The Most Important Milestone for Future Cancer Treatments: THE MUTATION CATALOG

Whole-genome analysis of cancer or pan-cancer analysis of whole genomes (Pan-Cancer Analysis of Whole Genomes, PCAWG) scientists published unprecedented data analyzing more than 2600 cancer genomes for mutation events. They cataloged and integrated a broad, comprehensive portrait of cancer-related mutations in coding and noncoding genome regions across 38 major tumor types and used this information to identify relevant biological pathways.

What Is a Genome?

The genome is the complete set of instructions in DNA. Each cell nucleus contains 2 DNA copies, one from the mother and one from the father. DNA folds and packages in a specific way to form chromosomes. Our cells contain 46 chromosomes of packaged DNA helices (23 from the mother and 23 from the father).

Summary of the Pan-Cancer Analysis of Whole Genomes Project:

• Approximately five driver mutations were identified in each cancer genome (% 5 had no driver mutation).
• Most driver mutations were found in the genome's coding region.
• A “molecular clock” can determine which mutations occurred early in cancer evolution.
• Genetic mutation patterns may help identify cancer type, assisting doctors with cancers of unknown primary / origin.

This collective effort involving collaboration across 37 countries between more than 1300 scientists and doctors resulted in Nature and associated journals simultaneously publishing 23 scientific papers.

The Pan-Cancer Analysis of Whole Genomes dataset is available to all cancer researchers worldwide. Its broad availability and quality will almost certainly generate a wave of biological insights and methodological advances.

For example, researchers' work analyzing whether mutations occurred earlier or later, using molecular clocks , may ultimately help early detection or prevention strategies.

Overall, the findings suggest that, years before cancer diagnosis, driver mutations may have implications for early detection and biomarker development. Driver mutations critical to cancer formation can arise years before diagnosis, which may be used for early detection.

However, inferential analyses such as whole-genome cancer analysis, while offering deeper understanding, are generally associated with greater complexity.

Pan-Cancer Analysis of Whole Genomes – Coding and Noncoding DNA Regions

Pan-Cancer Analysis of Whole Genomes (PCAWG) data and new understanding differ from conventional whole-exome sequencing .

Note: The exomeis the sum of DNA's protein-coding regions called exons and comprises only %1 of DNA, corresponding to 20 thousand protein-coding genes. Thus, whole-exome analysis analyzes mutations in only %2 of DNA. Introns are non-protein-coding DNA regions and until recently were called junk (insignificant) DNA .

So far, all targeted cancer drugshave been developed by examining DNA sequences from whole-exome sequencing, representing just %1 of the genome. Now, including noncoding regions has expanded coverage to %100.

Raw, processed, and interpreted Pan-Cancer Analysis of Whole Genomes project data are available to continue supplying other researchers for years to come on the PCAWG website online.

Pan-Cancer Analysis

The pan-cancer analysis used whole-genome sequencing data from 2605 primary tumors and 173 metastases or local recurrences , with cancerous tissue and matched normal tissue. Data were analyzed for 3 main mutation groups : somatic single-nucleotide variants (SNV); small insertions and deletions (indels); and copy-number changes and structural variants (SV), rearrangements of large DNA segments.

Impressively, cataloging revealed > 43 million SNVs, > 400.000 somatic multinucleotide variants, > 2.4 million somatic indels, nearly 290.000 somatic SVs, > 19.000 somatic retrotransposition events, and > 8000 de novo mitochondrial DNA mutations. The analysis reported substantial heterogeneity in somatic mutation burden across tumor types and patients.

Cataloging and understanding driver mutations has been a major cancer research goal for the past twenty years. Although driver mutations are limited in number, thousands or tens of thousands of combinations exist.

Researchers found that % 91 of cancers had at least one identifiable driver mutation, with 4.6 drivers per tumor. No driver mutation was reported for approximately % 5 of cancers.

By tumor type, driver SVs, structural variants, were more common in breast and ovarian adenocarcinomas. Driver point mutations played important roles in colorectal adenocarcinomas and mature B-cell lymphomas. Tumor suppressor genes were associated with many biallelic inactivating driver mutations: TP53mutations involved both alleles in % 77 of cases.

