The American Association for Cancer Research and the Japanese Cancer Association teamed up for the AACR-JCA Joint Session: Tracking Tumor Evolutionary Dynamics—From Initiation through Metastasis. Speakers in this major symposium discussed the dynamics of cancer evolution as it transforms from a normal cell, expands its population size, manifests as a tumor, and ultimately spreads through metastasis.
The presentation and related live panel discussion that took place on Tuesday provided an overview of recent progress in understanding of early cancer evolution in normal tissues, development of primary tumors, and evolution of metastatic potential. Registrants can watch a replay through June 21, 2021.
Somatic mutation development
Peter J. Campbell, PhD, Wellcome Sanger Institute, said that cancer evolution is a stepwise process with accumulation of mutations throughout life in all of our cells, with occasional mutationsin cancer genes. His presentation focused on what is happening at the beginning of cancer evolution using an example of phylogenetic reconstruction from a 59-year-old male’s prostate, noting that similar approaches can be used for other organ systems.
In their studies, Campbell and colleagues took the prostate, sliced it, and stained and imaged the slices, mapping individual ductal units from the urethra, through the branches and into the prostate. They performed whole genome sequences and phylogenetic tree reconstruction. They found that somatic mutations accumulated steadily and identified two waves of ductal morphogenesis.
They saw coalescences of branch points in phylogeny that seemed to cluster at embryonic development and pubertal maturation. The rudimentary ductal structure was laid down in embryonic development, and that stayed quiescent until puberty. At that stage, puberty drives formation of new side and terminal branches, seeded monoclonally by local stem cells scattered throughout the ductal tree. Tissue maintenance in adulthood is undertaken by local progenitor cells, leading to limited geographic scope of clones.
Although driver mutations are rare, they did see one driver mutation in 409 whole genomes. This clone was the only one to show branch point in the phylogenic tree timed to adulthood, and it spread across much wider geographic range than any other adult clone.
Seishi Ogawa, MD, PhD, Kyoto University, presented research on clonal evolution associated with chronic inflammation.
Clonal expansion in normal epithelium is almost obligatory in the esophagus and substantially promoted by heavy smoking and alcohol drinking, but rarely seen in normal colorectal epithelium, Ogawa said.
Ogawa and colleagues conducted a study using intensive multi-regional sampling to examine clonal expansion in the intestinal epithelia of patients with ulcerative colitis and compared it with controls. They found that the inflamed intestines in these patients underwent widespread remodeling by pervasive clones, and the majority of these clones were positively selected by acquiring mutations that involved genes often implicated in the downregulation of pro-inflammatory signals.
In inflamed colorectal epithelium of ulcerative colitis patients, clonal expansion was the rule and could be so extensive that a large area of colorectal epithelium was reconstituted by only a few clones, where IL-17 and other cytokine pathways play a major role in positive selection in ulcerative colitis epithelium.
The mechanism of positive selection in pre-cancerous fields may differ from what operates in cancer development, Ogawa said, raising a possibility that driver mutations detected in cancer samples by sequencing represent positive selection in precancerous fields but may not have a driver role in cancer development.
Modeling tumor evolution
Christina Curtis, PhD, MSc, Stanford University, spoke about modeling tumor evolution in patient samples and human organoids.
Clonal evolution is the driving force behind current public health issues such as infectious disease and cancer. To date, limited efforts have been invested in treatment and prevention from an evolutionary perspective because characterizing adaptive evolution is exceedingly difficult, Curtis said.
Curtis and colleagues have developed a model to trace premalignant clonal succession leading up to transformation in gastric organoids. This new platform for experimental evolution allowed them to map genotype to phenotype as single cell resolution. It should be applicable in other systems, also.
Using this model, they showed that TP53 deficiency induces gastric cancer in associated copy number alterations and transcriptional programs. In fact, these mirror patterns in the chromosomally unstable subgroup.
Lineage tracing by Curtis and colleagues demonstrated evidence for stringent selection for subclones with intrinsic fitness advantages, and this is accompanied by deterministic clonal expansion. By performing long-term evolution in organoids, they showed that they can mirror the latency of clonal succession that occurs in human tissue often over decades.
Clonal evolution modes
Tatsuhiro Shibata, MD, PhD, The Institute of Medical Science, The University of Tokyo, discussed temporal diversities of clonal evolution modes during colon carcinogenesis. According to Shibata, a classical linear, multi-step model of colon carcinogenesis has been proposed, but this model is a macroscopic viewpoint of tumor evolution. To infer the tumor heterogeneity, a microscopic view of clonal expansion during carcinogenesis is needed.
Shibata detailed a study that he said uncovered the history of clonal evolution from adenoma to metastasis. The study showed that clonal evolution modes and selective conditions temporarily change and influence tumor heterogeneity during colon carcinogenesis.
For example, looking at adenoma and early colorectal cancer (CRC) cases, they found that most driver genes are subclonal, suggesting Darwinian evolution is dominant at this stage. In advanced CRC cases, they found that major drivers such as APC, KRAS, and BRAF are clonal and in most cases shift to neutral evolution. This indicates there is a phase-transition point between these two types of evolutions.
Shibata also discussed how distinct molecular alterations—mutations, structural alterations, and host immune evasions—could be involved in this complex process.