Advertisment
ASCO Report 2014: The evolution of our molecular understanding of Colorectal Cancer
What we are doing now, what the future Holds, and how Tumour profiling is just the beginning
by Dr Paul Ross, Guy’s & St Thomas’ Hospitals, London. At the 50th annual meeting of the American Society of Clinical Oncology (ASCO) there was considerable focus on the role of predictive biomarkers and the use of same in both development of targeted therapies and in the selection of patients for established targeted therapies. The molecular understanding of colorectal cancer has evolved from the original Vogelstein molecular pathway applied to the adenoma-carcinoma sequence.
Sabine Tejpar reviewed established and putative biomarkers used in colorectal cancer. There are 2 established biomarkers RAS and microsatellite instability (MSI). It has been established and that stage II MSI-high tumours have a very good prognosis and do not require adjuvant chemotherapy. This was re-iterated in a presentation by Sargent this ASCO. Tejpar pointed out that even amongst this population further heterogeneity will exist.
The importance of tumour heterogeneity was the theme of her discussion regarding RAS. Mutations in exon 2 (codons 12, 13) of KRAS, occurring in 40% of colorectal cancers, are established negative predictive markers for use of anti-epidermal growth factor receptor (EGFR) monoclonal antibodies (MAb). Over the past year the importance of mutations in exons 3 and 4 (codons 59, 61, 117, 146) of KRAS and exons 2, 3 & 4 of NRAS have emerged. These additional RAS mutations account for approximately 15% of patients. The significance of these mutations has emerged from many trials including FIRE-3, CRYSTAL, OPUS and PRIME. The biological effects of each mutation may not be identical but the crucial observation is that all result in a disadvantage when treated with anti-EGFR MAb. On occasions, using such treatment in the presence of a mutation may actually confer harm. This was observed in the PRIME study evaluating the addition of panitumumab to FOLFOX chemotherapy where those with a RAS mutation had a shorter overall survival when treated with panitumumab compared to FOLFOX alone (15.6 v 19.2 months; HR=1.25, p=0.034). This illustrates how new paradigms are emerging where an agent is introduced based upon median effect and as tumour heterogeneity in respect to same becomes understood the population whom benefit from a treatment is refined. Indeed, there may be a role for such a process to examine the differential effects of chemotherapy by molecular sub-groups.
The growth in knowledge of the wide degree of tumour heterogeneity has resulted in a consensus group presenting a new paradigm for sub-types of colon cancer based on recurrent signals based on clinical, molecular, and pathway traits. The next step is to assess predictive value and differential drug sensitivity across these sub-types. Furthermore, addressing this complexity will require a new approach to clinical trial design.
Josep Tabernero continued the theme of molecular heterogeneity focussing on gene expression data and drug development. In the late 1990’s Vogelstein described the accumulation of genomic aberrations in colonic mucosa. From this starting point two patterns have been described: 1. Chromosomal instability and 2. MSI. MSI-high tumours may arise either from genetic predisposition or sporadically due to hypermethylation of the genes. Other molecular sub-types defined include CpG-island methylator phenotype and stable versus less stable karyotypes. In addition, there are two broad groups of tumours at the molecular level hyper-mutated and non-hyper-mutated. There are some genes that are activated in the majority of tumours eg Wnt. Globally, genetic abnormalities are consistent between primary and metastatic sites. Changes in balance of abnormalities can be observed to occur under the stress of treatment such as EGFR inhibitors. Emerging technologies may enable us to monitor such changes using circulating tumour DNA. The clinical benefits of such molecular are being explored in many early phase clinical trials. One such area of research has been in B-raf mutant colorectal where the raf inhibitor vemurafenib as a single agent was ineffective in colon cancer. Further laboratory research indicated that this was due to upregulation of EGFR in B-raf mutant tumour. Early phase clinical trials, presented in the poster highlights session, demonstrated significant promise for more effective treatment of this tumour sub-type.
In a talk that was focussed on the scientific background of this tumour heterogeneity Charles Swanton described not just the inter-tumour heterogeneity discussed by the previous speakers but also intra-tumour heterogeneity and intercellular heterogeneity. This arises in cancer, including colorectal cancer, due to branched evolution resulting in tumour heterogeneity. Heterogeneity supports evolutionary fitness and may be associated with poor clinical outcome. Consequently, predictive markers biopsies for different areas may give different results. Thus, to achieve maximal tumour control, we need to identify and target the trunkal drivers that are present in all clones and sub-clones of a tumour. Branched genetic events are present in some, but not all, cancer cells and occur dynamically during the course of the disease. Targeted therapy may prune some branches at the expense of others and be the driver for drug resistance. There are multiple routes to tumour heterogeneity in colorectal cancer. DNA repair and replication defects are the central causes of genome instability including: mismatch repair defects, base excision repair defects, abnormalities in polymerase epsilon and delta and chromosomal instability. One route to chromosomal instability may be through replication stress. Genome doubling (tetraploidy) is an important event that may accelerate cancer evolution and genomic diversification. In terms of drug development this means that it is important to understand Darwinian evolutionary pressure for cancer and have trial designs that target the trunkal and branch drivers. Ultimately, we need to seek drugs that may limit diversification.
In conclusion this session demonstrated the huge strides that have been made in the understanding of the chromosomal and genomic understanding of colorectal cancer. These advances pose new opportunities for treatment. However, current trial design demonstrates the median effect of a treatment in a population. Subsequent biomarker analysis will identify further sub-groups whom are actually benefitting from a treatment. Consequently, we will need to increasingly use adaptive trial designs to make further progress in the treatment of colorectal cancer.
