The legacy of a cancer gone rogue: The evolutionary trajectory of metastatic oesophageal adenocarcinoma


The advent of the new year is accompanied by a time of reflection and the question of what has ultimately defined our legacy in the last decade. For cancer research, whole genome sequencing and its application to personalised medicine has played a vital role. For a cancer itself, one may argue that metastasis, and the route that the cancer takes from the primary to secondary sites is it’s legacy as it is the metastasised tumour that lives to tell the tale and ultimately results in death. 

Oesophageal adenocarcinoma (OAC), the sixth most common cancer worldwide with a rising incidence, is a deadly cancer with an overall dismal five year survival of 15-20%(, I was surprised at how little was known about this cancer and how it spreads when my uncle died of this disease when I was a medical student. How different or similar were metastases compared to the primary tumour and were there any tailored therapies for metastatic disease?  A decade later we were still asking the same questions and we were none the wiser. Fundamentally, we didn’t understand how OAC spreads. One practical limitation was the lack of available tissue samples in a cancer where patients were too frail for multiple biopsies. Hence in the first year of my PhD I worked closely with the invaluable input of oncologists, palliative care physicians, pathologists, hospice staff, surgeons, nurses, patient groups and allied health professionals to set up a rapid autopsy programme. Patients were invited to be donors for cancer research and would undergo a post-mortem that would have to be completed within 6 hours of death in order to ensure that tissue samples had the appropriate integrity to undergo whole genome sequencing. These programmes had been established in the US, but this was the first of its kind in the UK, with obvious ethical and strategic considerations that we navigated with care. 

We divided the scientific experiments of the study in two parts. In the first part, we discovered cancer clones in 116 samples across all cases using whole genome sequencing. In the second part, to overcome the limitations of cost and to enable us to analyse samples in unprecedented detail, we developed a way to track these clones in time and space in a further 272 sites.

At first glance the phylogenetic trees of these cases had a striking stellate appearance, with a near synchronous spread of clones from the primary tumour to multiple sites. Secondly we noticed that these clones spread with complete disregard to tissue type (lymph node versus solid organ) and anatomical location (proximal versus distant). This was best illustrated on body maps recreated for 6 cases in detail (up to 48 samples per case) where clones were dotted around the body akin to Smarties, with no evidence of tropism (Figure 4 in main paper). There was some evidence of ongoing evolution associated with this event (more at the level of subclonal driver amplifications) along with an absence of mutational signature 1, which has been shown to have clock like properties and hence aids in inferring timing of events. These features together described a distinct model of metastasis that we termed clonal diaspora. These metastatic clones were present in a large proportion of earlier diagnostic samples (primary tissue and blood), suggesting that the bulk of the evolution had happened early (mainly point mutations), though this did continue on a smaller scale (copy number alterations).  Ultimately OAC is a rogue cancer, that does not follow the rules set out by our traditional staging algorithms.These staging methodologies have been argued to be insufficient by clinicians in accurately prognosticating patients, but it has been difficult to show this to date with the lack of available tissue. 

We interrogated which genomic factors, if any may be associated with this diaspora event. Mutations associated with signature 3 and hence a failure of DNA double strand break repair by homologous recombination were more prevalent post diaspora, though whether this was cause or effect remains to be seen and suggests that a failure of DNA repair may drive genomic heterogeneity.  In the future it would be interesting to look at extra genomic factors, including epigenetics to see if the trigger for this diaspora event can be identified, pre-empted and eventually therapeutically targeted. 

At the start of this study, one of our palliative care physicians told me it would be the health care professionals who would need convincing about such a study, not the patients. She was absolutely right. The patients were the ultimate advocates for this study, as were their families. They carried their donor card in their wallets and it was their pride in taking part which ensured that all healthcare members came together to respect their wishes. Primary care physicians came out in the middle of the night to do death certificates, our pathologist opened the doors to his mortuary at all times and our team of doctors, nurses, surgeons and laboratory staff were happy to be called day or night, on their own time, to contribute. It was the patients who mobilised us, and inspired us to work together to get this done.

On a personal level, I was initially apprehensive as to how patients might react when approached, especially when there was so little to offer them in return. I had no cure for the disease that had taken over their body, I had no answers for why this had happen.  As a surgeon, this  was an unfamiliar territory for me. However, patients showed remarkable understanding and they were altruistic and compassionate in way that I had never witnessed before. A cryptic cross word writer, a priest, a young family man, a businessman, a scientist’s husband amongst others. Cancer knows no boundaries and can affect us all.  I hope that this study and the avenue that it has opened up for further research urgently needed in this field, contributed in some small part to their immense legacy, and for this we are incredibly grateful to these men and their loving families. 


Please sign in or register for FREE

If you are a registered user on Nature Portfolio Cancer Community, please sign in