Longitudinal single cell clonal analysis reveals evolutionary stasis and predetermined malignant potential in non-dysplastic Barrett’s Oesophagus

Pierre Martinez1,Margriet R. Timmer2,Chiu T. Lau2,Sybren L. Meijer2,Fiebo J.W. ten Kate2,Rosalie C. Mallant-Hent3,9,Anton H.J. Naber4,9,Arnoud H.A.M. van Oijen5,9,Lubbertus C. Baak6,9,Pieter Scholten7,9,Clarisse J.M. Böhmer8,9,Paul Fockens2,Jacques J.G.H.M. Bergman2,9,Carlo C. Maley10,11,Trevor A. Graham1,Kausilia K. Krishnadath2,9

1Barts Cancer Institute – Queen Mary University of London, London, UK,2Academic Medical Center – University of Amsterdam, Amsterdam, The Netherlands,3Department of Gastroenterology and Hepatology, Flevoziekenhuis, Almere, The Netherlands,4Department of Gastroenterology and Hepatology, Tergooiziekenhuizen, Hilversum, The Netherlands,5Department of Gastroenterology and Hepatology, Medisch Centrum Alkmaar, Alkmaar, The Netherlands,6Department of Gastroenterology and Hepatology, Onze Lieve Vrouwe Gasthuis, Amsterdam, The Netherlands,7Department of Gastroenterology and Hepatology, Sint Lucas Andreas Ziekenhuis, Amsterdam, The Netherlands,8Department of Gastroenterology and Hepatology, Spaarne Ziekenhuis, Hoofddorp, The Netherlands,9Gastroenterological Association, Amsterdam, The Netherlands,10Center for Evolution and Cancer, University of California at San Francisco, San Francisco, CA, USA,11Biodesign Institute, School of Life Sciences, Arizona State University, Tempe, AZ, USA

Presenting date: Tuesday 3 November
Presenting time: 16.10-16.25


Endoscopic surveillance of the premalignant condition Barrett's Oesophagus provides a unique opportunity to study the evolution of a solid human neoplasm over both space and prolonged periods of time. Here we report the pattern of longitudinal clonal evolution occurring in a large cohort of 195 non-dysplastic Barrett's patients followed for a median duration of 43 (range: 11-130) months.


Multicolor fluorescence in situ hybridization was used to obtain single-cell genetic profiles from at least two different time points from each patient. We calculated several genetic diversity metrics and used uni- and multivariate analyses, Kaplan-Meier survival curves and resampling analyses to assess the relevance of our findings. We estimated clone size by multiplying clonal frequencies by segment area estimates.


We found that the level of genetic diversity in the population remained relatively stable over time with only one detectable clonal expansion every 36.8 patient years. These clones grew at an average of 2.1mm per year. Nevertheless, the non-dysplastic Barrett's lesions proved to be in a dynamic equilibrium whereby existing clones were continually lost and replaced by new clones. Strikingly, the baseline level of genetic diversity was a significant predictor of progression risk that outperformed all other single-marker measures.


Our results indicate a lack of strong selection for mutant clones within the Barrett's segment, while a high initial diversity status predicts future progression risk. These data challenge the existing stepwise model of carcinogenesis in solid tissues, and even argue against the prior model of neoplastic progression as a series of selective sweeps. Our results, consistent with other recent findings, suggest that clonal expansions are rare and slow growing, and that the malignant potential of a ‘benign' Barrett's lesion is predetermined within this observational timeframe. These observations have important consequences for managing cancer risk in Barrett's patients.