There is a commonly held belief that one of the primary reasons for such slow progress in the fight against cancer is financial. Without adequate funds, the story goes, scientists have not been able to make the positive moves towards new treatments that the ‘war on cancer’ has been promising for decades. The reality is, unfortunately, a lot more complicated. The truth is that there is an awful lot of money invested in cancer research – there are billions of dollars of funding from governments, private foundations, the drugs industry and members of the public donating to the cancer research charities. If it was just a matter of money the fight would have been won by now – or at least be further along than we currently are.
If it’s not just a question of funding, then how can we explain why new treatments are so slow in coming? And why are the new treatments that do appear so often disappointing?
One part of the answer, in my opinion, is that scientists have been focused on the wrong targets. As I have previously written, for example in the articles on How To Read A Cancer Paper, much early stage cancer research is done in vitro. Simply put, it means that researchers test new drugs in test tubes. Strictly speaking they’re not test tubes but flat dishes with a layer of growth medium (food and a nice environment for cells to grow on) on them. These flat dishes – often called Petri dishes – then have a layer of cancer cells grown on them. Mostly these cancer cells come from standardised cells lines that represent a particular type and sub-type of cancer – for example hormone resistant prostate carcinoma, or Her2+ breast cancer. While the distant ancestors of these cells will have come from patients, the cells supplied from the standard cell libraries have been grown in cultures from one generation to the next over many, many years. And this is one aspect of the problem.
Because these cells have been grown in cultures for so long they have adapted to these conditions. They have evolved to suit their surroundings. While they retain enough of their original characteristics to be cancer cells, they will have changed in lots of subtle ways from the cells that came out of a person.
Another key difference is to do with the fact that cancer cells in real people do not live in an environment that is even remotely like that in a test tube. In a body they are in contact with all kinds of other cells – including the cells from the tissues that surround the tumour, immune cells of various descriptions, blood cells and of course the other cells that are packed tight into a solid tumour. This is a long way from a thin layer of cells of the same type sitting on top of a layer of jelly containing nutrients that ensure it’s survival. In the body you have a complex eco-system of cells struggling for survival and competing with other types of tumour cells and with normal cells and the immune system. In the Petri dish you only have a pale imitation of this.
There should be nothing controversial or difficult about accepting any of this – it’s common sense and basic science. But the fact is that the test tube model of cancer has been what many scientists have been working with for decades. This model has been used both to test new therapies and also to increase our knowledge of how cancer operates. It’s used both for theoretical work in understanding cancer and in practical work for testing new drugs. And the problem is that it’s the wrong model. In science models are of crucial importance. With the right model you can get to the essence of a problem. The right model helps clarify thinking and enables you to test your predictions and understanding. But working with the wrong model and you can end up following blind alleys and obscuring the problem rather than clarifying it.
And this problem of working with standard cells lines isn’t just restricted to test tubes.
A better model for cancer is provided by living tissue. In medical research there are numerous animal models being used for in vivo research. Now aside from the rather obvious problem that mice and rats aren’t people, there is also the problem of what cells are used to seed the cancers in these experimental animals.
Cancers that are seeded from the standard cells lines differ in some quite subtle ways from naturally arising cancers. In real cancers, the tumours emerge in particular tissues surrounded by a bed of non-cancerous cells (the so-called tumour stroma), with blood and lymphatic vessels sprouting chaotically around them, attracting waves of immune cells. As the tumours grow the surrounding micro-environment is changed as the tumours release chemical signals and respond in turn to how the body reacts to these changes. There are all kinds of complex feedback mechanisms involved. In contrast when tumour cells from standard cells lines are implanted into animals they miss the initial growth pattern and although they become tumours, they may be missing key aspects of the stroma or tumour microenvironment that evolves with in 'natural' tumour growth.
Add to this the fact that to ensure that the implanted tumours grow and are not rejected, many experimental animals are often immune-compromised and are missing some elements of a natural immune response. For many years it was assumed that tumours were effectively invisible to the immune system, but we now know that this is simply not true. Tumours attract all kinds of immune cells including macrophages, neutrophils and other immune cell types. What happens however is that the immune process is subverted in the end by tumours, but it is also clear that the tumours will have been changed by this contact with the immune system.
So yet again we see that an inaccurate model of cancer can lead to much wasted or erroneous research. Drugs that work fantastically in animal models of cancer can turn out to be disappointing in practice because they've been targeting tumours that are missing important elements of 'natural' tumours.
A common thread in these examples is that the simplistic models that researchers have been using for decades ignore the effects of evolution. It’s hard to understand how it’s possible that one of the most fundamental discoveries in all of science can have been missed, but that is precisely what appears to have happened.
And it’s not just in test tube and animal models that the effects of evolution seem to have been a blind spot. The same thing seems to have happened with people too.
This is borne out in a recently published paper in the New England Journal of Medicine. Entitled Intratumor Heterogeneity and Branched Evolution Revealed by Multiregion Sequencing by M. Gerlinger and others, the authors took multiple tumour samples from four renal cell carcinoma patients and genetically analysed them looking for a wide range of cellular changes. What this revealed was that even within a single tumour sample there were different changes in different parts of the same tumour, let alone when comparing primary tumours and metastases. Simply put, even within a single tumour, there were different mutations taking place in different areas.
In an age when we are moving towards a situation when drug treatment choices are increasingly being made on the basis of the genetic makeup of patient’s tumours, the fact that tumours are sites of such rapid evolutionary processes is cause for concern. Patients may end up missing on beneficial treatments because the part of the tumour that was sampled in the biopsy didn’t express a marker that is actually present in the bulk of it. Or else patients may be treated with toxic drugs that do not effect the bulk of the tumour because the marker only existed in the tiny spot that was sampled.
We don’t even have to wait for the introduction of more ‘personalised treatments’ for this to become a problem. How many patients have suffered because doctors have assumed that a cancer recurrence must have the same genetic makeup as the original tumour?