Monthly Archives: September 2012

A very short introduction to intratumour heterogeneity.

We have known for a while that tumours are basically unique and different, barring very few cancers with a very simple aetiology (like retinoblastoma).This led to the formulation of the concept of personalised cancer medicine wherein therapies would be tailored to match the mutational makeup of a patient’s tumour.

You would think that was a great idea until someone realised that tumour evolution isn’t a ladderlike series of clonal expansions as imagined before.

The original conception of tumour evolution as a linear process.

However, if you actually look at what evolution is like, there are multiple species that evolve in parallel, and researchers started to investigate if this was the case with tumour biology as well, in line with what the only diagram in Charles Darwin’s magnum opus; The Origin of Species.

Parallel Evolution

Studies have since gone on to look at what happens to tumour cells with time, and they do duly evolve and there are multiple subclones with allele frequencies that change in response to chemotherapy, for instance.


Tumours change with time in response to chemotherapy.

Using DNA sequencing methods, Tim Ley and coworkers showed that tumour evolution with time could follow either of two models – either one of the subclones acquired new mutations in response to chemotherapy and evolved to form the relapse clone or a previously present clone was driven by the selective pressure of chemotherapy to expand into the relapse population. (Reference – )

While this study only showed that there was a genetically distinct relapse population compared to pre-chemotherapy disease, it did not focus on parallel evolution. But Charlie Swanton’s group at the CRUK’s LRI and Tariq Enver’ s group at the UCL Cancer Institute did.

Enver and colleagues used a technique called FISH to examine changes in the copy number of candidate genes (gains or losses in how many copies of a gene there are) in Acute Lymphoid Leukaemia. They found that there were multiple subclones present at the same time with copy number changes that were independently acquired in some cases and were part of a complex, branched hierarchy. (Reference – )

a) represents assumed linear architecture. b) represents an inferred complex architecture using FISH with moderate complexity. c) represents evidence for an even more complex architecture with eight subclones from a different patient.

Charlie Swanton’s group used multiregion sequencing to explore heterogeneity in renal cell carcinoma and again found evidence for a complex pattern of evolution with multiple subclones evolving in parallel. (Reference – )

Self-explanatory. Charlie Swanton’s group found evidence for a complex, branched pattern of tumour evolution using an approach that combined deep sequencing with multiple biopsies of both the primary tumour and metastases.

And to make matters worse, we have emerging evidence for heterogeneity on an epigenetic level as well, and we might begin to have to consider the implications of that for tumour evolution. (Reference – ).

We are therefore now required to consider multiple biopsies for every case and then develop strategies to find what drives all these subclones, and then devise therapies accordingly. Things have gotten just a little bit more complicated for personalised cancer medicine…