Patients with chromophobe renal cell carcinoma and pancreatic neuroendocrine cancer —tumors with notable chromosomal aneuploidy—had no cancer driver mutations detected. PCAWG researchers stated: “Certain combinations of whole-chromosome gains and losses may be sufficient to initiate a cancer in the absence of more targeted driver mutations.” This was their explanation.

Genetic differences between tumors explain why two patients with the same diagnosis respond completely differently to treatments.

The reasons are written in tumor genomes. The striking finding is how different one person's cancer genome is from another's: thousands of mutation combinations and more than 80 different processes that can cause cancer...

Approximately % 5 of all cancers are diagnosed as cancer of unknown primary (carcinoma of unknown origin). Mutation patterns and the genes affected may help identify the patient's cancer type.

A Closer Look at Mutations in Noncoding Regions

In analyses of noncoding somatic drivers, NEAT1 and MALAT1 long noncoding RNAand other previously reported noncoding drivers are investigated, while new noncoding drivers are also revealed.

PCAWG researchers found single-site hotspots in the TERT promoter associated with higher TERT expression (a promoteris a DNA segment that initiates gene transcription).

TOB1 had implications for stomach and breast cancers.

When TP53 promoter mutations were reported, no additional coding-region mutations were present, suggesting for the first time strong TP53 inactivation by noncoding mutations.

Researchers also reported new structural-variant mutation candidates for driving oncogenesis, including rearrangements involving AKR1C genes and BRD4.

However, noncoding driver mutations were rare. Researchers say larger datasets and technological advances will keep identifying new noncoding drivers, although at much lower frequencies than protein-coding drivers.

Molecular Clocks

Researchers also addressed whether mutations occurred early or late in cancer evolution. The molecular clock conceptwas explored by determining whether mutations were clonal, present in all cancer clones, or subclonal, present in some, and whether clonal events preceded subclonal ones.

In 2583 samples, among 47 million point mutations, % 22 were early clonal events, % 7 late clonal, % 53 unspecified clonal, and % 17 unspecified subclonal. In a panel of 453 cancer driver genes, they identified a distribution of % 29 early clonal, % 5 late clonal, % 56 unspecified clonal, and % 8 subclonal events among 5913 oncogenic point mutations.

Driver mutations in KRAS , TP53, and PIK3CA , and the noncoding driver TERT, were considered early events in cancer development. Researchers noted that the mutational signature changed over time in % 40 of tumors.

How Should a Doctor Use This Wealth of Information?

“The Pan-Cancer Project provides a roadmap for how genomes should be collected and analyzed. We developed high-quality standardized processes to analyze genomes and present them in an integrated way,” said project leader Dr. Lincoln D. Stein at a press conference. “This project identified the same pathways as exome sequencing, and we found many additional ways to refine them,” he added.

Furthermore, as our ability to understand which signaling / growth pathways and genes are altered in a particular cancer improves, we can more accurately see changes in a patient's tumor and use more targeted treatments .

Previously, looking only at exome sequences left us in the dark about tumor genomes in approximately a third of clinic patients. We would have no information to continue precision oncology .

However, an important limitation is the absence of clinical data linking patients to their treatments.

The ongoing ICGC ARGO project is addressing exactly this gap. In addition to PCAWG publications, they are collecting clinical and genomic data from 100.000 cancer patients.

Once sufficient data link genomic information with clinical outcomes, systems will begin to be developed to help doctors use patients’ molecular and genomic information to identify better treatment options, using machine learning and decision-support systems.

However, cancer genome testing will not soon be available in every clinic. Unlike currently available multigene panels, whole-genome sequencing laboratory processes and standardization must be transferred to genetic testing centers.

Running these new tools and analyzing their data also cost money. They are expensive now, but costs will fall. This technology will not arrive tomorrow, but will become more readily available in a few years.

Related Topic:

- What is next-generation sequencing (Next-Generation Sequencing, NGS)? Its use in oncology