Article Text

Download PDFPDF
PWE-065 Investigating clonal competition in Barrett's associated tumorigenesis using spatial maps of genetic heterogeneity
  1. T A Graham1,
  2. S Khan1,
  3. A Green2,
  4. D Oukrif3,
  5. L Gay2,
  6. H Barr4,
  7. M Novelli3,
  8. J Jankowski2,
  9. N A Shepherd5,
  10. S J Leedham1,
  11. S A C McDonald2,
  12. N A Wright1
  1. 1Histopathology Laboratory, Cancer Research UK London Research Institute, London, UK
  2. 2Centre for Digestive Diseases, Barts and the London School of Medicine and Dentistry, London, UK
  3. 3Department of Histopathology, University College London Hospitals, London, UK
  4. 4Upper GI Surgery, Gloucestershire Royal Hospitals, Gloucester, UK
  5. 5Department of Histopathology, Gloucestershire Royal Hospitals, London, UK


Introduction Genetic heterogeneity within Barrett's mucosa is significantly associated with the development of oesophageal adenocarcinoma (OA). Why genetic diversity promotes tumorigenesis is unclear. Pro-tumorigenic interaction between clones is a possible explanation. To look for evidence of clonal interaction within Barrett's lesions, we have undertaken a high-resolution clonal-ordering analysis to produce phylogenetic trees and spatial maps that illustrate the spread of genetically distinct clones in endoscopic mucosal resection specimens (EMRs).

Methods Entire EMR specimens were serially sectioned. Each crypt was classified according to histological grade. Aneuploidy was assessed using image cytometry. More than 50 individual crypts were then laser-capture microdissected from each EMR. The genetic mutation burden of individual crypts was assessed (K-ras, p53 and p16 point mutations and microsatellite loss of heterozygosity (LOH) data for up to16 loci on chromosomes 3p (FHIT), 5q (APC), 9p (p16), 17p (p53), 17q and 18q (SMAD4)). In each EMR, a point mutation in at least one gene was identified. The relative location of each analysed crypt was recorded and used to reconstruct a genetic and histological map of each EMR.

Results EMRs were rarely genetically homogeneous. Typically, dysplastic crypts within individual lesions were clonal for a founding p53 or p16 point-mutated clone, whereas the adjacent metaplastic (non-dysplastic) crypts analysed did not contain the point mutation. Mosaic subclones with distinct patterns of LOH (often 18q and/or 3p) were present. In one EMR, the same p53 and p16 point mutations and 5q and 17p LOH were detected across the entire EMR.

Conclusion Clonal expansion of mutated cells is frequently restricted to small numbers of crypts. Interactions between genetically distinct clones appears to be predominantly competitive: a crypt acquiring an additional pro-tumorigenic mutation can outcompete, and so clonally expand to the detriment of, the original crypts. Further investigation will reveal whether this progression is driven by mutation and selection or by additional mutations altering the behaviour of neighbouring pre-existing clones.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.