Involvement of MET/TWIST/APC combination or the potential role of ossification factors in pediatric high-grade osteosarcoma oncogenesis

Abstract

Dysregulated cell growth or differentiation due to misexpression of developmental critical factors seems to be a decisive event in oncogenesis. As osteosarcomas are histologically defined by malignant osteoblasts producing an osteoid component, we prospected in pediatric osteosarcomas treated with OS94 protocol the genomic status of several genes implied in ossification processes. In 91 osteosarcoma cases, we focused on the analysis of the fibroblast growth factor receptors (FGFRs) TWIST, APC, and MET by allelotyping, real-time quantitative polymerase chain reaction, gene sequencing, and protein polymorphism study. Our study supports the frequent role of TWIST, APC, and MET as osteosarcoma markers (50%, 62%, and 50%, respectively). TWIST and MET were mainly found to be deleted, and no additional APC mutation was identified. Surprisingly, FGFRs are abnormal in only < 30%. Most of these factors and their abnormalities seem to be linked more or less to one clinical subgroup, but the most significant correlation is the link of MET, TWIST, and APC abnormalities to a worse outcome and their combination within abnormal tumors. A wider cohort is mandatory to define more robust molecular conclusions, but these results are to be considered as the beginning of a more accurate basis for diagnosis, in search of targeted therapies, and to further characterize prognostic markers.

Keywords: APC; MET; Pediatric osteosarcoma; TWIST; osteogenesis.

Figure 1

Gene status within TWIST subgroups and statistical combinations. For each gene, we considered the proportion of normal and abnormal subpopulations in each TWIST subgroup (i.e., normal TWIST, TWIST deletion, and TWIST amplification patients). Black circles underline the main associations.

Figure 2

Gene status within MET subgroups. A combination of MET subpopulations (normal MET gene, MET deletion, and MET amplification) was made, and the percentages of normal and altered populations for each gene are given in each graph. Black circles underline the main associations.

Figure 3

EFS correlations considering TWIST (A), MET (B), and APC (C) molecular abnormalities. A statistical correlation for TWIST and MET analyses is found between a worse outcome and rearranged populations, and only a tendency for APC study.

Figure 4

OS correlations for TWIST (A), MET (B), and APC (C) molecular abnormalities. A significant correlation is validated between TWIST and APC alterations and a worse OS, and only a tendency is obtained for both MET deletion and amplification.