Title: Cancer evolution and identification of relapse-initiating cells
Christoph KLEIN (Germany) Fraunhofer Projektgruppe ITEM-R
Martin SCHULER (Germany) Department of Medical Oncology, West German Cancer Center, University Hospital Essen
Christoph BOCK (Austria) CeMM Research Center for Molecular Medicine of the Austrian Academy of Sciences, Vienna
Ehud SHAPIRO (Israel) Department of Computer Science & Applied Math and Department of biological chemistry Weizmann Institute of Science, Rehovot
Antonio CHIESI (Estonia) HANSABIOMED Ltd, Tallinn
a. Background and rationale.
Cancer is an evolutionary process where the units of selection, the cancer cells, are constantly reacting and changing in response to environmental and iatrogenic selection pressures. It is therefore not surprising that the previous focus on molecular studies of primary tumours in order to select therapy targets falls short with the need of striking a moving target.
Systemically disseminated cancer cells are often substantially different from the primary tumour. However, it is largely unknown which cells are able to survive systemic therapies and which mechanisms are involved in cell survival and progression to lethal metastasis in patients. Therefore, we need tools to monitor systemic cancer over the course of time. Secondly, to identify relapse-initiating cells, we need to unravel their origin. Are relapse-initiating cells derived from the most aberrant and most distant descendant of the cancerous progeny or are the seeds of relapsing metastases much closer to the root of the cancer cell kindred? Knowing the answer to this fundamental question will enable to select the right markers to monitor systemic cancer and eventually help to administer the most appropriate therapy for the actual stage of the disease.
CEVIR (ceviri, turk: translation) aims to translate an evolutionary understanding of cancer into monitoring assays that guide treatment decisions. Cancer cells and molecules such as nucleic acids can be found within the blood stream over the course of disease and provide valuable information about this evolutionary process. However, major hurdles to exploit blood-borne biomarkers are detection limits for circulating tumour cells (CTCs) or nucleic acids and the differentiation of disease-driving mechanisms from epiphenomena. CEVIR will address both hurdles.
We will improve CTC detection. Then, to monitor cancer evolution, it is essential to know what to monitor. For this, we will genetically determine the relapse initiating cells (RICs) by use of lineage tree analysis, generated from primary tumours, disseminated cancer cells from lymph nodes and CTCs from follow-up samples. Genetic and epigenetic profiling will unravel the molecular features of RICs. This information will then be used to establish monitoring assays, employing DNA mutations and methylation patterns isolated from plasma with and without extravesicular enrichment.
e. Expected results and potential impact.
We will answer whether RICs are derived from late arising rapidly dividing clones of the primary tumour or from rarely-dividing stem-like cells similar to some types of leukaemia and use this information for the development of blood-borne monitoring assays. By tracking RICs back to the earliest seeds of systemic disease, we will also develop monitoring assays for early systemic cancer and explore whether these characteristics are functionally relevant and possibly useful therapy targets.
Our CEVIR consortium brought together experts from different fields of academic and non-academic biomedical research aiming to improved detection of systemic cancer and study the evolutionary processes underlying systemic progression of non-small cell lung cancer (NSCLC). To achieve this goal, we recruited a cohort of advanced (n=52) and early-stage (n=250) NSCLC patients, adopted and optimized workflows for isolation of circulating tumor cells (CTCs) and circulating tumor DNA (ctDNA), as well as conducted a phylogenetic analysis of collected cells.
Isolation of ctDNA was most effective, when conducted on whole plasma samples compared to plasma-derived extracellular vesicles. To this end, the automated Maxwell RSC system proved to be the most reliable providing purified ctDNA in sufficient quantities. To study the profiles of genetic alteration in the ctDNA samples we utilized a commercial targeted NGS panel (Roche Avenio assay) and a custom real-time ddPCR assay. The ddPCR workflow enabled longitudinal study of tumor progression in NSCLC providing results well reflecting dynamics of the primary tumor growth during the course of therapy. Noteworthy, analysis of Roche Avenio datasets revealed presence of a EGFR pT790M mutation (conferring resistance to tyrosine kinase inhibitors) in treatment-naïve ctDNA samples that was not identified during genetic profiling of the corresponding PT specimens. In addition, we developed a MSRE-ddPCR assays allowing for epigenetic profiling of ctDNA allowing for accurate (AUC > 0.9) detection of tumor-associated methylation patterns in ctDNA samples of NSCLC patients. An integrative approach for genetic and epigenetic analysis of ctDNA enabled to monitor tumour progression over time in individual patients. This workflow was established and validated using an independent cohort of 222 ctDNA samples obtained from 73 patients with Ewing Sarcoma enabling to monitor genetic and epigenetic tumor progression in individual patients and can now be applied to NSCLC patients.
Isolation of CTCs was limited by the ability of the utilized methods to detect tumor cells in bloods samples of NSCLC patients. This most likely results from phenotypic heterogeneity of tumor cells in NSCLC and reflect limitations of the detection systems utilized in our study focusing tumor cells exhibiting epithelial phenotype. We optimized and adopted two CTC enrichment methods (i.e. CellSearch and Parsortix) but neither of the two provided significant yields of tumor cells. Fortunately, detection of disseminated tumor cells (DCCs) in lymph nodes and bone marrow proved to be effective yielding sufficient samples for downstream analyses. Phylogenetic analysis of the collected cell collective revealed that DCCs represent an earlier stage of tumor progression than cells derived from the corresponding primary tumors. We will continue to collaborate to further resolve the origin of metastatic founder and relapse-initiating cells. In summary, during the course of Cevir study we established a number of protocols suitable for detection and characterization of cfDNA and laid the ground for a large-longitudinal study to unravel the phylogenetic relationships of progression-driving cancer cells in NSCLC.
(Project funded under JTC 2014)