Skip to main content

Together we are beating cancer

Donate now
  • Health & Medicine

Improving the detection of prostate cancer

by Henry Scowcroft | Analysis

13 February 2009

0 comments 0 comments

Tiger or pussycat? New research could lead to a test to identify aggressive prostate cancers

Tiger or pussy cat? New research could lead to a test to identify aggressive prostate cancers

A paper published in Nature this week by a team of US researchers led to media reports about a new ‘urine test for prostate cancer‘ that can supposedly discriminate between aggressive and non-aggressive cancers.

This is a bit of a ‘holy grail’ in the field of prostate cancer diagnosis.

Prostate cancer, the fourth most common cancer in the UK, affects thousands of men. 34,000 were diagnosed with the disease in 2005 alone (the most recent year for which statistics are available).

But not all men who are diagnosed with prostate cancer will need treatment.

This is partly because the disease seems to exist in two forms – a fast-growing, aggressive form that may benefit more from treatment, and a slow-growing form that can require very little treatment other that regular monitoring.

The current test for prostate cancer – the PSA test – can’t tell the difference between the two types.

So if scientists can develop a non-invasive test that can tell the difference between the two it would be a huge advance, because it could spare many men the distressing experience and side effects of being tested and treated for prostate cancer.


The PSA blood test is used widely around the world and measures the levels of a protein called prostate-specific antigen.

In England, men can request a PSA test from their GP, who will discuss the pros and cons of the test (and its outcome). 

And men who experience symptoms such as difficulty passing urine, increased frequency of urination, or lower back pain will also be offered a PSA test, or a direct examination of their prostate.

Of course, these symptoms are, more often than not, due to something other than prostate cancer – so the PSA test is more a ‘useful indicator’ of whether a man should be referred to a specialist.

But is has also been very controversial. Men with high levels of the protein in their blood are ‘likely’ – not ‘certain’ – to have prostate cancer. And some men with low levels may actually have the disease.

This uncertainty has meant that many have challenged the wisdom of using PSA tests more widely, for example as a national screening programme.

Certainly, since the widespread introduction of the PSA test, prostate cancer rates have risen.

And as we’ve said, the PSA test can’t tell the difference between slow-growing cancers and aggressive ones. This is a serious criticism since it can lead to over-treatment for otherwise harmless cancers, and a lot of invasive biopsies that carries risks of incontinence and impotence.

So naturally, there’s a huge amount of interest in working out how to tell the difference between these ‘tigers’ and ‘pussy cats’ as early as possible.

Genetic markers

Some researchers, such as Cancer Research UK’s Dr Ros Eeles, have been trying to find genetic markers that predispose men to the disease. And other researchers have suggested that certain genes that are hyperactive in aggressive cancers could be used to identify them.

Others, like the team that published the paper in Nature, are looking for substances in the blood and urine that might be useful markers of aggressive cancers.


There are tens, if not hundreds, of papers published every year that find new proteins potentially linked to the disease. For example, UCL’s Professor Gareth Williams, whom we’ve funded since 1989, is looking at a protein called MCM5.

This protein is involved in DNA replication – Professor Williams’ area of expertise. Cells that are dividing quickly can release this protein into the bloodstream, urine or prostate fluid, and thus could be used to detect prostate and bladder cancer.

Other labs are studying other proteins. But as yet, none of these has been fully tested in proper clinical trials to prove that they’re better than PSA. That’s not to say this won’t happen in the future, but we’re not there yet.

The Nature paper

The ‘test’ covered in the media this week is similarly early in its development, but takes a new approach

Rather than looking for genes or proteins, researchers at the University of Michigan measured the levels of chemical by-products of the reactions inside our cells, known as ‘metabolites’, in 42 tissue samples, 110 blood samples, and the same number of urine samples.

The samples were taken from 14 patients with advanced prostate cancer, 12 with early prostate cancer, and a further 16 from people with benign disease.

In all they looked at levels of over a thousand different chemicals in each of these samples.

When they analysed their data, they found that 10 of chemicals these were found at much higher levels in prostate cancer than normal samples, and were present most often in samples from people with advanced disease.

One of these chemicals, sarcosine, really stood out.

Nearly eight out of ten advanced cancer samples contained sarcosine, compared with around four out of ten samples from men with early disease, and none in the benign samples.

On the back of this, the researchers performed more experiments with prostate cells in the laboratory. The results from these experiments suggested that sarcosine could actually encourage prostate cancer cells to start spreading.

This is an important finding, as it ties sarcosine into the disease’s biology.

And finally, the team found that sarcosine was effective at predicting which men with ‘intermediate’ PSA levels would go on to have positive biopsies.

This suggests that it could be more useful at detecting the disease than PSA, and help doctors decide how to treat men with the disease.

Early days

All of this is still a long way from being used in a doctor’s surgery. The authors themselves reckon that it’s at least 5 years before this could form the basis of a clinical test – and that’s assuming that the test clears every regulatory hurdle between now and then.

But, taken together with all the other results from labs around the world, a bigger picture starts to emerge.

Given the differences between every individual prostate cancer, it’s unlikely that anyone will ever discover a single molecular ‘message’ that says “this man has aggressive prostate cancer”.

But if scientists can discover enough different indicators, each of which is reasonably accurate, then the most likely scenario is that in a few years from now, doctors will have a whole panel of molecules they can test for – genes, proteins, metabolites and others – and the overall results will give an extremely accurate picture of the state of a man’s prostate.

That’s the vision, anyway. But as we often say at the end of these posts… more research is needed.



Arun Sreekumar et al. (2009). Metabolomic profiles delineate potential role for sarcosine in prostate cancer progression Nature, 457 (7231), 910-914 DOI: 10.1038/nature07762

Rosalind A Eeles et al. (2008). Multiple newly identified loci associated with prostate cancer susceptibility Nature Genetics, 40 (3), 316-321 DOI: 10.1038/ng.90

K. Stoeber et al. (2002). Diagnosis of Genito-Urinary Tract Cancer by Detection of Minichromosome Maintenance 5 Protein in Urine Sediments J Nat Cancer Inst, 94 (14), 1071-1079 DOI: 10.1093/jnci/94.14.1071

Christopher S Foster et al. (2004). Transcription factor E2F3 overexpressed in prostate cancer independently predicts clinical outcome Oncogene, 23 (35), 5871-5879 DOI: 10.1038/sj.onc.1207800