Abstract P3-07-06: Evidence for tumor heterogeneity and clonal evolution during invasive progression in breast cancer

Purpose: Intratumoral heterogeneity is well recognized to be an important driver of treatment resistance and metastasis. We undertook this N of three study to measure the degree of heterogeneity in three large preinvasive lesions, all with invasive components to determine the relationship between tumor heterogeneity, spatial distribution, clonal evolution, and invasive progression.

Methods: We identified patients A, B, C with extensive DCIS measuring 7.5 cm, 6 cm, and 7 cm associated with 0.3 cm, 3.8cm, and 3.4 cm of an invasive component and 0, 7 and 1 positive lymph node, respectively. We sequenced the tumor sample for Case A from 32 unique blocks with precise geospatial localization; invasive cancer was identified in 3 of 32 blocks. Case B had 26 blocks sequenced with invasive cancer in 13 of 26 blocks. Case C had 23 blocks sequenced with invasive in 11 of 23 blocks. For germline reference, we sequenced DNA from an uninvolved tissue from each case. NGS libraries were made from FFPE derived DNA (20-40ng) for full exome sequencing. Variant calling was performed by GATK HaplotypeCaller, Platypus and Mutect. Identified somatic mutations were annotated with Oncotator and pathway enrichment analysis was performed with Bioconductor. To investigate the clonal evolution and progression history, phylogenetic trees were constructed in R and sub-clonal analysis was performed with Treeomics.

Results: The sequence data was analyzed with Platypus, MuTect and GATK HaplotypeCaller. The somatic mutation sites were concatenated into one sequence for each sample. Both neighbor-joining trees and maximum parsimony trees were built for each case. Phylogenetic analysis and sub-clonal analysis support the multi-clonal invasion model of invasive cells, in which invasive cancer can evolve from multiple clades, either early or late in the evolutionary history, independently. Dense sampling allowed reconstruction of the temporal order of mutations that accumulated in the cell lineage of the invasive cancers. Furthermore, phylogeny and sub-clone spatial analysis revealed that distant regions may be closely genetically related and showed a weak spatial sub-clone clustering pattern, which is consistent with the predictions of Big Bang model. For driver genes, we find that except for SETD2 in Case B, the majority of driver gene mutations are sub-clonal. Somatic mutations on ATP-binding cassette (ABC) transporter pathway was found in all cases.

Conclusions: Extensive sampling and sequencing of tumors yields important insights about tumor heterogeneity and tumor progression of DCIS to invasive cancer. Variable invasive propensity was identified, with foci of invasion were geospatially associated with preinvasive regions of progressively higher mutational load.

Citation Format: Ding Y, Marks J, King L, Hardman T, Hall A, Mallo D, Rodrigo A, Maley C, Hwang S. Evidence for tumor heterogeneity and clonal evolution during invasive progression in breast cancer [abstract]. In: Proceedings of the 2018 San Antonio Breast Cancer Symposium; 2018 Dec 4-8; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2019;79(4 Suppl):Abstract nr P3-07-06.

Carbon dating cancer: defining the chronology of metastatic progression in colorectal cancer

Background

Patients often ask oncologists how long a cancer has been present before causing symptoms or spreading to other organs. The evolutionary trajectory of cancers can be defined using phylogenetic approaches but lack of chronological references makes dating the exact onset of tumours very challenging.

Patients and methods

Here, we describe the case of a colorectal cancer (CRC) patient presenting with synchronous lung metastasis and metachronous thyroid, chest wall and urinary tract metastases over the course of 5 years. The chest wall metastasis was caused by needle tract seeding, implying a known time of onset. Using whole genome sequencing data from primary and metastatic sites we inferred the complete chronology of the cancer by exploiting the time of needle tract seeding as an in vivo 'stopwatch'. This approach allowed us to follow the progression of the disease back in time, dating each ancestral node of the phylogenetic tree in the past history of the tumour. We used a Bayesian phylogenomic approach, which accounts for possible dynamic changes in mutational rate, to reconstruct the phylogenetic tree and effectively 'carbon date' the malignant progression.

Results

The primary colon cancer emerged between 5 and 8 years before the clinical diagnosis. The primary tumour metastasized to the lung and the thyroid within a year from its onset. The thyroid lesion presented as a tumour-to-tumour deposit within a benign Hurthle adenoma. Despite rapid metastatic progression from the primary tumour, the patient showed an indolent disease course. Primary cancer and metastases were microsatellite stable and displayed low chromosomal instability. Neo-antigen analysis suggested minimal immunogenicity.

Conclusion

Our data provide the first in vivo experimental evidence documenting the timing of metastatic progression in CRC and suggest that genomic instability might be more important than the metastatic potential of the primary cancer in dictating CRC fate.

Mutation rates as adaptations

In order to better understand life, it is helpful to look beyond the envelop of life as we know it. A simple model of coevolution was implemented with the addition of a gene for the mutation rate of the individual. This allowed the mutation rate itself to evolve in a lineage. The model shows that when the individuals interact in a sort of zero-sum game, the lineages maintain relatively high mutation rates. However, when individuals engage in interactions that have greater consequences for one individual in the interaction than the other, lineages tend to evolve relatively low mutation rates. This model suggests that one possible cause for differential mutation rates across genes may be the coevolutionary pressure of the various forms of interactions with other genes.