Send to

Choose Destination
Br J Cancer. 2016 Oct 11;115(8):1000-1007. doi: 10.1038/bjc.2016.280. Epub 2016 Sep 8.

Gene expression profiling identifies distinct molecular subgroups of leiomyosarcoma with clinical relevance.

Author information

Institute of Cancer Research, 15 Cotswold Road, Belmont, Sutton, London SM2 5NG, UK.
Department of Pharmacology and Pharmacy, The University of Hong Kong, Hong Kong, China.
Department of Pathology, The Royal Orthopaedic Hospital NHS Trust, Birmingham B31 2AP, UK.
Sarcoma Unit, The Royal Marsden Hospital NHS Trust, Fulham Road, London SW3 6JJ, UK.



Soft tissue sarcomas are heterogeneous and a major complication in their management is that the existing classification scheme is not definitive and is still evolving. Leiomyosarcomas, a major histologic category of soft tissue sarcomas, are malignant tumours displaying smooth muscle differentiation. Although defined as a single group, they exhibit a wide range of clinical behaviour. We aimed to carry out molecular classification to identify new molecular subgroups with clinical relevance.


We used gene expression profiling on 20 extra-uterine leiomyosarcomas and cross-study analyses for molecular classification of leiomyosarcomas. Clinical significance of the subgroupings was investigated.


We have identified two distinct molecular subgroups of leiomyosarcomas. One group was characterised by high expression of 26 genes that included many genes from the sub-classification gene cluster proposed by Nielsen et al. These sub-classification genes include genes that have importance structurally, as well as in cell signalling. Notably, we found a statistically significant association of the subgroupings with tumour grade. Further refinement led to a group of 15 genes that could recapitulate the tumour subgroupings in our data set and in a second independent sarcoma set. Remarkably, cross-study analyses suggested that these molecular subgroups could be found in four independent data sets, providing strong support for their existence.


Our study strongly supported the existence of distinct leiomyosarcoma molecular subgroups, which have clinical association with tumour grade. Our findings will aid in advancing the classification of leiomyosarcomas and lead to more individualised and better management of the disease.

[Indexed for MEDLINE]
Free PMC Article

Supplemental Content

Full text links

Icon for Nature Publishing Group Icon for PubMed Central
Loading ...
Support Center