Precision Medicine Experts Exploit Multiple Mutations to Improve Cancer Treatment

Highlights from MAP (13-14 October 2017, Zurich, Switzerland)

Experts in precision medicine are exploiting multiple mutations to improve treatment of cancer, as highlighted at the Molecular Analysis for Personalised therapy (MAP) Congress, held 13-14 October in Zurich, Switzerland.

The 2017 meeting focused on two aspects of precision medicine: the complexity of the molecular mechanisms that lead to cancer progression, and the use of technologies to decipher this complexity.

Fabrice André

Commenting on the first topic, Dr Fabrice André, co-chair of the MAP meeting and Chair of the ESMO Translational Research and Personalised Medicine Working Group, said: “We know now that some cancers need several mutations to grow. Precision medicine recognises that patient outcomes do not depend just on one mutation but on complex interactions between multiple mutations.”

During cancer progression, DNA alterations (e.g. mutations) appear randomly in cancer cells (due to replication errors or exogenous factors, for instance). Only the alterations conferring a survival advantage to the cancer cells are 'kept', following the well-known process called natural selection. This selection process is governed by 'evolutionary dependencies' between alterations, as the survival advantage of an alteration depends on the ones already present in the cancer.

A study presented at the MAP conference1,2 aimed to discover the evolutionary dependencies between alterations, and to understand how they affect cancer response to therapeutic agents. The researchers inferred hundreds of evolutionary dependencies by looking for patterns between alteration occurrences in a large cohort of more than 6,000 patients profiled by The Cancer Genome Atlas (TCGA) consortium. Next, they checked whether these dependencies were altering the response of cancer cell lines to therapeutic agents.

MAP 2017 Mina

“We found evidence supporting the existence of hundreds of evolutionary dependencies between alterations in cancer,” said lead author Dr Marco Mina University of Lausanne, Switzerland. “We also found evidence that these evolutionary dependencies affect the response of cancer cell lines to therapeutic agents.”

One of the strongest dependencies was the synergistic effect between mutations in ARID1A and RNF43 genes, potentially inducing cancer cells to proliferate faster in colorectal cancer and stomach adenocarcinoma. Cancer cell lines harbouring mutations in both ARID1A and RNF43 genes were significantly more sensitive to the VX-680 Aurora kinase inhibitor drug, paving the way towards a new and possibly promising therapy option. 

MAP 2017 RNF43 ARID1A

RNF43 – ARID1A co-occurrence

 

Mina said: “Dependencies between alterations should be considered when deciding on patient therapy. Our results suggest that when selecting the best therapeutic course of action for treating patients with stomach adenocarcinoma or colorectal cancer, it is important to determine whether the cancer harbours mutations in both ARID1A and RNF43 genes, as this might drastically change the response of the cancer to some therapeutic agents.” 

“This study illustrates how the co-occurrence of several mutations can lead to cancer progression,” said André. “Predicting outcome cannot be based on a single mutation. What is important is the overall landscape of mutations.”

Regarding the second topic of technology, André said that scientists know how to sequence DNA. The issues requiring more knowledge are how to transform raw sequencing data into biological information such as copy number alterations, and how to interpret the clinical meaning of the biological data.

When it comes to transforming raw sequencing data, researchers at the MAP conference showed3 how an algorithm they developed, called Excavator2, expands the number of copy number alterations that can be identified by small targeted sequencing panels, and therefore could be targeted for cancer treatment.

MAP 2017 Excavator 2

Excavator2 gives consistent results and allows to call off-target copy number alterations. © Luca Mazzarella 

 

Routine diagnosis of clinically actionable mutations is based on the use of targeted panels, which deplete genomic regions considered of secondary interest. A considerable amount of this ‘off-target’ genome sneaks through the enrichment process, but its sequencing is usually discarded in the final analysis. The researchers found a way to recover these off-target reads and use them to identify larger genomic alterations (structural variants like amplifications or deletions) extending outside the range of the targeted panel.

MAP 2017 Mazzarella

Lead author Dr Luca Mazzarella, European Institute of Oncology, Milan, Italy, said: “Our method enhances the identification of actionable structural alterations using panels that are routinely used in the clinic, expanding the chances of finding targeted treatments for patients. Additionally, it may identify structural alteration ‘signatures’ characterising the underlying biology of the tumour, which would be useful for instance to find tumours with functional loss of homologous recombination genes that benefit from PARP inhibitors.”

André said: “This study is an example of how we can use bioinformatic tools to fine-tune the analysis of genome sequences.”

References

1. Mina M, Raynaud F, Tavernari D, et al. Interrogating functional dependencies between genomic alterations can facilitate precision medicine approaches in cancer. Annals of Oncology, Volume 28, Issue suppl_7, 1 October 2017, mdx508.001, https://doi.org/10.1093/annonc/mdx508.001

2. Mina M, Raynaud F, Tavernari D, et al. Conditional Selection of Genomic Alterations Dictates Cancer Evolution and Oncogenic Dependencies. Cancer Cell 2017; 32(2): 155-168.

3. Mazzarella L, D’aurizio R, Frige D, et al. Genome-wide identification of actionable copy number alterations from targeted sequencing panels with Excavator2. Annals of Oncology, Volume 28, Issue suppl_7, 1 October 2017, mdx508.002, https://doi.org/10.1093/annonc/mdx508.002