Medical and health sciences
- Anatomical pathology
- Analysis of next-generation sequence data
- Computational biomodelling and machine learning
Three to five percent of people who are confronted with cancer will be diagnosed with Cancer of Unknown Primary (CUP); a metastasized cancer of which the tissue-of-origin cannot be determined. If possible, a biopsy of the tumor will be taken for pathological examination, but this can take up to two weeks to plan, execute and analyze. Diagnosis usually requires multiple rounds of immunostaining, a slow and costly method that leaves about 34% of CUPs without definite diagnosis. Because treatment options are limited to aspecific chemotherapy, CUP is the fourth most common cause of cancer death. The tumor DNA methylation pattern is a unique ‘fingerprint’ that can be used to determine the tissue-of-origin in CUPs. A novel technique, called cfRRBS, developed at the UGent-VIB allows the identification of the methylation profile starting from minimal amounts of highly fragmented DNA. With this technique not only DNA extracted from paraffin-embedded tumor tissue can be used, but also cell-free DNA isolated from liquid biopsies such as blood. In this research project, cfRRBS will be turned into a clinically applicable protocol. A computational analysis pipeline with reference data sets will be built to classify tumors according to their methylation profile. The processing of patient samples and sequencing data will be optimized to achieve a flexible and reliable classifier. The performance of this classifier will then be prospectively validated as a diagnostic tool for CUPs.