Curious Dr. George | Plumbing the Core and Nibbling at the Margins of Cancer

Seeing the Forest for the Trees: How Cancer Evolution Affects Treatment

Lisandra E. West, PhD, Senior Scientific Knowledge Engineer, CollabRx.

Q: How might current thinking on tumor evolution/heterogeneity and sequential mutational profiling inform optimal therapy selection?
A: I’d like to share some learnings from ASCO, 2016:
To help think about tumor evolution over the course of time and treatment of cancer, I’m a fan of the palm tree analogy (made by Charles Swanton, ASCO 2016). Early genetic alterations that occur in the developing tumor are “trunk” genetic events that are present in every cancer cell in the tumor. Trunk driver mutations may be predictive of drug sensitivity or intrinsic resistance. As the tumor continues to grow some cells will develop additional mutations that will be shared by their progeny cells within the tumor. To continue the palm tree analogy, this results in a branching effect, with each branch representing a sub-clonal population of tumor cells that have new acquired mutations not shared with the original trunk cells or cells in branches that evolved earlier. Swanton cites an earlier study (Landau et al, Cell, 2013) that demonstrated that acquisition of subclonal driver mutations (occurring in the tree branches) results in significantly reduced survival over time as compared to patients whose tumors lack subclonal drivers.
How does subclonal heterogeneity (many tree branches) affect response to targeted therapy? Swanton cited a study (Pearson et al, Cancer Discovery, (2016)) that showed that the best response to FGFR inhibitor therapy occurred in patients whose tumors were homogenous (having few or no branches) for FGFR amplification. Non-responding tumors had sub-clonal heterogeneous FGFR amplification AND presence of FGFR non-amplified tumor cells. This work indicates that targeting trunk or early clonal events is important because response to targeted therapy is better when a greater proportion of cancer cells in the patient share genetic alterations. Interestingly, the overlap of mutations from two different metastatic sites biopsied from the same patient showed that percentage of mutations shared between sites decreases after therapy (venn diagrams presented by Alexandra Snyder, MD, at ASCO 2016). That is to say, there is higher diversity in mutational profiles from different metastatic sites after treatment. This indicates cancer evolution over the course of treatment(s), and leads to questions about whether molecular data from a single biopsy site can be sufficient to inform treatment decision making as cancer progresses.
Thus, therapy drives tumor heterogeneity, which then fosters polyclonal cancer therapy resistance. In a study of patients with metastatic CRC treated with EGFR antibodies, liquid biopsy and sequencing demonstrated that KRAS, NRAS, BRAF, and EGFR mutations were present in post-treatment circulating tumor DNA, whereas they had been absent in pretreatment cfDNA (Bettegowda et al, Science Translational Medicine (2014)). Treatment with a PI3K inhibitor has been shown to select for sub-clones that harbor inactivating PTEN mutations which confer resistance to PIK3CA inhibitor treatment (Juric et al, Nature (2013)). Chemotherapy damages DNA and causes mutations in tumor cells that persist after treatment. These cells go on to form sub-clonal populations that are resistant to drug treatment. In a heterogeneous tumor, drug resistant subclones take over and become the dominant cancer clones that ultimately cause cancer recurrence. Quite remarkably (at least to me), Swanton suggested consideration of treating a patient to stable disease, versus maximal tumor response, based on the idea that treating to maximal tumor response may provoke the rise of a resistant and untreatable subclonal cancer cell population – to the detriment of the patient.
An emerging mechanism fueling tumor diversity and subclonal evolution is genomic DNA cytosine deamination catalyzed by members of the AID/APOBEC family of DNA deaminases which induce genomic damage through their DNA deaminating activity. Deregulation of APOBEC enzymes causes a general mutator phenotype that gives rise to the non-random somatic mutations observed in cancer, ultimately manifesting as diverse and heterogeneous tumor subclones (Mcgranahan et al, Science Translational Medicine (2015); De Bruin et al, Science (2014)). Systemic drug treatment provides selective pressure that gives clones harboring resistance-conferring mutations a competitive growth advantage. We are beginning to understand that it will be necessary to predict and target tumor evolution over the course of treatment. This will require measuring (through ongoing liquid biopsies?) cancer’s subclonal structure that mitigates drug resistance.
Importantly we are beginning to find evidence that lots of mutations in cancer (mutational load/burden) or heterogeneity itself may represent an “Achilles heel” in cancer (Charles Swanton, ASCO 2016). Mutational burden is emerging as a predictive biomarker for sensitivity to checkpoint inhibitor immunotherapy in several cancer types including non-small cell lung cancer and melanoma. Detecting mutational burden, monitoring tumor evolution, and timing and sequencing of immunotherapy with targeted therapies seem likely to be areas of intense investigation moving forward in cancer research.
Copyright: This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.