Preview

PDF, English - main document Download (52MB) | Terms of use

Abstract

Chronic lymphocytic leukemia (CLL) is an indolent B-cell malignancy with a very heterogeneous clinical course. Even though many aspects of the biology of CLL have been thoroughly described, the underlying molecular cause for this heterogeneity has still not been completely understood. To fill this gap, this thesis presents a comprehensive analysis of cancer cell-intrinsic and extrinsic factors which modify drug response phenotypes and patient outcome in a cohort of 81 primary CLL patient samples.

Some cancer cell-intrinsic factors, like the genome or transcriptome of CLL, have been comprehensively explored. However, proteomic profiling of a large CLL patient cohort and its integration with other molecular layers is currently lacking. Therefore, this study performed a thorough characterisation of multiple CLL cell-intrinsic factors, including the proteome, the transcriptome and the genome. These were additionally linked to ex-vivo drug response profiles (43 drugs). This revealed associations between the different layers and functional consequences for drug response and clinical outcome. nsupervised clustering of protein levels uncovered a previously unappreciated poor prognosis CLL subgroup, which was independent of established risk factors and characterised by a distinct protein and drug response profile. The existence of this subgroup could be validated in an external cohort. This comprehensive multi-omics analysis represents the first proteogenomic study of a large CLL patient cohort.

CLL cells additionally depend on cell-extrinsic signals provided by the microenvironment, such as the bone marrow niche. Such signals can modify and reduce the activity of selected drugs. However, a systematic analysis of how the bone marrow microenvironment influences drug response and resistance is lacking, because appropriate bone marrow model systems for high-throughput drug screening do currently not exist. To this end, a high-throughput co-culture drug-sensitivity testing platform was established. During the careful evaluation of different stroma cells as CLL cell support for the system, an unexpected phenomenon was discovered. Some bone marrow stroma cells had the ability to phagocytose apoptotic cells in large amounts. Phagocytosis decreased the total amount of cells and, thus, artificially increased the percentage of alive cells. This has implications for co-culture studies in general, as phagocytosis can cause a systematic bias and the misinterpretation of results if left unconsidered. Consequently, nonphagocytic stroma cells were chosen for the final screening platform.

Using this optimised system, responses to 43 different drugs were measured. A linear model was employed to distinguish between the effect of stroma cells on spontaneous and on druginduced apoptosis of CLL cells. In accordance with the literature, stroma cells protected CLL cells from spontaneous apoptosis ex-vivo. Interestingly, effect sizes varied between patients and especially samples with unmutated immunoglobulin heavy chain variable region and high degrees of spontaneous apoptosis profited from co-culturing. Moreover, the influence of stroma cells on drug responses was systematically assessed. While some drugs, like chemotherapeutics, were less active in co-cultures, other drugs had unchanged activity or were even more effcient in the context of stroma cells. Especially Janus kinase inhibitors could overcome the protective effect by stroma cells and kill CLL cells despite the presence of stroma. The systematic analysis of the impact of the bone marrow niche on drug response can help to understand and overcome microenvironment-induced resistances.

In conclusions, this thesis provides a systematic overview of how leukemia cell-intrinsic layers of CLL and the microenvironment determine drug response and patient outcome.