Similarity and diversity of the tumor microenvironment in multiple metastases: critical implications for overall and progression-free survival of high-grade serous ovarian cancer
Metrics: PDF 1672 views | HTML 3358 views | ?
Andreas Heindl1,2,3, Chunyan Lan4,5, Daniel Nava Rodrigues6, Konrad Koelble3,7 and Yinyin Yuan1,2,3
1 Centre for Evolution and Cancer, The Institute of Cancer Research, London, UK
2 Centre for Molecular Pathology, Royal Marsden Hospital, London, UK
3 Division of Molecular Pathology, The Institute of Cancer Research, London, UK
4 Department of Gynecologic Oncology, Sun Yat-sen University Cancer Centre, Guangzhou, China
5 State Key Laboratory of Oncology in South China, Collaborative Innovation Centre for Cancer Medicine, Guangzhou, China
6 Division of Cancer Therapeutics, The Institute of Cancer Research, London, UK
7 Department of Histopathology, Royal Marsden Hospital, London, UK
Yinyin Yuan, email:
Keywords: high-grade serous ovarian cancer, locally advanced disease, tumor microenvironment, automated image analysis, ecological diversity
Received: March 24, 2016 Accepted: August 24, 2016 Published: September 19, 2016
The tumor microenvironment is pivotal in influencing cancer progression and metastasis. Different cells co-exist with high spatial diversity within a patient, yet their combinatorial effects are poorly understood. We investigate the similarity of the tumor microenvironment of 192 local metastatic lesions in 61 ovarian cancer patients. An ecologically inspired measure of microenvironmental diversity derived from multiple metastasis sites is correlated with clinicopathological characteristics and prognostic outcome. We demonstrate a high accuracy of our automated analysis across multiple sites. A low level of similarity in microenvironmental composition is observed between ovary tumor and corresponding local metastases (stromal ratio r = 0.30, lymphocyte ratio r = 0.37). We identify a new measure of microenvironmental diversity derived from Shannon entropy that is highly predictive of poor overall (p = 0.002, HR = 3.18, 95% CI = 1.51-6.68) and progression-free survival (p = 0.0036, HR = 2.83, 95% CI = 1.41-5.7), independent of and stronger than clinical variables, subtype stratifications based on single cell types alone and number of sites. Although stromal influence in ovary tumors is known to have significant clinical implications, our findings reveal an even stronger impact orchestrated by diverse cell types. Quantitative histology-based measures can further enable objective selection of patients who are in urgent need of new therapeutic strategies such as combinatorial treatments targeting heterogeneous tumor microenvironment.
All site content, except where otherwise noted, is licensed under a Creative Commons Attribution 3.0 License.