On the heterogeneity of tumor sequencing

Abstract

Cancer is caused by damage to a cells’ DNA. This damage can be as small as a change or deletion of one single DNA base, as big as duplications or deletions of a full chromosome, or more complex such as erroneous reassembly of broken chromosomes. All these events can lead to unbridled cell proliferation and defective cell repair mechanisms: a tumor cell. Because of DNA sequencing, we now know several genes that are frequently affected by DNA damage in cancer, and also know the effect of such events on the cell. This has led to the development of so called ‘targeted treatments’, treatments that are specifically aimed at the defective gene. By applying these types of treatment, we can fight the tumor more effectively because they block the exact processes that make a specific cell cancerous.

Unfortunately, targeted treatment has not been as successful as expected, especially not for patients with metastatic disease. In many cases, patients initially respond really well to treatment, but their tumors become resistant after some time, or a subset of their tumors does respond to therapy while other lesions continue to grow. This is likely due to genetic differences between tumors within a single patient, or even between cells within the same tumor: tumor heterogeneity. As a tumor continues to grow, the DNA damage builds up. At each cell division, the damaged DNA is copied to the new cell, and additional defects can occur on top of that which finally leads to large differences between cells. Obviously, these genetic differences will cause the cells to behave differently as well.

In this thesis, we investigate the extent of tumor heterogeneity in patients with advanced metastatic disease, by comprehensively characterizing samples from multiple tumor sites or multiple time points. We combine several techniques to capture the various types of DNA damage, and aim to unravel the effect of the genetic changes on the behavior of the tumor by adding gene expression analysis or therapy response measurements. We also present a technique to extract more information from a single sequencing experiment, which increases the efficiency of routine screening for targetable genetic defects. Our results demonstrate the progress that has been made in the past few years, both on the technical and the clinical front, but also reveal the challenges that still lay ahead on the path to stop cancer.