications/gains were defined as a fold change ≥1.5 and <2.0 [19, 20]. This assay is currently approved by the United States Food and Drug Administration as a class II in vitro diagnostic test. As part of routine 468 gene panel MSK-IMPACT test, osteosarcomas were profiled for 6p12–21 copy number alterations using targeted probes for VEGFA and CCND3 (Figure 1A, C).

Figure 1: Histopathology, FISH and Copy Number Alterations.

A representative H&E stained image (A), with corresponding FISH results demonstrating amplification of RUNX2 (B) and next generation sequencing based results of copy number analysis have been depicted (C). Specifically, an osteosarcoma profiled using MSK IMPACT is shown, with relative (Log2) tumor/normal ratios on the y-axis and corresponding chromosomes on the x-axis. Amplifications at 6p21 (VEGFA/CCND3) have been depicted. Additional copy number changes include amplification of TERT (5p15), CARD11/PMS2/RAC1 (7p22), CDK4/MDM2 (12q14–15) and losses of JAK2/CD274/PDCD1LG2/PTPRD (9p24) and PIK3C3 (18q12). Each blue dot represents an individual probe region and amplified regions (6p21.1) are shown in red. MSK IMPACT, Memorial Sloan Kettering Cancer Center Integrated Mutation Profiling of Actionable Cancer Targets.

2.3. Fluorescence in situ Hybridization Based Copy Number Assessment

We hypothesized that the RUNX2 gene (not included in the MSK-IMPACT panel) was included in the amplified locus. RUNX2 amplification status was therefore assessed using fluorescence in situ hybridization (FISH). FISH was performed as previously described using probes for RUNX2 (labeled with a red fluorophore; Empire Genomics, Buffalo, NY) and a centromere probe for chromosome 6 (labelled with an aqua fluorophore; CEP6, Abbott Molecular, Des Plaines, IL) [19]. Signal analysis was performed in combination with morphology correlation, and at least 100 interphase cells within marked tumor areas were evaluated and imaged using a Zeiss fluorescence microscope coupled with Metasystems ISIS software (Newton, MA). A positive result was determined if more than 10% of cells showed amplification, and >10:1 ratioof RUNX2 signals to the reference centromeric probe.

3.0. Results

Of 111 osteosarcomas profiled using next generation sequencing-based copy number assessment, amplification at 6p12–21 was documented for 24 of 111 cases using probes for VEGFA/CCND3 (21.6%; Table 1). Of these 24 cases, 18 had material available for confirmatory downstream analysis and amplification of RUNX2 was confirmed in 17 (of 18) cases using FISH (Figure 1B, Table 1). No amplification was seen in a single case (Case 24) which had a 1.8-fold copy number gain of VEGFA without a concurrent amplification of CCND3 and it is possible that RUNX2 was not a part of the amplified locus in this case. Five of 24 specimens (20.8%) had alterations of TP53 (cases 3, 6 and 17) and RB1 (cases 16 and 22) and only 2 of these cases had confirmatory testing for RUNX2 amplification using FISH.

Table 1.

RUNX2 Amplification in Osteosarcoma

Case Number	Age (yrs)	Sex	Primary/Metastasis	Histologic Subtype	Treatment Status	Copy Number: VEGFA	Copy Number: CCND3	FISH: RUNX2
1	36	Male	Metastasis	Telangiectatic	Post treatment	2.8	8.9	Amplified
2	22	Male	Primary	Osteoblastic	-	7.4	5.5	Amplified
3	27	Male	Metastasis	Telangiectatic	Post treatment	4.1	7.3	-
4	12	Male	Metastasis	Telangiectatic	Post treatment	2.4	6.3	Amplified
5	14	Female	-	NOS	-	2.5	6.0	-
6	19	Male	Primary	Osteoblastic	-	4.4	4.4	-
7	17	Male	Metastasis	Osteoblastic with giant cell rich component	-	3.6	3.6	Amplified
8	18	Female	Metastasis	Chondroblastic	Post treatment	3.5	3.5	Amplified
9	18	Female	Metastasis	Osteoblastic	Post treatment	3.3	No gain	Amplified
10	22	Male	Primary	Osteoblastic, focally chondroblastic	Post treatment	3.0	No gain	Amplified
11	59	Male	Primary	Giant cell rich with telangiectatic features	-	2.9	2.9	Amplified
12	14	Female	Metastasis	Osteoblastic and chondroblastic	-	2.9	2.9	Amplified
13	19	Male	Metastasis	Osteoblastic	-	2.5	2.5	Amplified
14	11	Male	Primary	Osteoblastic	-	2.1	2.5	-
15	26	Female	Primary	Osteoblastic	-	2.3	2.3	Amplified
16	11	Male	Metastasis	Osteoblastic	-	2.3	2.3	-
17	17	Male	Primary	Fibroblastic	-	2.2	2.2	Amplified
18	8	Female	Primary	Osteoblastic	-	No gain	2.1	-
19	15	Female	Metastasis	Osteoblastic and chondroblastic	Post treatment	1.8	1.8	Amplified
20	19	Male	Metastasis	Osteoblastic	Post treatment	1.7	1.7	Amplified
21	18	Female	Metastasis	Osteoblastic	-	1.7	1.7	Amplified
22	13	Male	Primary	Telangiectatic	-	1.7	1.4	Amplified
23	11	Male	Primary	Osteoblastic	Post treatment	1.6	1.6	Amplified
24	19	Male	Metastasis	Osteoblastic	Post treatment	1.8	1.1	Not Amplified
Summarized Data
24 Cases	Mean: 19.4 (Range: 8–59)	Male: Female- 16:8	Primary: Metastasis- 10:13	Osteoblastic: 12; Chondroblastic: 1; Telangiectatic: 4; Others: 7	Post treatment: 10 cases	Mean: 2.8	Mean: 3.4	Amplified: 17/18

Yrs: years; NOS: not otherwise specified. 17/72 cases previously reported to have 6p21.1 amplifications without assessment of RUNX2 status (Suehara et al¹¹).

