Cancer – that dreaded and mysterious disease that manifests in so many diverse ways and affects nearly everyone in some way or another. As different as leukemia might seem from prostate cancer or breast cancer, these three and all other cancers are fundamentally genetic disorders. One of the stunning discoveries of the 1970s was the finding that certain retroviruses expressed single proteins that were sufficient to cause infected cells to become cancerous. Even more remarkable was the subsequent finding that these so-called viral oncogenes were actually our own genes captured by these viruses and turned against us. This insight provides an underlying and unifying model that rogue proteins cause cancer.
Our DNA encodes thousands of proteins that interact in a meticulously coordinated manner to mediate all the normal functions of our cells and organs. Mutations in our DNA can cause changes in protein function and/or amount, either of which can lead to disease, such as mutations in the hemoglobin gene causing sickle-cell anemia. When mutations occur in genes that code for certain critical proteins, the aberrant expression of these proteins can convert normal cells into tumor cells. Some people are born with mutations in critical genes that predispose them to cancer development, but for most of us the cancer-initiating mutations occur randomly in our cells as we age. The daunting goal for scientists and clinicians over the last 40 years has been to identify all the critical oncogenic proteins and their functions. Starting with the first human genome sequence that was completed in 2001 , the cost of genomic sequencing has declined and the speed of this sequencing has increased greatly. Because of these technological advances, about 10 years ago the Pan-Cancer Analysis of Whole Genomes (PCAWG) Consortium formed to undertake a massive, worldwide evaluation of mutations in cancers. The group sequenced 2,658 whole genomes of cancer cells, along with matched samples of non-cancerous cells from the same patients. Altogether 38 different types of cancers were included in the study, and the first round of analyses was published this month in 17 papers. There is far too much information to discuss in this blog, but one seminal observation from this first round of analyses concerned so-called driver mutations. These are mutations that provide a selective advantage for the cancer cell over normal cells. Five percent of tumor samples had no detectable driver mutations, indicating that there are still unknown pathways to cancer that need to be identified. The remaining tumors had an average of 4-5 driver mutations that accumulated over years to decades before the cancer diagnosis. This long cumulative process suggests that it may be possible to develop earlier detection methods and even treatment approaches that target these aberrant cells before they become cancerous. It was also intriguing that many cancers have specific mutational signatures that can be used to classify both primary and metastatic cancers. This refined genetic analysis may someday be useful for tailoring individually-adapted therapies after sequencing the patient’s tumor cells. While there is much more information to be mined from the data assembled by the consortium, this first round of studies is a remarkable achievement that greatly advances our understanding of cancer genetics.