EpiMet-OS: EMT Diagnostics in Oncology Scoring
The developmental conserved mechanism, epithelial-mesenchymal transition (EMT), has been shown to mediate carcinoma progression, but it is still unclear if EMT can have real translational and clinical value in cancer management. We established an EpiMet programme dedicated to translating the EMT mechanism into cancer diagnostics and therapeutics.
Large-scale analysis of gene expression profiles on 15 cancer types (13,000+ samples) identified a universal generic EMT Signature allowing quantitative scoring for cancer samples. The resulting EMT score further predicts overall survival in ovarian, gastric, pancreatic cancers, and glioblastoma; this is aimed at developing a cost-effective and robust prognostic system for personalized cancer treatment.
Molecular Subtype-Specific Ovarian Cancer Management
Cancer management has entered the era for precision management. A promising approach is the clustering of patients based on the gene expression profiling. Ovarian cancer is known to be molecularly heterogeneous with several gene expression molecular subtypes (GEMS) reported; we identified 5 molecular subtypes that showed clinical survival significance; among these the Stem-A subtype which confers the worst survival outcome as an independent prognostic factor.
In vivo data demonstrated sensitivity of platinum-refractory Stem-A patients being sensitive to vinorelbine, leading to an international multicenter Phase II single-arm clinical trial initiated by the Gynaecologic Oncology Group Singapore (GOGS) and Australian New Zealand Gynaecologic Oncology Group (ANZGOG), the first in the world to prospectively stratify platinum-resistant OvCa patients by GEMS followed by Stem-A patients receiving vinorelbine. Pipelines for novel therapeutic targets for other GEMS are currently in development.
Haptoglobin as biomarker for ovarian cancer
(Epithelial ovarian cancer (EOC) is the second most common gynaecological malignancy in Singapore. Most patients are diagnosed at advanced stage as there are no clinical symptoms presented in the early stage of this disease. This provides a key challenge in finding effective screening tests for this ‘silent killer’. However, ovarian cysts are commonly present in women of all ages; therefore there is a need to identify and discriminate between benign and malignant cysts at surgery for appropriate care. Intra-operative diagnosis of an ovarian cyst has an impact on prognosis, fertility preservation and the need for additional surgery for these patients.
The current practice in evaluating ovarian cysts during surgery entails the need for frozen section examination of the cyst wall. This process is time-consuming, laborious, expensive and requires a highly trained pathologist on standby. This histopathological service is also scarce in certain developing nations resulting in a lack of intra-operative diagnosis.
Ovarian cystic fluid can be collected intra-operatively through either laparoscopic or open surgery. Over the years our group has identified haptoglobin as a protein marker in the differentiation of benign and malignant epithelial ovarian tumours. This novel bench to bed side finding has enabled us to develop a rapid diagnostic kit OvaCisTm that measures haptoglobin using a rapid colorimetric assay. Leveraging on our preliminary clinical data we hypothesize that OvaCisTm is able to detect malignancy of the ovarian cyst as accurately as the standard frozen section. This could be an alternative cost-effective and rapid method of detection at a typical operative setting for patients undergoing ovarian cystectomy.
Our current study will enable us to clinically validate OvaCisTM as a rapid and accurate diagnostic kit in predicting ovarian malignancy in cyst fluid versus frozen section of the tumor tissue in a hospital setting. These novel findings will help us provide a cost-effective and rapid diagnosis in centres where frozen sections are not readily available as well as a cost-effective alternative option in other tertiary hospitals.