Predicting evolutionary futures

Photo credit: Living in Monrovia / Foter.com / CC BY-SA

Photo credit: Living in Monrovia / Foter.com / CC BY-SA

One of the striking achievements of cancer genomics and its allied bioinformatics has been to construct phylogenetic trees depicting the trajectories of sub-clones in cancers and their ancestral relationships. It’s like taking a peek back in time at the origin and prior evolutionary history of the malignancy.

But what about the converse? Is it possible to infer, from features of cancer cells, what their future potential or ability to evolve into more malignant, metastatic or drug-resistant phenotypes may be? There’s no doubt this would be extremely useful, particularly in the context of early diagnosis and intervention.

The view that early intervention in cancer increases the prospects for cure or control, is generally accepted; but there are also difficulties here. The majority of tumours evolve very slowly or even stop progressing, so may never become life-threatening 1. Coupled to this, we haven’t had the tools to predict or distinguish the tumours that are the bad players. If cancer surveillance in a population of seemingly healthy individuals were to pick up substantial numbers of sub-clinical or covert tumours, as it might well do, what should we then do about it, given that most may be, in effect, false positives? It’s a problem that Cancer Research UK have recognised as one of their designated seven ‘Grand Challenges’.

A ‘wait and see’ philosophy might seem reasonable but then, once malignancy does evolve, it could be too late. Early surgical intervention might be prudent but this is invasive and can come at a cost in collateral damage. Clearly there is a need for better prognostic markers.

But can evolution be predictable?

‘Evolution has no eyes to the future’
George Williams, 1966

All evolution of individual organisms, including cancer cells, rests on random shuffling within the genome and the natural selection of those variants that, serendipitously, already have the best-adapted phenotypes. Their ‘fitness’ or ‘adaptation’ is contingent upon ecological circumstances ‘of the moment’ and the prevailing, local selective pressures. And these can and do change. Winners can become losers.

Evaluating the ‘fitness’ of tumour cells might tell us how they came to be where they are as a dominant sub-clone in a diagnostic sample but probably little about their future prospects. The cells that cause havoc in metastasis or drug-resistant recurrence are mostly very rare entities in the primary tumours 2.

In the past, we’ve tried to prognosticate using specific genetic lesions or angiogenic phenotypes that are known to be associated with poor clinical outcome. But these tactics have limited accuracy and there is significant variation in outcome within sub-groups so identified.

Given that evolution is fuelled by ‘driver’ mutations distilled from essentially random genetic noise and therefore imbued with chance, you might imagine that it is essentially unpredictable. The evolutionary biologist Stephen Jay Gould once said that if we ran the tape of all life again, it would come up with very different answers. But is this really the case? New evolutionary innovations are not created from a blank slate but by tweaking what already exists. Hence starting conditions in a cancer may influence likely trajectories. Some tumours may be born to be bad 3. There are also limits on what is likely to impart fitness within an ecosystem. Evolution is, for this reason, frequently convergent for both organisms and cells, coming up with the same fitness trick many times independently 4.

So maybe evolution of cancers towards a more malignant status is, to some extent at least, predictable.

Simpler evolutionary systems lend themselves both to predicting and to steering future trajectories in a probabilistic fashion, for example, with respect to the emergence of antibiotic resistance in bacteria 5.

It is certainly more complicated in cancer but something similar should be possible.

In principle, the probabilistic likelihood of the progression of disease or mutation-based drug resistance should be determined by the type of ‘substrate’ available for selection (as in bacteria with respect to antibiotic resistance). I’ve argued that the critical cellular substrate or unit of evolutionary selection in cancer is the population of cells with the potential for self-renewal, i.e. the stem cells 6. The size of the pool of stem cells in cancer can be extremely variable, but that parameter plus the inherent genetic diversity within the pool should provide a quantitative measure for the substrate available for selection.

To date, these are parameters that have been evaluated separately. The frequency or pool size of cancer stem cells (measured by limiting dilution xenotransplantation or quantitative gene expression) does seem to be associated with progression or clinical outcome in acute leukaemia, colorectal cancer and breast cancer 6 .

The extent of genetic diversity of whole tumour populations will probably mirror the underlying diversity of stem cells whose activity generates the clones making up that tumour population. Measures of genetic diversity in oesophageal cancer, lung cancer, breast cancer, paediatric nephroblastoma and acute leukaemia are also significantly associated with progression of disease or clinical outcome 6.

Other evolutionary parameters may also have predictive value. Theoretical considerations suggest that a heterogeneous ecosystem should encourage more diversification and therefore more potential for progression 7. Some measures of micro-environmental diversity in cancer do appear to support this contention 8. Components of ecosystem diversity might even be discernible via whole body imaging and that would provide practical possibilities for screening for prognostic purposes 9.

Once we make a start on this evolutionary prognostication and are able to test its practical value, the most ambitious but exciting task will then be to manipulate or steer the evolutionary futures in cancer clones towards lower fitness for malignancy. This should be achievable through manipulation of ecosystem pressures or, in the context of drug resistance, by appropriately selected drug schedules 10, 11.

Mel

References

  1. Greaves M (2014) Does everyone develop covert cancer? Nat Rev Cancer, 14: 209-210.

  2. Klein CA (2013) Selection and adaptation during metastatic cancer progression. Nature, 501: 365-372.

  3. Sottoriva A, Kang H, Ma Z, Graham TA, Salomon MP, Zhao J, Marjoram P, Siegmund K, Press MF, Shibata D, Curtis C (2015) A Big Bang model of human colorectal tumor growth. Nat Genet, 47: 209-216.

  4. Conway Morris S (2003) Life’s solution: inevitable humans in a lonely universe. Cambridge Univ Press.

  5. Palmer AC, Kishony R (2013) Understanding, predicting and manipulating the genotypic evolution of antibiotic resistance. Nat Rev Genet, 14: 243-248.

  6. Greaves M (2015) Evolutionary determinants of cancer. Cancer Disc, 5: 806-820.

  7. Anderson ARA, Weaver AM, Cummings PT, Quaranta V (2006) Tumor morphology and phenotypic evolution driven by selective pressure from the microenvironment. Cell, 127: 905-915.

  8. Yuan Y, Failmezger H, Rueda OM, Ali HR, Graf S, Chin SF, Schwarz RF, Curtis C, Dunning MJ, Bardwell H, Johnson N, Doyle S, Turashvili G, Provenzano E, Aparicio S, Caldas C, Markowetz F (2012) Quantitative image analysis of cellular heterogeneity in breast tumors complements genomic profiling. Sci Transl Med, 4: 157ra143.

  9. Win T, Miles KA, Janes SM, Ganeshan B, Shastry M, Endozo R, Meagher M, Shortman RI, Wan S, Kayani I, Ell PJ, Groves AM (2013) Tumor heterogeneity and permeability as measured on the CT component of PET/CT predict survival in patients with non-small cell lung cancer. Clin Cancer Res, 19: 3591-3599.

  10. Nichol D, Jeavons P, Fletcher AG, et al (2015) Steering evolution with sequential therapy to prevent the emergence of bacterial antibiotic resistance. PLoS Comput Biol, 11: e1004493.

  11. Chmielecki J, Foo J, Oxnard GR, Hutchinson K, Ohashi K, Somwar R, Wang L, Amato KR, Arcila M, Sos ML, Socci ND, Viale A, de Stanchina E, Ginsberg MS, Thomas RK, Kris MG, Akira I, Ladanyi M, Miller VA, Michor F, Pao W (2011) Optimization of dosing for EGFR-mutant non-small cell lung cancer with evolutionary cancer modelling. Sci Transl Med, 3: 90ra59.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s