RUNX2 (6p21.1) amplifications were observed in a wide range of clinical specimens including primary and metastatic tumors and included multiple (n=10) post-chemotherapy specimens (Table 1). Of note, these amplifications were noted twice as frequently in females compared to males (16:8) and this suggests a gender predilection which needs to be confirmed in future studies.

4.0. Discussion

There is a paucity of functional studies regarding the identification of driver alterations at the 6p12–21 locus. While there is interest in the presence of VEGFA amplification at this locus due to the availability of targeted therapies, it is important to identify other oncogenes at this locus [11]. Although prior studies by Lue et al had suggested RUNX2, CDC5L and CCND3 as important oncogenic drivers at this locus, there have been limited systematic follow up studies [10]. Given the vast body of literature pertaining to the role of RUNX2 in osteoblastic differentiation, the identification of potentially dysregulated RUNX2 expression in these cases is significant because it helps us understand the underlying biology of these tumors [1–4]. For instance, as RUNX2 may directly induce vascular endothelial growth factor gene expression and angiogenesis, the documentation of combined RUNX2/VEGFA amplifications may help us better understand the pathogenic alterations that drive tumor biology in such cases [11, 21–23].

Overall, the estimated frequency of genomic amplification events at the 6p12–21 locus, after taking into consideration both the current study (24 of 111 cases, 21.6%) as well as previous studies by Lau et al (7 of 25 cases, 28%), Lu et al (12 of 48 cases, 25%) and Yang et al (32 of 50 cases, 64%), is likely to be approximately 32.1% (75 of 234 cases) [9, 10, 12]. From our results it could be hypothesized that most of these cases with 6p12–21 amplification have dysregulated expression of the RUNX2 gene which contributes to osteosarcoma pathogenesis. Oncogenes that have been implicated in other tumor types with 6p12–21 amplifications include TFEB/VEGFA in renal cell carcinomas and DEK/E2F3 in retinoblastomas, while the pathogenic role of genes such as CDC5L and CCND3 in osteosarcomas needs to be further defined [10, 24–27].

In addition, it is possible that poorly defined structural rearrangement events at the 6p12–21 locus have a significant role in pathogenesis [9, 24]. Of note, recent studies have suggested that neoantigens generated secondary to gene fusions may be positively correlated with response to immunotherapy [28–30]. A recent study demonstrated the presence of RUNX2 fusions in osteosarcoma in the background of high levels of RUNX2 mRNA transcript expression [28]. On the basis of our current results, we hypothesize that the high levels of RUNX2 expression in these cases may represent osteosarcoma with 6p12–21 amplification. Therefore, the presence of poorly defined structural rearrangement events at the 6p12–21 locus which may drive response to immunotherapy is an area that needs to be further investigated in future studies.

Of note, RUNX2 amplifications were identified in both primary and post-chemotherapy specimens supporting this event being a primary/ de novo alteration rather than a secondary alteration. Similarly, the presence of these amplifications in both primary and metastatic specimens argues against these amplifications being a late event in tumor progression. RUNX2 is thought to have a synergistic effect on cell cycle progression in the presence of TP53 and RB1 alterations [6, 16, 17]. Five of 24 specimens (20.8%) had alterations of TP53 and RB1 and these cases included primary and metastatic tumors as well as post-chemotherapy specimens. Due to the limited number of 6p12–21 amplified cases with co-alterations involving the TP53 and RB1 genes, no definitive conclusions can be drawn regarding their prognostic impact.

Identification of 6p12–21 amplifications as well as candidate oncogenes such as RUNX2 will aid in understanding the biology of osteoblastic differentiation and future studies with larger cohorts are needed to better define the prognostic impact of these genomic alterations. FISH for RUNX2 therefore represents a rapid and widely-available diagnostic modality to identify such cases, while more comprehensive molecular profiling including global copy number analysis is likely to help in better defining the individual genes and alterations present at this amplicon. At present, VEGFA amplifications at this locus may represent an immediate avenue for targeted therapy using anti-angiogenic agents as has been recently proposed [11].

Highlights.

• 6p12–21 amplification is frequently seen in osteosarcoma (24 of 111 cases, 21.6%).

• FISH confirmation of RUNX2 within the amplified locus seen in 17/18 (94.4%) cases.

• RUNX2 is a known transcriptional regulator of osteoblastic differentiation.

• In addition, VEGFA amplification at 6p21 is a potentially targetable alteration.

• RUNX2 FISH is a widely available diagnostic modality to identify such cases.