The Smaller Picture

It is often thought that cancer cells follow a Darwinian model of evolution – but a recent study challenges this view. Lead study author (1) Chung-I Wu and colleagues decided to take a look at the smaller picture: studying just one small tumor (slightly smaller in size than a ping-pong ball) derived from a hepatocellular carcinoma. They found that it contained over 100 million mutations within coding regions of its DNA – a much higher mutation rate than they expected. We spoke with Wu about the implications of this discovery…

Why study just one tumor?

Every population geneticist’s first question about any population is “how much genetic diversity is there?” and “how is the diversity distributed?” The amount and distribution of diversity informs us about the demographic history of the population (how big it is, how much it has been growing/declining, how much subdivision there is, and so on), the influence of natural selection, and the possible evolutionary trajectory of the population in the future.

The cancer cells within just one tumor represent a very interesting population, and we want to study the evolution of each case of cancer with the same rigor as if we were studying one single species. This is in sharp contrast to the prevailing strategy in cancer biology, which advocates studying large numbers of cases with minimal depth for each case. We plan to study no more than 10 cases in the years to come, with very extensive sampling within and between tumors of the same patient.

What did this first study tell you?

First, the diversity within a tumor accumulates in a non-Darwinian mode. In cancer biology, it is almost always assumed that Darwinian selection dominates cellular evolution. This thinking echoes evolutionary biology 50 years ago. But in 1968, the advent of the neutral theory of molecular evolution changed this view drastically. Much of the evolution within natural populations is perceived as neutral, and not driven by Darwinian selection. Our key finding is that non-Darwinian evolution appears to be as dominant at the cellular level as it is outdoors in natural populations.

Second, the level of genetic diversity is orders of magnitude higher under non-Darwinian than under Darwinian evolution. This is intuitively understandable when we look at cancer cells, as Darwinian selection would throw out bad mutations and keep a very small number of the best mutations, whereas almost every mutation is kept under the non-Darwinian mode (Figure 1).

What could your findings mean for oncology?

The near certainty of drug resistance is the inevitable consequence of non-Darwinian evolution when there are hundreds of millions of coding region mutations. High genetic diversity is seen even when the tumors are microscopic. The medical implication is this: if complete eradication is untenable, what would be the rational strategy to control and contain the growth of tumors? There are many options, but acceptance of the pre-existing resistance is necessary before such options can be evaluated. An immediate question to look into is finding the best treatment strategy right after surgery, when the tumors remaining are microscopic, but also highly genetically diverse.