The largest ever study of cancer genetics, involving more than 1300 researchers across four continents, has catalogued a trove of cancer mutations in a vast, open access, computer cloud database.
The Pan-Cancer Analysis of Whole Genomes Consortium (PCAWG) gathered its findings meticulously over more than a decade to inform precision medicine, which aims to tailor treatment to a person’s specific cancer.
The results are detailed in six articles published in a special edition of the journal Nature.
“The entire Pan-Cancer work is helping to answer a long-standing medical difficulty: why two patients with what appears to be the same cancer can respond differently to the same drug,” says senior author Peter Campbell from the Wellcome Trust Sanger Institute in the UK.
“We show that the reasons for these different effects of treatment are written in the DNA,” he says.
The group used whole genome sequencing to tease out the unique genetics of tumours, analysing DNA from more than 2000 tumours of 38 types, including breast, brain, liver, prostate, lung and pancreas.
The researchers explain how the ruinous course of those cancers – the disease kills nearly 10 million a year and is the second biggest cause of death globally – follows the blueprint of Darwinian evolution.
Mutations arise in the genes of cells that confer a survival advantage, allowing those cells to replicate faster than their neighbours, breach tissue boundaries, and dodge the immune cells that try to wipe them out.
Mutations, the team says, come in two types. “Passenger” mutations are just along for the ride and don’t propel a tumour. “Driver” mutations, however, fuel the devastating fitness of cancer cells that enables their spread through the body.
Rooting out driver mutations and developing treatments that target them is the main game in cancer therapy – but where to look in a genome built from around three billion base pairs?
One finding of the consortium is that most driver mutations sit in the 2% of the genome that codes for proteins, the so-called “exome”.
“To us, this was an unexpected and important result,” says co-author Iñigo Martincorena, from the Wellcome Sanger Institute.
“For cancer patients, this means that the vast majority of clinically-relevant mutations in a cancer are likely to be found in protein-coding sequences, which will simplify efforts for the clinical use of genome sequencing in cancer,” she says.
That result could only happen, of course, after an exhaustive scouring of the remaining 98% of the cancer genome, the non-coding region sometimes dubbed “dark matter”.
The magnitude of that task was summed up in a perspective piece by Marcin Cieslik and Arul Chinnaiyan, from the Rogel Cancer Centre at the University of Michigan in the US.
“This required hundreds of terabytes of data, spread across multiple data centres, and probably millions of processing hours — all facilitated by cloud computing,” they write.
The group also uncovered a host of previously unknown mutations, including insertions or deletions of genetic code – some caused by exposure to UV light or tobacco – that act as the DNA fingerprint of a tumour.
“Some types of these DNA fingerprints, or mutational signatures, reflect how the cancer could respond to drugs,” says Steven Rozen, a senior author from Duke-NUS Medical School in Singapore.
“Further research into this could help to diagnose some cancers and what drugs they might respond to,” he says.
The research also established a critical timeline for when driver mutations happen, a kind of biological “carbon dating” of a cancer’s evolution.
“Overall, the group’s findings suggest that driver mutations can occur years before cancer is diagnosed, which has implications for early detection and biomarker development,” write Cieslik and Chinnaiyan.
They do, however, say the research is limited in one pivotal respect. The teams lacked access to clinical outcome data, so couldn’t link genetic findings with real world patient results.
Other research is, however, on to that.
The International Cancer Genome Consortium–Accelerate Research in Genomic Oncology (ICGC–ARGO) aims to link more than 100,000 cancer genomes with data on how patients actually fare in the clinic.
One of the more tectonic aspects of this research is to question the relevance of a cancer’s tissue type – for example, skin, brain or lung – as knowledge of its genetics grows.
“Cancer is a genetic disease, and the type of mutations is often more important than where the cancer originates in the body,” says co-author Joachim Weischenfeldt, from the University of Copenhagen in Denmark.
“For example, we may have a type of breast cancer and prostate cancer where the driver mutations are similar. This means that the patient with prostate cancer may benefit from the same treatment as the one you would give the breast cancer patient, because the two types share an important driver mutation,” he says.
In 2012 there were 14 million new cancer cases worldwide, a number expected to exceed 23 million by 2030. Logging its genetic fingerprint will be a massive task, but Campbell remains optimistic.
“The genome of each patient’s cancer is unique, but there are a finite set of recurring patterns in the DNA, so with large enough studies we can identify all these patterns to optimise cancer diagnosis and treatment,” he says.