Better treatment selection through biomarker discovery

Accurate treatment selection in metastatic prostate patients is crucial to control cancer spread and to minimize patient symptoms and drug side effects. The lab of prof. Ost at Ghent University Hospital was interested in developing a minimally invasive, blood-based test to select the correct treatment strategy. To this end, Biogazelle performed a miRNA biomarker discovery study on serum samples using its small RNA sequencing workflow optimized for body fluids.

The discovery phase identified several candidate miRNA biomarker signatures, which were subsequently confirmed by Biogazelle using RT-qPCR technology. The signatures are currently being validated in a larger, independent patient cohort. The next step aims to implement this test in the clinic, optimizing treatment selection and therefore improving patients’ quality of life.

"Biogazelle has offered a complete project management pipeline for biomarker discovery and validation with frequent go / no-go decisions. In addition, they offer advanced data-analysis envisaging clinical applications of the discovered biomarkers. A transparent communication allowed for easy follow-up of the project."

- prof. dr. Piet Ost, radiation oncologist, Ghent University Hospital, Belgium

Biomarker driven disease status determination

Gut inflammation is commonly present in arthritis patients, linked to extensive disease and representing an important risk factor for the development of Crohn’s disease. Gut inflammation is evaluated through colonoscopy, a disturbing test for patients. Seeking for alternatives to reduce the number of colonoscopies, we supported the research of prof. Elewaut towards the development of a blood-based test that can predict gut inflammation in arthritis patients.

Biogazelle’s contribution included the profiling of a large panel of miRNA biomarker candidates in serum samples from arthritis patients. A biomarker signature was identified and validated in a second, independent patient cohort, paving the path towards the development of an RT-qPCR test to be implemented in the clinic.

Mode of action analysis of drug treatments

Whole genome transcriptome profiling is a valuable tool for the understanding of drug mode of action. Identifying drug targets and affected pathways upon perturbation or treatment and assigning functions to specific genes are, amongst other, questions that can be tackled through transcriptome analysis.

Biogazelle is a pioneer in the study of long non-coding RNAs. While more than 60 000 lncRNAs have been identified in the human genome, just a tiny fraction of them has been functionally characterized to date. With the aim of establishing a catalog of human lncRNA functions, we performed a systematic study where we analyzed the transcriptome of cancer cells upon treatment with 90 different chemical compounds or 90 siRNAs targeting transcription factors. In total, over 200 samples underwent total RNA sequencing, providing comprehensive information on both protein coding and lncRNA expression levels.

The generated data, and its integration with publicly available datasets such as The Cancer Genome Atlas (TCGA) has spurred the development of LNCarta, a proprietary catalog of human lncRNA functions. Further, perturbation and pathway specific lncRNAs were catalogued and used as one of the integrated information layers for therapeutic target selection in Biogazelle’s drug development programs (antisense therapies against oncogenic lncRNAs in lung, liver and colon cancer).

Disease characterization

Colorectal cancer (CRC) is one of most deadly cancer types. The implementation of more precise treatment strategies is needed to help reduce the mortality rates. To this end, several attempts for gene expression-based patient classification are pursued.

Biogazelle contributed in a large multicentric study led by prof. Tejpar (University Hospital Leuven), in which 700 CRC patient tumor biopsies were analyzed for the expression of panel of more than 900 miRNAs. Based on the miRNA expression profiles, 4 consensus molecular subtypes with distinguishing features were identified. This is in line with results obtained for The Cancer Genome Atlas data (Guinnei et al., Nature Methods, 2015) and provides the most robust classification system for CRC to date.