Câncer como uma corrida evolucionária

Estou para iniciar uma orientação sobre modelagem de câncer e tratamentos. Talvez o Mairon deveria dar uma olhada neste artigo:

Evolution Could Explain Cancer Relapse Mystery

Posted: 23 Jul 2009 09:10 PM PDT

Evolution may explain why treated tumors sometimes spread more aggressively than untreated ones.

Most cancer therapies work by distinguishing between normal cells and cancerous ones. For example, radiation therapy distinguishes between tumour cells that are unable to repair themselves and healthy cells which maintain this capability. Chemotherapy is more harmful to tumour cells because they replicate more frequently and so on.

But perhaps there's a cleverer way to target tumours by thinking about them in a different way. Now Branislav Brutovsky at P. J. Safarik University in Slovakia and his pal Dragos Horvath outline just such a new way of thinking about cancer and say that it could have profound implications for cancer treatment. The key is to treat tumour growth as an evolutionary process and to think of therapy as a way of decreasing the evolutionary efficiency to an extent that the growth cannot survive.

Here's the background. According to Brutovsky and Horvath, the process of tumour growth is one of evolutionary optimisation in which competing cells replicate with errors and in numbers that depend on their fitness qualities. Various research seems to indicate that certain cancerous tumours consist of cells with a huge variation in genomes. Some researchers have gone as far as to say that there is no typical genotype for breast and renal cancers. It's this kind of evidence that suggests that cancer is somehow solving an evolutionary optimisation problem.

That immediately points to different way of tackling tumours. Instead of distinguishing between healthy and cancerous cells as conventional treatments do, the evolutionary view suggests that tumours can be inhibited by changing the evolutionary attributes of the entire system. Brutovsky and Horvath suggest strategies such as reducing mutation rates, reducing the effective population size and increasing the generation time of the self-renewing cells. Another idea is to reduce the relative fitness of cancerous cells.

Those all sound like sensible suggestions but the more difficult task is to translate this into real world therapies. That's much more difficult because it has hard to measure or model the evolutionary process at work. What exactly would the fitness landscape look like for cancerous cells and how might it vary with time? Questions like this are likely to remain puzzles for some time.

But the new thinking does offer some interesting insight. Cancer specialists have long noticed that cancer cells that survive therapy are often more aggressive than those that were killed off and that the regrowth of treated tumours is much faster than untreated ones.

Brutovsky and Horvath point out that when evolutionary algorithms are used to solve problems, engineers often remove a certain population of solutions to accelerate the convergence towards a specific goal. This may be exactly what is happening in therapies that do not remove a decisive population of cells. Instead it may just simplify the fitness landscape in away that allows the surviving population to thrive.

As Brutovsky and Horvath put it: "A paradoxical consequence of the optimization facet of carcinogenesis is that applied anti-cancer therapy can be, in fact, the causative factor of its acceleration on longer time scales."

Sobering stuff.

Ref: arxiv.org/abs/0907.2004: Optimization Aspects of Carcinogenesis

Comentários

Postagens mais visitadas deste blog

O SEMCIÊNCIA mudou de casa

Aborto: um passo por vez

Wormholes